basic.py 178 KB
Newer Older
wxchan's avatar
wxchan committed
1
# coding: utf-8
2
"""Wrapper for C API of LightGBM."""
3
import abc
wxchan's avatar
wxchan committed
4
import ctypes
5
import json
wxchan's avatar
wxchan committed
6
import warnings
7
from collections import OrderedDict
8
from copy import deepcopy
9
from enum import Enum
10
from functools import wraps
11
from os import SEEK_END, environ
12
13
from os.path import getsize
from pathlib import Path
14
from tempfile import NamedTemporaryFile
15
from typing import Any, Callable, Dict, Iterable, List, Optional, Set, Tuple, Union
wxchan's avatar
wxchan committed
16
17
18
19

import numpy as np
import scipy.sparse

20
from .compat import PANDAS_INSTALLED, concat, dt_DataTable, pd_CategoricalDtype, pd_DataFrame, pd_Series
wxchan's avatar
wxchan committed
21
22
from .libpath import find_lib_path

23
24
25
26
27
28
29
30
31
__all__ = [
    'Booster',
    'Dataset',
    'LGBMDeprecationWarning',
    'LightGBMError',
    'register_logger',
    'Sequence',
]

32
_DatasetHandle = ctypes.c_void_p
33
_LGBM_EvalFunctionResultType = Tuple[str, float, bool]
34
_LGBM_BoosterBestScoreType = Dict[str, Dict[str, float]]
35
_LGBM_BoosterEvalMethodResultType = Tuple[str, str, float, bool]
36
37
_LGBM_CategoricalFeatureConfiguration = Union[List[str], List[int], str]
_LGBM_FeatureNameConfiguration = Union[List[str], str]
38
39
40
41
42
43
_LGBM_LabelType = Union[
    list,
    np.ndarray,
    pd_Series,
    pd_DataFrame
]
44

45
46
47
ZERO_THRESHOLD = 1e-35


48
49
50
51
def _is_zero(x: float) -> bool:
    return -ZERO_THRESHOLD <= x <= ZERO_THRESHOLD


52
def _get_sample_count(total_nrow: int, params: str) -> int:
53
54
55
    sample_cnt = ctypes.c_int(0)
    _safe_call(_LIB.LGBM_GetSampleCount(
        ctypes.c_int32(total_nrow),
56
        _c_str(params),
57
58
59
60
        ctypes.byref(sample_cnt),
    ))
    return sample_cnt.value

wxchan's avatar
wxchan committed
61

62
63
64
65
66
67
class _MissingType(Enum):
    NONE = 'None'
    NAN = 'NaN'
    ZERO = 'Zero'


68
class _DummyLogger:
69
    def info(self, msg: str) -> None:
70
71
        print(msg)

72
    def warning(self, msg: str) -> None:
73
74
75
        warnings.warn(msg, stacklevel=3)


76
77
78
_LOGGER: Any = _DummyLogger()
_INFO_METHOD_NAME = "info"
_WARNING_METHOD_NAME = "warning"
79
80


81
82
83
84
def _has_method(logger: Any, method_name: str) -> bool:
    return callable(getattr(logger, method_name, None))


85
86
87
def register_logger(
    logger: Any, info_method_name: str = "info", warning_method_name: str = "warning"
) -> None:
88
89
90
91
    """Register custom logger.

    Parameters
    ----------
92
    logger : Any
93
        Custom logger.
94
95
96
97
    info_method_name : str, optional (default="info")
        Method used to log info messages.
    warning_method_name : str, optional (default="warning")
        Method used to log warning messages.
98
    """
99
100
101
102
103
104
    if not _has_method(logger, info_method_name) or not _has_method(logger, warning_method_name):
        raise TypeError(
            f"Logger must provide '{info_method_name}' and '{warning_method_name}' method"
        )

    global _LOGGER, _INFO_METHOD_NAME, _WARNING_METHOD_NAME
105
    _LOGGER = logger
106
107
    _INFO_METHOD_NAME = info_method_name
    _WARNING_METHOD_NAME = warning_method_name
108
109


110
def _normalize_native_string(func: Callable[[str], None]) -> Callable[[str], None]:
111
    """Join log messages from native library which come by chunks."""
112
    msg_normalized: List[str] = []
113
114

    @wraps(func)
115
    def wrapper(msg: str) -> None:
116
117
118
119
120
121
122
123
124
125
126
        nonlocal msg_normalized
        if msg.strip() == '':
            msg = ''.join(msg_normalized)
            msg_normalized = []
            return func(msg)
        else:
            msg_normalized.append(msg)

    return wrapper


127
def _log_info(msg: str) -> None:
128
    getattr(_LOGGER, _INFO_METHOD_NAME)(msg)
129
130


131
def _log_warning(msg: str) -> None:
132
    getattr(_LOGGER, _WARNING_METHOD_NAME)(msg)
133
134
135


@_normalize_native_string
136
def _log_native(msg: str) -> None:
137
    getattr(_LOGGER, _INFO_METHOD_NAME)(msg)
138
139


140
def _log_callback(msg: bytes) -> None:
141
    """Redirect logs from native library into Python."""
142
    _log_native(str(msg.decode('utf-8')))
143
144


145
def _load_lib() -> ctypes.CDLL:
146
    """Load LightGBM library."""
wxchan's avatar
wxchan committed
147
148
149
    lib_path = find_lib_path()
    lib = ctypes.cdll.LoadLibrary(lib_path[0])
    lib.LGBM_GetLastError.restype = ctypes.c_char_p
150
    callback = ctypes.CFUNCTYPE(None, ctypes.c_char_p)
151
    lib.callback = callback(_log_callback)  # type: ignore[attr-defined]
152
    if lib.LGBM_RegisterLogCallback(lib.callback) != 0:
153
        raise LightGBMError(lib.LGBM_GetLastError().decode('utf-8'))
wxchan's avatar
wxchan committed
154
155
    return lib

wxchan's avatar
wxchan committed
156

157
158
159
160
161
162
163
# we don't need lib_lightgbm while building docs
_LIB: ctypes.CDLL
if environ.get('LIGHTGBM_BUILD_DOC', False):
    from unittest.mock import Mock  # isort: skip
    _LIB = Mock(ctypes.CDLL)  # type: ignore
else:
    _LIB = _load_lib()
wxchan's avatar
wxchan committed
164

wxchan's avatar
wxchan committed
165

166
_NUMERIC_TYPES = (int, float, bool)
167
_ArrayLike = Union[List, np.ndarray, pd_Series]
168
169


170
def _safe_call(ret: int) -> None:
171
172
    """Check the return value from C API call.

wxchan's avatar
wxchan committed
173
174
175
    Parameters
    ----------
    ret : int
176
        The return value from C API calls.
wxchan's avatar
wxchan committed
177
178
    """
    if ret != 0:
179
        raise LightGBMError(_LIB.LGBM_GetLastError().decode('utf-8'))
wxchan's avatar
wxchan committed
180

wxchan's avatar
wxchan committed
181

182
def _is_numeric(obj: Any) -> bool:
183
    """Check whether object is a number or not, include numpy number, etc."""
wxchan's avatar
wxchan committed
184
185
186
    try:
        float(obj)
        return True
wxchan's avatar
wxchan committed
187
188
189
    except (TypeError, ValueError):
        # TypeError: obj is not a string or a number
        # ValueError: invalid literal
wxchan's avatar
wxchan committed
190
191
        return False

wxchan's avatar
wxchan committed
192

193
def _is_numpy_1d_array(data: Any) -> bool:
194
    """Check whether data is a numpy 1-D array."""
195
    return isinstance(data, np.ndarray) and len(data.shape) == 1
wxchan's avatar
wxchan committed
196

wxchan's avatar
wxchan committed
197

198
def _is_numpy_column_array(data: Any) -> bool:
199
200
201
202
203
204
205
    """Check whether data is a column numpy array."""
    if not isinstance(data, np.ndarray):
        return False
    shape = data.shape
    return len(shape) == 2 and shape[1] == 1


206
def _cast_numpy_array_to_dtype(array: np.ndarray, dtype: np.dtype) -> np.ndarray:
207
    """Cast numpy array to given dtype."""
208
209
210
211
212
    if array.dtype == dtype:
        return array
    return array.astype(dtype=dtype, copy=False)


213
def _is_1d_list(data: Any) -> bool:
214
    """Check whether data is a 1-D list."""
215
    return isinstance(data, list) and (not data or _is_numeric(data[0]))
wxchan's avatar
wxchan committed
216

wxchan's avatar
wxchan committed
217

218
219
220
def _is_1d_collection(data: Any) -> bool:
    """Check whether data is a 1-D collection."""
    return (
221
        _is_numpy_1d_array(data)
222
        or _is_numpy_column_array(data)
223
        or _is_1d_list(data)
224
225
226
227
        or isinstance(data, pd_Series)
    )


228
def _list_to_1d_numpy(data, dtype=np.float32, name='list'):
229
    """Convert data to numpy 1-D array."""
230
    if _is_numpy_1d_array(data):
231
        return _cast_numpy_array_to_dtype(data, dtype)
232
    elif _is_numpy_column_array(data):
233
234
        _log_warning('Converting column-vector to 1d array')
        array = data.ravel()
235
        return _cast_numpy_array_to_dtype(array, dtype)
236
    elif _is_1d_list(data):
wxchan's avatar
wxchan committed
237
        return np.array(data, dtype=dtype, copy=False)
238
    elif isinstance(data, pd_Series):
239
        _check_for_bad_pandas_dtypes(data.to_frame().dtypes)
240
        return np.array(data, dtype=dtype, copy=False)  # SparseArray should be supported as well
wxchan's avatar
wxchan committed
241
    else:
242
243
        raise TypeError(f"Wrong type({type(data).__name__}) for {name}.\n"
                        "It should be list, numpy 1-D array or pandas Series")
wxchan's avatar
wxchan committed
244

wxchan's avatar
wxchan committed
245

246
247
248
249
250
251
252
def _is_numpy_2d_array(data: Any) -> bool:
    """Check whether data is a numpy 2-D array."""
    return isinstance(data, np.ndarray) and len(data.shape) == 2 and data.shape[1] > 1


def _is_2d_list(data: Any) -> bool:
    """Check whether data is a 2-D list."""
253
    return isinstance(data, list) and len(data) > 0 and _is_1d_list(data[0])
254
255
256
257
258
259
260
261
262
263
264
265
266
267


def _is_2d_collection(data: Any) -> bool:
    """Check whether data is a 2-D collection."""
    return (
        _is_numpy_2d_array(data)
        or _is_2d_list(data)
        or isinstance(data, pd_DataFrame)
    )


def _data_to_2d_numpy(data: Any, dtype: type = np.float32, name: str = 'list') -> np.ndarray:
    """Convert data to numpy 2-D array."""
    if _is_numpy_2d_array(data):
268
        return _cast_numpy_array_to_dtype(data, dtype)
269
270
271
    if _is_2d_list(data):
        return np.array(data, dtype=dtype)
    if isinstance(data, pd_DataFrame):
272
        _check_for_bad_pandas_dtypes(data.dtypes)
273
        return _cast_numpy_array_to_dtype(data.values, dtype)
274
275
276
277
    raise TypeError(f"Wrong type({type(data).__name__}) for {name}.\n"
                    "It should be list of lists, numpy 2-D array or pandas DataFrame")


278
def _cfloat32_array_to_numpy(cptr: "ctypes._Pointer", length: int) -> np.ndarray:
279
    """Convert a ctypes float pointer array to a numpy array."""
wxchan's avatar
wxchan committed
280
    if isinstance(cptr, ctypes.POINTER(ctypes.c_float)):
281
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
wxchan's avatar
wxchan committed
282
    else:
283
        raise RuntimeError('Expected float pointer')
wxchan's avatar
wxchan committed
284

Guolin Ke's avatar
Guolin Ke committed
285

286
def _cfloat64_array_to_numpy(cptr: "ctypes._Pointer", length: int) -> np.ndarray:
287
    """Convert a ctypes double pointer array to a numpy array."""
Guolin Ke's avatar
Guolin Ke committed
288
    if isinstance(cptr, ctypes.POINTER(ctypes.c_double)):
289
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
Guolin Ke's avatar
Guolin Ke committed
290
291
292
    else:
        raise RuntimeError('Expected double pointer')

wxchan's avatar
wxchan committed
293

294
def _cint32_array_to_numpy(cptr: "ctypes._Pointer", length: int) -> np.ndarray:
295
    """Convert a ctypes int pointer array to a numpy array."""
wxchan's avatar
wxchan committed
296
    if isinstance(cptr, ctypes.POINTER(ctypes.c_int32)):
297
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
wxchan's avatar
wxchan committed
298
    else:
299
300
301
        raise RuntimeError('Expected int32 pointer')


302
def _cint64_array_to_numpy(cptr: "ctypes._Pointer", length: int) -> np.ndarray:
303
304
    """Convert a ctypes int pointer array to a numpy array."""
    if isinstance(cptr, ctypes.POINTER(ctypes.c_int64)):
305
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
306
307
    else:
        raise RuntimeError('Expected int64 pointer')
wxchan's avatar
wxchan committed
308

wxchan's avatar
wxchan committed
309

310
def _c_str(string: str) -> ctypes.c_char_p:
311
    """Convert a Python string to C string."""
wxchan's avatar
wxchan committed
312
313
    return ctypes.c_char_p(string.encode('utf-8'))

wxchan's avatar
wxchan committed
314

315
def _c_array(ctype: type, values: List[Any]) -> ctypes.Array:
316
    """Convert a Python array to C array."""
317
    return (ctype * len(values))(*values)  # type: ignore[operator]
wxchan's avatar
wxchan committed
318

wxchan's avatar
wxchan committed
319

320
def _json_default_with_numpy(obj: Any) -> Any:
321
322
323
324
325
326
327
328
329
    """Convert numpy classes to JSON serializable objects."""
    if isinstance(obj, (np.integer, np.floating, np.bool_)):
        return obj.item()
    elif isinstance(obj, np.ndarray):
        return obj.tolist()
    else:
        return obj


330
331
332
333
334
335
336
337
def _to_string(x: Union[int, float, str, List]) -> str:
    if isinstance(x, list):
        val_list = ",".join(str(val) for val in x)
        return f"[{val_list}]"
    else:
        return str(x)


338
def _param_dict_to_str(data: Optional[Dict[str, Any]]) -> str:
339
    """Convert Python dictionary to string, which is passed to C API."""
340
    if data is None or not data:
wxchan's avatar
wxchan committed
341
342
343
        return ""
    pairs = []
    for key, val in data.items():
344
        if isinstance(val, (list, tuple, set)) or _is_numpy_1d_array(val):
345
            pairs.append(f"{key}={','.join(map(_to_string, val))}")
346
        elif isinstance(val, (str, Path, _NUMERIC_TYPES)) or _is_numeric(val):
347
            pairs.append(f"{key}={val}")
348
        elif val is not None:
349
            raise TypeError(f'Unknown type of parameter:{key}, got:{type(val).__name__}')
wxchan's avatar
wxchan committed
350
    return ' '.join(pairs)
351

wxchan's avatar
wxchan committed
352

353
class _TempFile:
354
355
    """Proxy class to workaround errors on Windows."""

356
357
358
    def __enter__(self):
        with NamedTemporaryFile(prefix="lightgbm_tmp_", delete=True) as f:
            self.name = f.name
359
            self.path = Path(self.name)
360
        return self
wxchan's avatar
wxchan committed
361

362
    def __exit__(self, exc_type, exc_val, exc_tb):
363
364
        if self.path.is_file():
            self.path.unlink()
365

wxchan's avatar
wxchan committed
366

367
class LightGBMError(Exception):
368
369
    """Error thrown by LightGBM."""

370
371
372
    pass


373
374
375
376
377
378
379
380
# DeprecationWarning is not shown by default, so let's create our own with higher level
class LGBMDeprecationWarning(UserWarning):
    """Custom deprecation warning."""

    pass


class _ConfigAliases:
381
382
383
384
    # lazy evaluation to allow import without dynamic library, e.g., for docs generation
    aliases = None

    @staticmethod
385
    def _get_all_param_aliases() -> Dict[str, List[str]]:
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
        buffer_len = 1 << 20
        tmp_out_len = ctypes.c_int64(0)
        string_buffer = ctypes.create_string_buffer(buffer_len)
        ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
        _safe_call(_LIB.LGBM_DumpParamAliases(
            ctypes.c_int64(buffer_len),
            ctypes.byref(tmp_out_len),
            ptr_string_buffer))
        actual_len = tmp_out_len.value
        # if buffer length is not long enough, re-allocate a buffer
        if actual_len > buffer_len:
            string_buffer = ctypes.create_string_buffer(actual_len)
            ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
            _safe_call(_LIB.LGBM_DumpParamAliases(
                ctypes.c_int64(actual_len),
                ctypes.byref(tmp_out_len),
                ptr_string_buffer))
        aliases = json.loads(
            string_buffer.value.decode('utf-8'),
405
            object_hook=lambda obj: {k: [k] + v for k, v in obj.items()}
406
407
        )
        return aliases
408
409

    @classmethod
410
411
412
    def get(cls, *args) -> Set[str]:
        if cls.aliases is None:
            cls.aliases = cls._get_all_param_aliases()
413
414
        ret = set()
        for i in args:
415
            ret.update(cls.get_sorted(i))
416
417
        return ret

418
419
420
421
422
423
    @classmethod
    def get_sorted(cls, name: str) -> List[str]:
        if cls.aliases is None:
            cls.aliases = cls._get_all_param_aliases()
        return cls.aliases.get(name, [name])

424
    @classmethod
425
426
427
    def get_by_alias(cls, *args) -> Set[str]:
        if cls.aliases is None:
            cls.aliases = cls._get_all_param_aliases()
428
429
430
431
        ret = set(args)
        for arg in args:
            for aliases in cls.aliases.values():
                if arg in aliases:
432
                    ret.update(aliases)
433
434
435
                    break
        return ret

436

437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
def _choose_param_value(main_param_name: str, params: Dict[str, Any], default_value: Any) -> Dict[str, Any]:
    """Get a single parameter value, accounting for aliases.

    Parameters
    ----------
    main_param_name : str
        Name of the main parameter to get a value for. One of the keys of ``_ConfigAliases``.
    params : dict
        Dictionary of LightGBM parameters.
    default_value : Any
        Default value to use for the parameter, if none is found in ``params``.

    Returns
    -------
    params : dict
        A ``params`` dict with exactly one value for ``main_param_name``, and all aliases ``main_param_name`` removed.
        If both ``main_param_name`` and one or more aliases for it are found, the value of ``main_param_name`` will be preferred.
    """
    # avoid side effects on passed-in parameters
    params = deepcopy(params)

458
459
    aliases = _ConfigAliases.get_sorted(main_param_name)
    aliases = [a for a in aliases if a != main_param_name]
460
461

    # if main_param_name was provided, keep that value and remove all aliases
462
    if main_param_name in params.keys():
463
464
465
        for param in aliases:
            params.pop(param, None)
        return params
466

467
468
469
470
471
    # if main param name was not found, search for an alias
    for param in aliases:
        if param in params.keys():
            params[main_param_name] = params[param]
            break
472

473
474
475
476
477
478
479
    if main_param_name in params.keys():
        for param in aliases:
            params.pop(param, None)
        return params

    # neither of main_param_name, aliases were found
    params[main_param_name] = default_value
480
481
482
483

    return params


484
_MAX_INT32 = (1 << 31) - 1
485

486
"""Macro definition of data type in C API of LightGBM"""
487
488
489
490
_C_API_DTYPE_FLOAT32 = 0
_C_API_DTYPE_FLOAT64 = 1
_C_API_DTYPE_INT32 = 2
_C_API_DTYPE_INT64 = 3
Guolin Ke's avatar
Guolin Ke committed
491

492
"""Matrix is row major in Python"""
493
_C_API_IS_ROW_MAJOR = 1
wxchan's avatar
wxchan committed
494

495
"""Macro definition of prediction type in C API of LightGBM"""
496
497
498
499
_C_API_PREDICT_NORMAL = 0
_C_API_PREDICT_RAW_SCORE = 1
_C_API_PREDICT_LEAF_INDEX = 2
_C_API_PREDICT_CONTRIB = 3
wxchan's avatar
wxchan committed
500

501
"""Macro definition of sparse matrix type"""
502
503
_C_API_MATRIX_TYPE_CSR = 0
_C_API_MATRIX_TYPE_CSC = 1
504

505
"""Macro definition of feature importance type"""
506
507
_C_API_FEATURE_IMPORTANCE_SPLIT = 0
_C_API_FEATURE_IMPORTANCE_GAIN = 1
508

509
"""Data type of data field"""
510
511
512
513
514
515
_FIELD_TYPE_MAPPER = {
    "label": _C_API_DTYPE_FLOAT32,
    "weight": _C_API_DTYPE_FLOAT32,
    "init_score": _C_API_DTYPE_FLOAT64,
    "group": _C_API_DTYPE_INT32
}
wxchan's avatar
wxchan committed
516

517
"""String name to int feature importance type mapper"""
518
519
520
521
_FEATURE_IMPORTANCE_TYPE_MAPPER = {
    "split": _C_API_FEATURE_IMPORTANCE_SPLIT,
    "gain": _C_API_FEATURE_IMPORTANCE_GAIN
}
522

wxchan's avatar
wxchan committed
523

524
def _convert_from_sliced_object(data: np.ndarray) -> np.ndarray:
525
    """Fix the memory of multi-dimensional sliced object."""
526
    if isinstance(data, np.ndarray) and isinstance(data.base, np.ndarray):
527
        if not data.flags.c_contiguous:
528
529
            _log_warning("Usage of np.ndarray subset (sliced data) is not recommended "
                         "due to it will double the peak memory cost in LightGBM.")
530
531
532
533
            return np.copy(data)
    return data


534
def _c_float_array(data):
535
    """Get pointer of float numpy array / list."""
536
    if _is_1d_list(data):
wxchan's avatar
wxchan committed
537
        data = np.array(data, copy=False)
538
    if _is_numpy_1d_array(data):
539
        data = _convert_from_sliced_object(data)
540
        assert data.flags.c_contiguous
wxchan's avatar
wxchan committed
541
542
        if data.dtype == np.float32:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_float))
543
            type_data = _C_API_DTYPE_FLOAT32
wxchan's avatar
wxchan committed
544
545
        elif data.dtype == np.float64:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_double))
546
            type_data = _C_API_DTYPE_FLOAT64
wxchan's avatar
wxchan committed
547
        else:
548
            raise TypeError(f"Expected np.float32 or np.float64, met type({data.dtype})")
wxchan's avatar
wxchan committed
549
    else:
550
        raise TypeError(f"Unknown type({type(data).__name__})")
551
    return (ptr_data, type_data, data)  # return `data` to avoid the temporary copy is freed
wxchan's avatar
wxchan committed
552

wxchan's avatar
wxchan committed
553

554
def _c_int_array(data):
555
    """Get pointer of int numpy array / list."""
556
    if _is_1d_list(data):
wxchan's avatar
wxchan committed
557
        data = np.array(data, copy=False)
558
    if _is_numpy_1d_array(data):
559
        data = _convert_from_sliced_object(data)
560
        assert data.flags.c_contiguous
wxchan's avatar
wxchan committed
561
562
        if data.dtype == np.int32:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_int32))
563
            type_data = _C_API_DTYPE_INT32
wxchan's avatar
wxchan committed
564
565
        elif data.dtype == np.int64:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_int64))
566
            type_data = _C_API_DTYPE_INT64
wxchan's avatar
wxchan committed
567
        else:
568
            raise TypeError(f"Expected np.int32 or np.int64, met type({data.dtype})")
wxchan's avatar
wxchan committed
569
    else:
570
        raise TypeError(f"Unknown type({type(data).__name__})")
571
    return (ptr_data, type_data, data)  # return `data` to avoid the temporary copy is freed
wxchan's avatar
wxchan committed
572

wxchan's avatar
wxchan committed
573

574
def _is_allowed_numpy_dtype(dtype) -> bool:
575
    float128 = getattr(np, 'float128', type(None))
576
577
578
579
    return (
        issubclass(dtype, (np.integer, np.floating, np.bool_))
        and not issubclass(dtype, (np.timedelta64, float128))
    )
580
581


582
def _check_for_bad_pandas_dtypes(pandas_dtypes_series: pd_Series) -> None:
583
584
    bad_pandas_dtypes = [
        f'{column_name}: {pandas_dtype}'
585
        for column_name, pandas_dtype in pandas_dtypes_series.items()
586
        if not _is_allowed_numpy_dtype(pandas_dtype.type)
587
588
589
590
    ]
    if bad_pandas_dtypes:
        raise ValueError('pandas dtypes must be int, float or bool.\n'
                         f'Fields with bad pandas dtypes: {", ".join(bad_pandas_dtypes)}')
591
592


593
594
595
596
597
598
def _data_from_pandas(
    data,
    feature_name: Optional[_LGBM_FeatureNameConfiguration],
    categorical_feature: Optional[_LGBM_CategoricalFeatureConfiguration],
    pandas_categorical: Optional[List[List]]
):
599
    if isinstance(data, pd_DataFrame):
600
601
        if len(data.shape) != 2 or data.shape[0] < 1:
            raise ValueError('Input data must be 2 dimensional and non empty.')
602
        if feature_name == 'auto' or feature_name is None:
603
            data = data.rename(columns=str, copy=False)
604
        cat_cols = [col for col, dtype in zip(data.columns, data.dtypes) if isinstance(dtype, pd_CategoricalDtype)]
605
        cat_cols_not_ordered = [col for col in cat_cols if not data[col].cat.ordered]
606
607
608
609
610
        if pandas_categorical is None:  # train dataset
            pandas_categorical = [list(data[col].cat.categories) for col in cat_cols]
        else:
            if len(cat_cols) != len(pandas_categorical):
                raise ValueError('train and valid dataset categorical_feature do not match.')
611
            for col, category in zip(cat_cols, pandas_categorical):
612
613
                if list(data[col].cat.categories) != list(category):
                    data[col] = data[col].cat.set_categories(category)
614
        if len(cat_cols):  # cat_cols is list
615
            data = data.copy(deep=False)  # not alter origin DataFrame
616
            data[cat_cols] = data[cat_cols].apply(lambda x: x.cat.codes).replace({-1: np.nan})
617
618
619
        if categorical_feature is not None:
            if feature_name is None:
                feature_name = list(data.columns)
620
            if categorical_feature == 'auto':  # use cat cols from DataFrame
621
                categorical_feature = cat_cols_not_ordered
622
623
            else:  # use cat cols specified by user
                categorical_feature = list(categorical_feature)
624
625
        if feature_name == 'auto':
            feature_name = list(data.columns)
626
        _check_for_bad_pandas_dtypes(data.dtypes)
627
628
629
        df_dtypes = [dtype.type for dtype in data.dtypes]
        df_dtypes.append(np.float32)  # so that the target dtype considers floats
        target_dtype = np.find_common_type(df_dtypes, [])
630
631
632
633
634
635
636
637
638
639
        try:
            # most common case (no nullable dtypes)
            data = data.to_numpy(dtype=target_dtype, copy=False)
        except TypeError:
            # 1.0 <= pd version < 1.1 and nullable dtypes, least common case
            # raises error because array is casted to type(pd.NA) and there's no na_value argument
            data = data.astype(target_dtype, copy=False).values
        except ValueError:
            # data has nullable dtypes, but we can specify na_value argument and copy will be made
            data = data.to_numpy(dtype=target_dtype, na_value=np.nan)
640
641
642
643
644
645
    else:
        if feature_name == 'auto':
            feature_name = None
        if categorical_feature == 'auto':
            categorical_feature = None
    return data, feature_name, categorical_feature, pandas_categorical
646
647


648
649
650
651
def _dump_pandas_categorical(
    pandas_categorical: Optional[List[List]],
    file_name: Optional[Union[str, Path]] = None
) -> str:
652
    categorical_json = json.dumps(pandas_categorical, default=_json_default_with_numpy)
653
    pandas_str = f'\npandas_categorical:{categorical_json}\n'
654
655
656
657
658
659
    if file_name is not None:
        with open(file_name, 'a') as f:
            f.write(pandas_str)
    return pandas_str


660
661
662
def _load_pandas_categorical(
    file_name: Optional[Union[str, Path]] = None,
    model_str: Optional[str] = None
663
) -> Optional[List[List]]:
664
665
    pandas_key = 'pandas_categorical:'
    offset = -len(pandas_key)
666
    if file_name is not None:
667
        max_offset = -getsize(file_name)
668
669
670
671
        with open(file_name, 'rb') as f:
            while True:
                if offset < max_offset:
                    offset = max_offset
672
                f.seek(offset, SEEK_END)
673
674
675
676
                lines = f.readlines()
                if len(lines) >= 2:
                    break
                offset *= 2
677
        last_line = lines[-1].decode('utf-8').strip()
678
        if not last_line.startswith(pandas_key):
679
            last_line = lines[-2].decode('utf-8').strip()
680
    elif model_str is not None:
681
682
683
684
685
686
        idx = model_str.rfind('\n', 0, offset)
        last_line = model_str[idx:].strip()
    if last_line.startswith(pandas_key):
        return json.loads(last_line[len(pandas_key):])
    else:
        return None
687
688


689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
class Sequence(abc.ABC):
    """
    Generic data access interface.

    Object should support the following operations:

    .. code-block::

        # Get total row number.
        >>> len(seq)
        # Random access by row index. Used for data sampling.
        >>> seq[10]
        # Range data access. Used to read data in batch when constructing Dataset.
        >>> seq[0:100]
        # Optionally specify batch_size to control range data read size.
        >>> seq.batch_size

    - With random access, **data sampling does not need to go through all data**.
    - With range data access, there's **no need to read all data into memory thus reduce memory usage**.

709
710
    .. versionadded:: 3.3.0

711
712
713
714
715
716
717
718
719
    Attributes
    ----------
    batch_size : int
        Default size of a batch.
    """

    batch_size = 4096  # Defaults to read 4K rows in each batch.

    @abc.abstractmethod
720
    def __getitem__(self, idx: Union[int, slice, List[int]]) -> np.ndarray:
721
722
723
724
725
726
727
        """Return data for given row index.

        A basic implementation should look like this:

        .. code-block:: python

            if isinstance(idx, numbers.Integral):
728
                return self._get_one_line(idx)
729
            elif isinstance(idx, slice):
730
731
                return np.stack([self._get_one_line(i) for i in range(idx.start, idx.stop)])
            elif isinstance(idx, list):
732
                # Only required if using ``Dataset.subset()``.
733
                return np.array([self._get_one_line(i) for i in idx])
734
            else:
735
                raise TypeError(f"Sequence index must be integer, slice or list, got {type(idx).__name__}")
736
737
738

        Parameters
        ----------
739
        idx : int, slice[int], list[int]
740
741
742
743
            Item index.

        Returns
        -------
744
        result : numpy 1-D array or numpy 2-D array
745
            1-D array if idx is int, 2-D array if idx is slice or list.
746
747
748
749
750
751
752
753
754
        """
        raise NotImplementedError("Sub-classes of lightgbm.Sequence must implement __getitem__()")

    @abc.abstractmethod
    def __len__(self) -> int:
        """Return row count of this sequence."""
        raise NotImplementedError("Sub-classes of lightgbm.Sequence must implement __len__()")


755
class _InnerPredictor:
756
757
758
759
760
    """_InnerPredictor of LightGBM.

    Not exposed to user.
    Used only for prediction, usually used for continued training.

Nikita Titov's avatar
Nikita Titov committed
761
762
763
    .. note::

        Can be converted from Booster, but cannot be converted to Booster.
Guolin Ke's avatar
Guolin Ke committed
764
    """
765

766
767
768
769
770
771
    def __init__(
        self,
        model_file: Optional[Union[str, Path]] = None,
        booster_handle: Optional[ctypes.c_void_p] = None,
        pred_parameter: Optional[Dict[str, Any]] = None
    ):
772
        """Initialize the _InnerPredictor.
wxchan's avatar
wxchan committed
773
774
775

        Parameters
        ----------
776
        model_file : str, pathlib.Path or None, optional (default=None)
wxchan's avatar
wxchan committed
777
            Path to the model file.
778
779
780
        booster_handle : object or None, optional (default=None)
            Handle of Booster.
        pred_parameter: dict or None, optional (default=None)
781
            Other parameters for the prediction.
wxchan's avatar
wxchan committed
782
783
784
785
786
        """
        self.handle = ctypes.c_void_p()
        self.__is_manage_handle = True
        if model_file is not None:
            """Prediction task"""
Guolin Ke's avatar
Guolin Ke committed
787
            out_num_iterations = ctypes.c_int(0)
wxchan's avatar
wxchan committed
788
            _safe_call(_LIB.LGBM_BoosterCreateFromModelfile(
789
                _c_str(str(model_file)),
wxchan's avatar
wxchan committed
790
791
                ctypes.byref(out_num_iterations),
                ctypes.byref(self.handle)))
Guolin Ke's avatar
Guolin Ke committed
792
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
793
794
795
796
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.num_class = out_num_class.value
797
            self.num_total_iteration = out_num_iterations.value
798
            self.pandas_categorical = _load_pandas_categorical(file_name=model_file)
wxchan's avatar
wxchan committed
799
        elif booster_handle is not None:
Guolin Ke's avatar
Guolin Ke committed
800
            self.__is_manage_handle = False
wxchan's avatar
wxchan committed
801
            self.handle = booster_handle
Guolin Ke's avatar
Guolin Ke committed
802
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
803
804
805
806
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.num_class = out_num_class.value
807
            self.num_total_iteration = self.current_iteration()
808
            self.pandas_categorical = None
wxchan's avatar
wxchan committed
809
        else:
810
            raise TypeError('Need model_file or booster_handle to create a predictor')
wxchan's avatar
wxchan committed
811

812
        pred_parameter = {} if pred_parameter is None else pred_parameter
813
        self.pred_parameter = _param_dict_to_str(pred_parameter)
cbecker's avatar
cbecker committed
814

815
    def __del__(self) -> None:
816
817
818
819
820
        try:
            if self.__is_manage_handle:
                _safe_call(_LIB.LGBM_BoosterFree(self.handle))
        except AttributeError:
            pass
wxchan's avatar
wxchan committed
821

822
    def __getstate__(self) -> Dict[str, Any]:
823
824
825
826
        this = self.__dict__.copy()
        this.pop('handle', None)
        return this

827
828
829
830
831
832
833
834
835
836
837
    def predict(
        self,
        data,
        start_iteration: int = 0,
        num_iteration: int = -1,
        raw_score: bool = False,
        pred_leaf: bool = False,
        pred_contrib: bool = False,
        data_has_header: bool = False,
        validate_features: bool = False
    ):
838
        """Predict logic.
wxchan's avatar
wxchan committed
839
840
841

        Parameters
        ----------
842
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame or scipy.sparse
843
            Data source for prediction.
844
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM).
845
846
        start_iteration : int, optional (default=0)
            Start index of the iteration to predict.
847
848
849
850
851
852
853
854
855
856
857
        num_iteration : int, optional (default=-1)
            Iteration used for prediction.
        raw_score : bool, optional (default=False)
            Whether to predict raw scores.
        pred_leaf : bool, optional (default=False)
            Whether to predict leaf index.
        pred_contrib : bool, optional (default=False)
            Whether to predict feature contributions.
        data_has_header : bool, optional (default=False)
            Whether data has header.
            Used only for txt data.
858
859
860
        validate_features : bool, optional (default=False)
            If True, ensure that the features used to predict match the ones used to train.
            Used only if data is pandas DataFrame.
wxchan's avatar
wxchan committed
861
862
863

        Returns
        -------
864
        result : numpy array, scipy.sparse or list of scipy.sparse
865
            Prediction result.
866
            Can be sparse or a list of sparse objects (each element represents predictions for one class) for feature contributions (when ``pred_contrib=True``).
wxchan's avatar
wxchan committed
867
        """
wxchan's avatar
wxchan committed
868
        if isinstance(data, Dataset):
869
            raise TypeError("Cannot use Dataset instance for prediction, please use raw data instead")
870
871
872
873
874
875
876
877
878
879
880
        elif isinstance(data, pd_DataFrame) and validate_features:
            data_names = [str(x) for x in data.columns]
            ptr_names = (ctypes.c_char_p * len(data_names))()
            ptr_names[:] = [x.encode('utf-8') for x in data_names]
            _safe_call(
                _LIB.LGBM_BoosterValidateFeatureNames(
                    self.handle,
                    ptr_names,
                    ctypes.c_int(len(data_names)),
                )
            )
881
        data = _data_from_pandas(data, None, None, self.pandas_categorical)[0]
882
        predict_type = _C_API_PREDICT_NORMAL
wxchan's avatar
wxchan committed
883
        if raw_score:
884
            predict_type = _C_API_PREDICT_RAW_SCORE
wxchan's avatar
wxchan committed
885
        if pred_leaf:
886
            predict_type = _C_API_PREDICT_LEAF_INDEX
887
        if pred_contrib:
888
            predict_type = _C_API_PREDICT_CONTRIB
wxchan's avatar
wxchan committed
889
        int_data_has_header = 1 if data_has_header else 0
cbecker's avatar
cbecker committed
890

891
        if isinstance(data, (str, Path)):
892
            with _TempFile() as f:
wxchan's avatar
wxchan committed
893
894
                _safe_call(_LIB.LGBM_BoosterPredictForFile(
                    self.handle,
895
                    _c_str(str(data)),
Guolin Ke's avatar
Guolin Ke committed
896
897
                    ctypes.c_int(int_data_has_header),
                    ctypes.c_int(predict_type),
898
                    ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
899
                    ctypes.c_int(num_iteration),
900
901
                    _c_str(self.pred_parameter),
                    _c_str(f.name)))
902
903
                preds = np.loadtxt(f.name, dtype=np.float64)
                nrow = preds.shape[0]
wxchan's avatar
wxchan committed
904
        elif isinstance(data, scipy.sparse.csr_matrix):
905
906
907
908
909
910
            preds, nrow = self.__pred_for_csr(
                csr=data,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
Guolin Ke's avatar
Guolin Ke committed
911
        elif isinstance(data, scipy.sparse.csc_matrix):
912
913
914
915
916
917
            preds, nrow = self.__pred_for_csc(
                csc=data,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
wxchan's avatar
wxchan committed
918
        elif isinstance(data, np.ndarray):
919
920
921
922
923
924
            preds, nrow = self.__pred_for_np2d(
                mat=data,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
925
926
927
        elif isinstance(data, list):
            try:
                data = np.array(data)
928
            except BaseException:
929
                raise ValueError('Cannot convert data list to numpy array.')
930
931
932
933
934
935
            preds, nrow = self.__pred_for_np2d(
                mat=data,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
936
        elif isinstance(data, dt_DataTable):
937
938
939
940
941
942
            preds, nrow = self.__pred_for_np2d(
                mat=data.to_numpy(),
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
wxchan's avatar
wxchan committed
943
944
        else:
            try:
945
                _log_warning('Converting data to scipy sparse matrix.')
wxchan's avatar
wxchan committed
946
                csr = scipy.sparse.csr_matrix(data)
947
            except BaseException:
948
                raise TypeError(f'Cannot predict data for type {type(data).__name__}')
949
950
951
952
953
954
            preds, nrow = self.__pred_for_csr(
                csr=csr,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
wxchan's avatar
wxchan committed
955
956
        if pred_leaf:
            preds = preds.astype(np.int32)
957
        is_sparse = scipy.sparse.issparse(preds) or isinstance(preds, list)
958
        if not is_sparse and preds.size != nrow:
wxchan's avatar
wxchan committed
959
            if preds.size % nrow == 0:
960
                preds = preds.reshape(nrow, -1)
wxchan's avatar
wxchan committed
961
            else:
962
                raise ValueError(f'Length of predict result ({preds.size}) cannot be divide nrow ({nrow})')
wxchan's avatar
wxchan committed
963
964
        return preds

965
966
967
968
969
970
971
    def __get_num_preds(
        self,
        start_iteration: int,
        num_iteration: int,
        nrow: int,
        predict_type: int
    ) -> int:
972
        """Get size of prediction result."""
973
        if nrow > _MAX_INT32:
974
            raise LightGBMError('LightGBM cannot perform prediction for data '
975
                                f'with number of rows greater than MAX_INT32 ({_MAX_INT32}).\n'
976
                                'You can split your data into chunks '
977
                                'and then concatenate predictions for them')
Guolin Ke's avatar
Guolin Ke committed
978
979
980
        n_preds = ctypes.c_int64(0)
        _safe_call(_LIB.LGBM_BoosterCalcNumPredict(
            self.handle,
Guolin Ke's avatar
Guolin Ke committed
981
982
            ctypes.c_int(nrow),
            ctypes.c_int(predict_type),
983
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
984
            ctypes.c_int(num_iteration),
Guolin Ke's avatar
Guolin Ke committed
985
986
            ctypes.byref(n_preds)))
        return n_preds.value
wxchan's avatar
wxchan committed
987

988
989
990
991
992
993
994
995
996
997
998
999
1000
    def __inner_predict_np2d(
        self,
        mat: np.ndarray,
        start_iteration: int,
        num_iteration: int,
        predict_type: int,
        preds: Optional[np.ndarray]
    ) -> Tuple[np.ndarray, int]:
        if mat.dtype == np.float32 or mat.dtype == np.float64:
            data = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
        else:  # change non-float data to float data, need to copy
            data = np.array(mat.reshape(mat.size), dtype=np.float32)
        ptr_data, type_ptr_data, _ = _c_float_array(data)
1001
1002
1003
1004
1005
1006
        n_preds = self.__get_num_preds(
            start_iteration=start_iteration,
            num_iteration=num_iteration,
            nrow=mat.shape[0],
            predict_type=predict_type
        )
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
        if preds is None:
            preds = np.empty(n_preds, dtype=np.float64)
        elif len(preds.shape) != 1 or len(preds) != n_preds:
            raise ValueError("Wrong length of pre-allocated predict array")
        out_num_preds = ctypes.c_int64(0)
        _safe_call(_LIB.LGBM_BoosterPredictForMat(
            self.handle,
            ptr_data,
            ctypes.c_int(type_ptr_data),
            ctypes.c_int32(mat.shape[0]),
            ctypes.c_int32(mat.shape[1]),
            ctypes.c_int(_C_API_IS_ROW_MAJOR),
            ctypes.c_int(predict_type),
            ctypes.c_int(start_iteration),
            ctypes.c_int(num_iteration),
            _c_str(self.pred_parameter),
            ctypes.byref(out_num_preds),
            preds.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
        if n_preds != out_num_preds.value:
            raise ValueError("Wrong length for predict results")
        return preds, mat.shape[0]

    def __pred_for_np2d(
        self,
        mat: np.ndarray,
        start_iteration: int,
        num_iteration: int,
        predict_type: int
    ) -> Tuple[np.ndarray, int]:
1036
        """Predict for a 2-D numpy matrix."""
wxchan's avatar
wxchan committed
1037
        if len(mat.shape) != 2:
1038
            raise ValueError('Input numpy.ndarray or list must be 2 dimensional')
wxchan's avatar
wxchan committed
1039

1040
        nrow = mat.shape[0]
1041
1042
        if nrow > _MAX_INT32:
            sections = np.arange(start=_MAX_INT32, stop=nrow, step=_MAX_INT32)
1043
            # __get_num_preds() cannot work with nrow > MAX_INT32, so calculate overall number of predictions piecemeal
1044
            n_preds = [self.__get_num_preds(start_iteration, num_iteration, i, predict_type) for i in np.diff([0] + list(sections) + [nrow])]
1045
            n_preds_sections = np.array([0] + n_preds, dtype=np.intp).cumsum()
1046
            preds = np.empty(sum(n_preds), dtype=np.float64)
1047
1048
            for chunk, (start_idx_pred, end_idx_pred) in zip(np.array_split(mat, sections),
                                                             zip(n_preds_sections, n_preds_sections[1:])):
1049
                # avoid memory consumption by arrays concatenation operations
1050
1051
1052
1053
1054
1055
1056
                self.__inner_predict_np2d(
                    mat=chunk,
                    start_iteration=start_iteration,
                    num_iteration=num_iteration,
                    predict_type=predict_type,
                    preds=preds[start_idx_pred:end_idx_pred]
                )
1057
            return preds, nrow
wxchan's avatar
wxchan committed
1058
        else:
1059
1060
1061
1062
1063
1064
1065
            return self.__inner_predict_np2d(
                mat=mat,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type,
                preds=None
            )
wxchan's avatar
wxchan committed
1066

1067
1068
1069
    def __create_sparse_native(
        self,
        cs: Union[scipy.sparse.csc_matrix, scipy.sparse.csr_matrix],
1070
1071
1072
1073
1074
1075
        out_shape: np.ndarray,
        out_ptr_indptr: "ctypes._Pointer",
        out_ptr_indices: "ctypes._Pointer",
        out_ptr_data: "ctypes._Pointer",
        indptr_type: int,
        data_type: int,
1076
        is_csr: bool
1077
    ) -> Union[List[scipy.sparse.csc_matrix], List[scipy.sparse.csr_matrix]]:
1078
1079
1080
        # create numpy array from output arrays
        data_indices_len = out_shape[0]
        indptr_len = out_shape[1]
1081
        if indptr_type == _C_API_DTYPE_INT32:
1082
            out_indptr = _cint32_array_to_numpy(out_ptr_indptr, indptr_len)
1083
        elif indptr_type == _C_API_DTYPE_INT64:
1084
            out_indptr = _cint64_array_to_numpy(out_ptr_indptr, indptr_len)
1085
1086
        else:
            raise TypeError("Expected int32 or int64 type for indptr")
1087
        if data_type == _C_API_DTYPE_FLOAT32:
1088
            out_data = _cfloat32_array_to_numpy(out_ptr_data, data_indices_len)
1089
        elif data_type == _C_API_DTYPE_FLOAT64:
1090
            out_data = _cfloat64_array_to_numpy(out_ptr_data, data_indices_len)
1091
1092
        else:
            raise TypeError("Expected float32 or float64 type for data")
1093
        out_indices = _cint32_array_to_numpy(out_ptr_indices, data_indices_len)
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
        # break up indptr based on number of rows (note more than one matrix in multiclass case)
        per_class_indptr_shape = cs.indptr.shape[0]
        # for CSC there is extra column added
        if not is_csr:
            per_class_indptr_shape += 1
        out_indptr_arrays = np.split(out_indptr, out_indptr.shape[0] / per_class_indptr_shape)
        # reformat output into a csr or csc matrix or list of csr or csc matrices
        cs_output_matrices = []
        offset = 0
        for cs_indptr in out_indptr_arrays:
            matrix_indptr_len = cs_indptr[cs_indptr.shape[0] - 1]
            cs_indices = out_indices[offset + cs_indptr[0]:offset + matrix_indptr_len]
            cs_data = out_data[offset + cs_indptr[0]:offset + matrix_indptr_len]
            offset += matrix_indptr_len
            # same shape as input csr or csc matrix except extra column for expected value
            cs_shape = [cs.shape[0], cs.shape[1] + 1]
            # note: make sure we copy data as it will be deallocated next
            if is_csr:
                cs_output_matrices.append(scipy.sparse.csr_matrix((cs_data, cs_indices, cs_indptr), cs_shape))
            else:
                cs_output_matrices.append(scipy.sparse.csc_matrix((cs_data, cs_indices, cs_indptr), cs_shape))
        # free the temporary native indptr, indices, and data
        _safe_call(_LIB.LGBM_BoosterFreePredictSparse(out_ptr_indptr, out_ptr_indices, out_ptr_data,
                                                      ctypes.c_int(indptr_type), ctypes.c_int(data_type)))
        if len(cs_output_matrices) == 1:
            return cs_output_matrices[0]
        return cs_output_matrices

1122
1123
1124
1125
1126
1127
1128
1129
1130
    def __inner_predict_csr(
        self,
        csr: scipy.sparse.csr_matrix,
        start_iteration: int,
        num_iteration: int,
        predict_type: int,
        preds: Optional[np.ndarray]
    ) -> Tuple[np.ndarray, int]:
        nrow = len(csr.indptr) - 1
1131
1132
1133
1134
1135
1136
        n_preds = self.__get_num_preds(
            start_iteration=start_iteration,
            num_iteration=num_iteration,
            nrow=nrow,
            predict_type=predict_type
        )
1137
1138
1139
1140
1141
        if preds is None:
            preds = np.empty(n_preds, dtype=np.float64)
        elif len(preds.shape) != 1 or len(preds) != n_preds:
            raise ValueError("Wrong length of pre-allocated predict array")
        out_num_preds = ctypes.c_int64(0)
wxchan's avatar
wxchan committed
1142

1143
1144
        ptr_indptr, type_ptr_indptr, _ = _c_int_array(csr.indptr)
        ptr_data, type_ptr_data, _ = _c_float_array(csr.data)
1145

1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
        assert csr.shape[1] <= _MAX_INT32
        csr_indices = csr.indices.astype(np.int32, copy=False)

        _safe_call(_LIB.LGBM_BoosterPredictForCSR(
            self.handle,
            ptr_indptr,
            ctypes.c_int(type_ptr_indptr),
            csr_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
            ptr_data,
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csr.indptr)),
            ctypes.c_int64(len(csr.data)),
            ctypes.c_int64(csr.shape[1]),
            ctypes.c_int(predict_type),
            ctypes.c_int(start_iteration),
            ctypes.c_int(num_iteration),
            _c_str(self.pred_parameter),
            ctypes.byref(out_num_preds),
            preds.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
        if n_preds != out_num_preds.value:
            raise ValueError("Wrong length for predict results")
        return preds, nrow

    def __inner_predict_csr_sparse(
        self,
        csr: scipy.sparse.csr_matrix,
        start_iteration: int,
        num_iteration: int,
        predict_type: int
1175
    ) -> Tuple[Union[List[scipy.sparse.csc_matrix], List[scipy.sparse.csr_matrix]], int]:
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
        ptr_indptr, type_ptr_indptr, __ = _c_int_array(csr.indptr)
        ptr_data, type_ptr_data, _ = _c_float_array(csr.data)
        csr_indices = csr.indices.astype(np.int32, copy=False)
        matrix_type = _C_API_MATRIX_TYPE_CSR
        if type_ptr_indptr == _C_API_DTYPE_INT32:
            out_ptr_indptr = ctypes.POINTER(ctypes.c_int32)()
        else:
            out_ptr_indptr = ctypes.POINTER(ctypes.c_int64)()
        out_ptr_indices = ctypes.POINTER(ctypes.c_int32)()
        if type_ptr_data == _C_API_DTYPE_FLOAT32:
            out_ptr_data = ctypes.POINTER(ctypes.c_float)()
        else:
            out_ptr_data = ctypes.POINTER(ctypes.c_double)()
        out_shape = np.empty(2, dtype=np.int64)
        _safe_call(_LIB.LGBM_BoosterPredictSparseOutput(
            self.handle,
            ptr_indptr,
            ctypes.c_int(type_ptr_indptr),
            csr_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
            ptr_data,
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csr.indptr)),
            ctypes.c_int64(len(csr.data)),
            ctypes.c_int64(csr.shape[1]),
            ctypes.c_int(predict_type),
            ctypes.c_int(start_iteration),
            ctypes.c_int(num_iteration),
            _c_str(self.pred_parameter),
            ctypes.c_int(matrix_type),
            out_shape.ctypes.data_as(ctypes.POINTER(ctypes.c_int64)),
            ctypes.byref(out_ptr_indptr),
            ctypes.byref(out_ptr_indices),
            ctypes.byref(out_ptr_data)))
        matrices = self.__create_sparse_native(
            cs=csr,
            out_shape=out_shape,
            out_ptr_indptr=out_ptr_indptr,
            out_ptr_indices=out_ptr_indices,
            out_ptr_data=out_ptr_data,
            indptr_type=type_ptr_indptr,
            data_type=type_ptr_data,
            is_csr=True
        )
        nrow = len(csr.indptr) - 1
        return matrices, nrow

1222
1223
1224
1225
1226
1227
1228
    def __pred_for_csr(
        self,
        csr: scipy.sparse.csr_matrix,
        start_iteration: int,
        num_iteration: int,
        predict_type: int
    ) -> Tuple[np.ndarray, int]:
1229
        """Predict for a CSR data."""
1230
        if predict_type == _C_API_PREDICT_CONTRIB:
1231
1232
1233
1234
1235
1236
            return self.__inner_predict_csr_sparse(
                csr=csr,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
1237
        nrow = len(csr.indptr) - 1
1238
1239
        if nrow > _MAX_INT32:
            sections = [0] + list(np.arange(start=_MAX_INT32, stop=nrow, step=_MAX_INT32)) + [nrow]
1240
            # __get_num_preds() cannot work with nrow > MAX_INT32, so calculate overall number of predictions piecemeal
1241
            n_preds = [self.__get_num_preds(start_iteration, num_iteration, i, predict_type) for i in np.diff(sections)]
1242
            n_preds_sections = np.array([0] + n_preds, dtype=np.intp).cumsum()
1243
            preds = np.empty(sum(n_preds), dtype=np.float64)
1244
1245
            for (start_idx, end_idx), (start_idx_pred, end_idx_pred) in zip(zip(sections, sections[1:]),
                                                                            zip(n_preds_sections, n_preds_sections[1:])):
1246
                # avoid memory consumption by arrays concatenation operations
1247
1248
1249
1250
1251
1252
1253
                self.__inner_predict_csr(
                    csr=csr[start_idx:end_idx],
                    start_iteration=start_iteration,
                    num_iteration=num_iteration,
                    predict_type=predict_type,
                    preds=preds[start_idx_pred:end_idx_pred]
                )
1254
1255
            return preds, nrow
        else:
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
            return self.__inner_predict_csr(
                csr=csr,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type,
                preds=None
            )

    def __inner_predict_sparse_csc(
        self,
1266
1267
1268
1269
        csc: scipy.sparse.csc_matrix,
        start_iteration: int,
        num_iteration: int,
        predict_type: int
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
    ):
        ptr_indptr, type_ptr_indptr, __ = _c_int_array(csc.indptr)
        ptr_data, type_ptr_data, _ = _c_float_array(csc.data)
        csc_indices = csc.indices.astype(np.int32, copy=False)
        matrix_type = _C_API_MATRIX_TYPE_CSC
        if type_ptr_indptr == _C_API_DTYPE_INT32:
            out_ptr_indptr = ctypes.POINTER(ctypes.c_int32)()
        else:
            out_ptr_indptr = ctypes.POINTER(ctypes.c_int64)()
        out_ptr_indices = ctypes.POINTER(ctypes.c_int32)()
        if type_ptr_data == _C_API_DTYPE_FLOAT32:
            out_ptr_data = ctypes.POINTER(ctypes.c_float)()
        else:
            out_ptr_data = ctypes.POINTER(ctypes.c_double)()
        out_shape = np.empty(2, dtype=np.int64)
        _safe_call(_LIB.LGBM_BoosterPredictSparseOutput(
            self.handle,
            ptr_indptr,
            ctypes.c_int(type_ptr_indptr),
            csc_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
            ptr_data,
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csc.indptr)),
            ctypes.c_int64(len(csc.data)),
            ctypes.c_int64(csc.shape[0]),
            ctypes.c_int(predict_type),
            ctypes.c_int(start_iteration),
            ctypes.c_int(num_iteration),
            _c_str(self.pred_parameter),
            ctypes.c_int(matrix_type),
            out_shape.ctypes.data_as(ctypes.POINTER(ctypes.c_int64)),
            ctypes.byref(out_ptr_indptr),
            ctypes.byref(out_ptr_indices),
            ctypes.byref(out_ptr_data)))
        matrices = self.__create_sparse_native(
            cs=csc,
            out_shape=out_shape,
            out_ptr_indptr=out_ptr_indptr,
            out_ptr_indices=out_ptr_indices,
            out_ptr_data=out_ptr_data,
            indptr_type=type_ptr_indptr,
            data_type=type_ptr_data,
            is_csr=False
        )
        nrow = csc.shape[0]
        return matrices, nrow
Guolin Ke's avatar
Guolin Ke committed
1316

1317
1318
1319
1320
1321
1322
1323
    def __pred_for_csc(
        self,
        csc: scipy.sparse.csc_matrix,
        start_iteration: int,
        num_iteration: int,
        predict_type: int
    ) -> Tuple[np.ndarray, int]:
1324
        """Predict for a CSC data."""
Guolin Ke's avatar
Guolin Ke committed
1325
        nrow = csc.shape[0]
1326
        if nrow > _MAX_INT32:
1327
1328
1329
1330
1331
1332
            return self.__pred_for_csr(
                csr=csc.tocsr(),
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
1333
        if predict_type == _C_API_PREDICT_CONTRIB:
1334
1335
1336
1337
1338
1339
            return self.__inner_predict_sparse_csc(
                csc=csc,
                start_iteration=start_iteration,
                num_iteration=num_iteration,
                predict_type=predict_type
            )
1340
1341
1342
1343
1344
1345
        n_preds = self.__get_num_preds(
            start_iteration=start_iteration,
            num_iteration=num_iteration,
            nrow=nrow,
            predict_type=predict_type
        )
1346
        preds = np.empty(n_preds, dtype=np.float64)
Guolin Ke's avatar
Guolin Ke committed
1347
1348
        out_num_preds = ctypes.c_int64(0)

1349
1350
        ptr_indptr, type_ptr_indptr, __ = _c_int_array(csc.indptr)
        ptr_data, type_ptr_data, _ = _c_float_array(csc.data)
Guolin Ke's avatar
Guolin Ke committed
1351

1352
        assert csc.shape[0] <= _MAX_INT32
1353
        csc_indices = csc.indices.astype(np.int32, copy=False)
1354

Guolin Ke's avatar
Guolin Ke committed
1355
1356
1357
        _safe_call(_LIB.LGBM_BoosterPredictForCSC(
            self.handle,
            ptr_indptr,
1358
            ctypes.c_int(type_ptr_indptr),
1359
            csc_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
Guolin Ke's avatar
Guolin Ke committed
1360
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1361
1362
1363
1364
1365
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csc.indptr)),
            ctypes.c_int64(len(csc.data)),
            ctypes.c_int64(csc.shape[0]),
            ctypes.c_int(predict_type),
1366
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
1367
            ctypes.c_int(num_iteration),
1368
            _c_str(self.pred_parameter),
Guolin Ke's avatar
Guolin Ke committed
1369
            ctypes.byref(out_num_preds),
wxchan's avatar
wxchan committed
1370
            preds.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
wxchan's avatar
wxchan committed
1371
        if n_preds != out_num_preds.value:
1372
            raise ValueError("Wrong length for predict results")
wxchan's avatar
wxchan committed
1373
1374
        return preds, nrow

1375
    def current_iteration(self) -> int:
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
        """Get the index of the current iteration.

        Returns
        -------
        cur_iter : int
            The index of the current iteration.
        """
        out_cur_iter = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterGetCurrentIteration(
            self.handle,
            ctypes.byref(out_cur_iter)))
        return out_cur_iter.value

wxchan's avatar
wxchan committed
1389

1390
class Dataset:
wxchan's avatar
wxchan committed
1391
    """Dataset in LightGBM."""
1392

1393
1394
1395
    def __init__(
        self,
        data,
1396
        label: Optional[_LGBM_LabelType] = None,
1397
1398
1399
1400
        reference: Optional["Dataset"] = None,
        weight=None,
        group=None,
        init_score=None,
1401
1402
        feature_name: _LGBM_FeatureNameConfiguration = 'auto',
        categorical_feature: _LGBM_CategoricalFeatureConfiguration = 'auto',
1403
1404
1405
        params: Optional[Dict[str, Any]] = None,
        free_raw_data: bool = True
    ):
1406
        """Initialize Dataset.
1407

wxchan's avatar
wxchan committed
1408
1409
        Parameters
        ----------
1410
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame, scipy.sparse, Sequence, list of Sequence or list of numpy array
wxchan's avatar
wxchan committed
1411
            Data source of Dataset.
1412
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM) or a LightGBM Dataset binary file.
1413
        label : list, numpy 1-D array, pandas Series / one-column DataFrame or None, optional (default=None)
1414
1415
1416
            Label of the data.
        reference : Dataset or None, optional (default=None)
            If this is Dataset for validation, training data should be used as reference.
1417
        weight : list, numpy 1-D array, pandas Series or None, optional (default=None)
1418
            Weight for each instance. Weights should be non-negative.
1419
        group : list, numpy 1-D array, pandas Series or None, optional (default=None)
1420
1421
1422
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
1423
1424
            For example, if you have a 100-document dataset with ``group = [10, 20, 40, 10, 10, 10]``, that means that you have 6 groups,
            where the first 10 records are in the first group, records 11-30 are in the second group, records 31-70 are in the third group, etc.
1425
        init_score : list, list of lists (for multi-class task), numpy array, pandas Series, pandas DataFrame (for multi-class task), or None, optional (default=None)
1426
            Init score for Dataset.
1427
        feature_name : list of str, or 'auto', optional (default="auto")
1428
1429
            Feature names.
            If 'auto' and data is pandas DataFrame, data columns names are used.
1430
        categorical_feature : list of str or int, or 'auto', optional (default="auto")
1431
1432
            Categorical features.
            If list of int, interpreted as indices.
1433
            If list of str, interpreted as feature names (need to specify ``feature_name`` as well).
1434
            If 'auto' and data is pandas DataFrame, pandas unordered categorical columns are used.
1435
            All values in categorical features will be cast to int32 and thus should be less than int32 max value (2147483647).
1436
            Large values could be memory consuming. Consider using consecutive integers starting from zero.
1437
            All negative values in categorical features will be treated as missing values.
1438
            The output cannot be monotonically constrained with respect to a categorical feature.
1439
            Floating point numbers in categorical features will be rounded towards 0.
Nikita Titov's avatar
Nikita Titov committed
1440
        params : dict or None, optional (default=None)
1441
            Other parameters for Dataset.
Nikita Titov's avatar
Nikita Titov committed
1442
        free_raw_data : bool, optional (default=True)
1443
            If True, raw data is freed after constructing inner Dataset.
wxchan's avatar
wxchan committed
1444
        """
1445
        self.handle: Optional[_DatasetHandle] = None
wxchan's avatar
wxchan committed
1446
1447
1448
1449
1450
        self.data = data
        self.label = label
        self.reference = reference
        self.weight = weight
        self.group = group
1451
        self.init_score = init_score
1452
1453
        self.feature_name: _LGBM_FeatureNameConfiguration = feature_name
        self.categorical_feature: _LGBM_CategoricalFeatureConfiguration = categorical_feature
1454
        self.params = deepcopy(params)
wxchan's avatar
wxchan committed
1455
        self.free_raw_data = free_raw_data
1456
        self.used_indices: Optional[List[int]] = None
1457
        self._need_slice = True
1458
        self._predictor: Optional[_InnerPredictor] = None
1459
        self.pandas_categorical = None
1460
        self._params_back_up = None
1461
        self.version = 0
1462
        self._start_row = 0  # Used when pushing rows one by one.
wxchan's avatar
wxchan committed
1463

1464
    def __del__(self) -> None:
1465
1466
1467
1468
        try:
            self._free_handle()
        except AttributeError:
            pass
1469

1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
    def _create_sample_indices(self, total_nrow: int) -> np.ndarray:
        """Get an array of randomly chosen indices from this ``Dataset``.

        Indices are sampled without replacement.

        Parameters
        ----------
        total_nrow : int
            Total number of rows to sample from.
            If this value is greater than the value of parameter ``bin_construct_sample_cnt``, only ``bin_construct_sample_cnt`` indices will be used.
            If Dataset has multiple input data, this should be the sum of rows of every file.

        Returns
        -------
        indices : numpy array
            Indices for sampled data.
        """
1487
        param_str = _param_dict_to_str(self.get_params())
1488
1489
        sample_cnt = _get_sample_count(total_nrow, param_str)
        indices = np.empty(sample_cnt, dtype=np.int32)
1490
        ptr_data, _, _ = _c_int_array(indices)
1491
1492
1493
1494
        actual_sample_cnt = ctypes.c_int32(0)

        _safe_call(_LIB.LGBM_SampleIndices(
            ctypes.c_int32(total_nrow),
1495
            _c_str(param_str),
1496
1497
1498
            ptr_data,
            ctypes.byref(actual_sample_cnt),
        ))
1499
1500
        assert sample_cnt == actual_sample_cnt.value
        return indices
1501

1502
1503
1504
1505
1506
    def _init_from_ref_dataset(
        self,
        total_nrow: int,
        ref_dataset: _DatasetHandle
    ) -> 'Dataset':
1507
1508
1509
1510
1511
1512
        """Create dataset from a reference dataset.

        Parameters
        ----------
        total_nrow : int
            Number of rows expected to add to dataset.
1513
1514
        ref_dataset : object
            Handle of reference dataset to extract metadata from.
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539

        Returns
        -------
        self : Dataset
            Constructed Dataset object.
        """
        self.handle = ctypes.c_void_p()
        _safe_call(_LIB.LGBM_DatasetCreateByReference(
            ref_dataset,
            ctypes.c_int64(total_nrow),
            ctypes.byref(self.handle),
        ))
        return self

    def _init_from_sample(
        self,
        sample_data: List[np.ndarray],
        sample_indices: List[np.ndarray],
        sample_cnt: int,
        total_nrow: int,
    ) -> "Dataset":
        """Create Dataset from sampled data structures.

        Parameters
        ----------
1540
        sample_data : list of numpy array
1541
            Sample data for each column.
1542
        sample_indices : list of numpy array
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
            Sample data row index for each column.
        sample_cnt : int
            Number of samples.
        total_nrow : int
            Total number of rows for all input files.

        Returns
        -------
        self : Dataset
            Constructed Dataset object.
        """
        ncol = len(sample_indices)
        assert len(sample_data) == ncol, "#sample data column != #column indices"

        for i in range(ncol):
            if sample_data[i].dtype != np.double:
                raise ValueError(f"sample_data[{i}] type {sample_data[i].dtype} is not double")
            if sample_indices[i].dtype != np.int32:
                raise ValueError(f"sample_indices[{i}] type {sample_indices[i].dtype} is not int32")

        # c type: double**
        # each double* element points to start of each column of sample data.
        sample_col_ptr = (ctypes.POINTER(ctypes.c_double) * ncol)()
        # c type int**
        # each int* points to start of indices for each column
        indices_col_ptr = (ctypes.POINTER(ctypes.c_int32) * ncol)()
        for i in range(ncol):
1570
1571
            sample_col_ptr[i] = _c_float_array(sample_data[i])[0]
            indices_col_ptr[i] = _c_int_array(sample_indices[i])[0]
1572
1573

        num_per_col = np.array([len(d) for d in sample_indices], dtype=np.int32)
1574
        num_per_col_ptr, _, _ = _c_int_array(num_per_col)
1575
1576

        self.handle = ctypes.c_void_p()
1577
        params_str = _param_dict_to_str(self.get_params())
1578
1579
1580
1581
1582
1583
1584
        _safe_call(_LIB.LGBM_DatasetCreateFromSampledColumn(
            ctypes.cast(sample_col_ptr, ctypes.POINTER(ctypes.POINTER(ctypes.c_double))),
            ctypes.cast(indices_col_ptr, ctypes.POINTER(ctypes.POINTER(ctypes.c_int32))),
            ctypes.c_int32(ncol),
            num_per_col_ptr,
            ctypes.c_int32(sample_cnt),
            ctypes.c_int32(total_nrow),
1585
            ctypes.c_int64(total_nrow),
1586
            _c_str(params_str),
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
            ctypes.byref(self.handle),
        ))
        return self

    def _push_rows(self, data: np.ndarray) -> 'Dataset':
        """Add rows to Dataset.

        Parameters
        ----------
        data : numpy 1-D array
            New data to add to the Dataset.

        Returns
        -------
        self : Dataset
            Dataset object.
        """
        nrow, ncol = data.shape
        data = data.reshape(data.size)
1606
        data_ptr, data_type, _ = _c_float_array(data)
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618

        _safe_call(_LIB.LGBM_DatasetPushRows(
            self.handle,
            data_ptr,
            data_type,
            ctypes.c_int32(nrow),
            ctypes.c_int32(ncol),
            ctypes.c_int32(self._start_row),
        ))
        self._start_row += nrow
        return self

1619
    def get_params(self) -> Dict[str, Any]:
1620
1621
1622
1623
        """Get the used parameters in the Dataset.

        Returns
        -------
1624
        params : dict
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
            The used parameters in this Dataset object.
        """
        if self.params is not None:
            # no min_data, nthreads and verbose in this function
            dataset_params = _ConfigAliases.get("bin_construct_sample_cnt",
                                                "categorical_feature",
                                                "data_random_seed",
                                                "enable_bundle",
                                                "feature_pre_filter",
                                                "forcedbins_filename",
                                                "group_column",
                                                "header",
                                                "ignore_column",
                                                "is_enable_sparse",
                                                "label_column",
1640
                                                "linear_tree",
1641
1642
1643
1644
                                                "max_bin",
                                                "max_bin_by_feature",
                                                "min_data_in_bin",
                                                "pre_partition",
Nikita Titov's avatar
Nikita Titov committed
1645
                                                "precise_float_parser",
1646
1647
1648
1649
1650
                                                "two_round",
                                                "use_missing",
                                                "weight_column",
                                                "zero_as_missing")
            return {k: v for k, v in self.params.items() if k in dataset_params}
1651
1652
        else:
            return {}
1653

1654
    def _free_handle(self) -> "Dataset":
1655
        if self.handle is not None:
1656
            _safe_call(_LIB.LGBM_DatasetFree(self.handle))
1657
            self.handle = None
1658
        self._need_slice = True
Guolin Ke's avatar
Guolin Ke committed
1659
1660
        if self.used_indices is not None:
            self.data = None
Nikita Titov's avatar
Nikita Titov committed
1661
        return self
wxchan's avatar
wxchan committed
1662

1663
1664
1665
1666
1667
1668
    def _set_init_score_by_predictor(
        self,
        predictor: Optional[_InnerPredictor],
        data,
        used_indices: Optional[List[int]]
    ):
Guolin Ke's avatar
Guolin Ke committed
1669
        data_has_header = False
1670
        if isinstance(data, (str, Path)):
Guolin Ke's avatar
Guolin Ke committed
1671
            # check data has header or not
1672
            data_has_header = any(self.params.get(alias, False) for alias in _ConfigAliases.get("header"))
Guolin Ke's avatar
Guolin Ke committed
1673
        num_data = self.num_data()
1674
1675
1676
        if predictor is not None:
            init_score = predictor.predict(data,
                                           raw_score=True,
1677
1678
                                           data_has_header=data_has_header)
            init_score = init_score.ravel()
1679
            if used_indices is not None:
1680
                assert not self._need_slice
1681
                if isinstance(data, (str, Path)):
1682
                    sub_init_score = np.empty(num_data * predictor.num_class, dtype=np.float64)
1683
                    assert num_data == len(used_indices)
1684
1685
                    for i in range(len(used_indices)):
                        for j in range(predictor.num_class):
1686
1687
1688
1689
                            sub_init_score[i * predictor.num_class + j] = init_score[used_indices[i] * predictor.num_class + j]
                    init_score = sub_init_score
            if predictor.num_class > 1:
                # need to regroup init_score
1690
                new_init_score = np.empty(init_score.size, dtype=np.float64)
1691
1692
                for i in range(num_data):
                    for j in range(predictor.num_class):
1693
1694
1695
                        new_init_score[j * num_data + i] = init_score[i * predictor.num_class + j]
                init_score = new_init_score
        elif self.init_score is not None:
1696
            init_score = np.zeros(self.init_score.shape, dtype=np.float64)
1697
1698
        else:
            return self
Guolin Ke's avatar
Guolin Ke committed
1699
1700
        self.set_init_score(init_score)

1701
1702
1703
    def _lazy_init(
        self,
        data,
1704
        label: Optional[_LGBM_LabelType] = None,
1705
1706
1707
1708
1709
1710
1711
1712
1713
        reference: Optional["Dataset"] = None,
        weight=None,
        group=None,
        init_score=None,
        predictor=None,
        feature_name='auto',
        categorical_feature='auto',
        params: Optional[Dict[str, Any]] = None
    ) -> "Dataset":
wxchan's avatar
wxchan committed
1714
1715
        if data is None:
            self.handle = None
Nikita Titov's avatar
Nikita Titov committed
1716
            return self
Guolin Ke's avatar
Guolin Ke committed
1717
1718
1719
        if reference is not None:
            self.pandas_categorical = reference.pandas_categorical
            categorical_feature = reference.categorical_feature
1720
1721
1722
1723
        data, feature_name, categorical_feature, self.pandas_categorical = _data_from_pandas(data,
                                                                                             feature_name,
                                                                                             categorical_feature,
                                                                                             self.pandas_categorical)
Guolin Ke's avatar
Guolin Ke committed
1724

1725
        # process for args
wxchan's avatar
wxchan committed
1726
        params = {} if params is None else params
1727
1728
1729
        args_names = (getattr(self.__class__, '_lazy_init')
                      .__code__
                      .co_varnames[:getattr(self.__class__, '_lazy_init').__code__.co_argcount])
1730
        for key in params.keys():
1731
            if key in args_names:
1732
1733
                _log_warning(f'{key} keyword has been found in `params` and will be ignored.\n'
                             f'Please use {key} argument of the Dataset constructor to pass this parameter.')
1734
        # get categorical features
1735
1736
1737
1738
1739
1740
        if categorical_feature is not None:
            categorical_indices = set()
            feature_dict = {}
            if feature_name is not None:
                feature_dict = {name: i for i, name in enumerate(feature_name)}
            for name in categorical_feature:
1741
                if isinstance(name, str) and name in feature_dict:
1742
                    categorical_indices.add(feature_dict[name])
1743
                elif isinstance(name, int):
1744
1745
                    categorical_indices.add(name)
                else:
1746
                    raise TypeError(f"Wrong type({type(name).__name__}) or unknown name({name}) in categorical_feature")
1747
            if categorical_indices:
1748
1749
                for cat_alias in _ConfigAliases.get("categorical_feature"):
                    if cat_alias in params:
1750
                        # If the params[cat_alias] is equal to categorical_indices, do not report the warning.
1751
                        if not (isinstance(params[cat_alias], list) and set(params[cat_alias]) == categorical_indices):
1752
                            _log_warning(f'{cat_alias} in param dict is overridden.')
1753
                        params.pop(cat_alias, None)
1754
                params['categorical_column'] = sorted(categorical_indices)
1755

1756
        params_str = _param_dict_to_str(params)
1757
        self.params = params
1758
        # process for reference dataset
wxchan's avatar
wxchan committed
1759
        ref_dataset = None
wxchan's avatar
wxchan committed
1760
        if isinstance(reference, Dataset):
1761
            ref_dataset = reference.construct().handle
wxchan's avatar
wxchan committed
1762
1763
        elif reference is not None:
            raise TypeError('Reference dataset should be None or dataset instance')
1764
        # start construct data
1765
        if isinstance(data, (str, Path)):
wxchan's avatar
wxchan committed
1766
1767
            self.handle = ctypes.c_void_p()
            _safe_call(_LIB.LGBM_DatasetCreateFromFile(
1768
1769
                _c_str(str(data)),
                _c_str(params_str),
wxchan's avatar
wxchan committed
1770
1771
1772
1773
                ref_dataset,
                ctypes.byref(self.handle)))
        elif isinstance(data, scipy.sparse.csr_matrix):
            self.__init_from_csr(data, params_str, ref_dataset)
Guolin Ke's avatar
Guolin Ke committed
1774
1775
        elif isinstance(data, scipy.sparse.csc_matrix):
            self.__init_from_csc(data, params_str, ref_dataset)
wxchan's avatar
wxchan committed
1776
1777
        elif isinstance(data, np.ndarray):
            self.__init_from_np2d(data, params_str, ref_dataset)
1778
1779
1780
1781
1782
1783
1784
1785
1786
        elif isinstance(data, list) and len(data) > 0:
            if all(isinstance(x, np.ndarray) for x in data):
                self.__init_from_list_np2d(data, params_str, ref_dataset)
            elif all(isinstance(x, Sequence) for x in data):
                self.__init_from_seqs(data, ref_dataset)
            else:
                raise TypeError('Data list can only be of ndarray or Sequence')
        elif isinstance(data, Sequence):
            self.__init_from_seqs([data], ref_dataset)
1787
        elif isinstance(data, dt_DataTable):
1788
            self.__init_from_np2d(data.to_numpy(), params_str, ref_dataset)
wxchan's avatar
wxchan committed
1789
1790
1791
1792
        else:
            try:
                csr = scipy.sparse.csr_matrix(data)
                self.__init_from_csr(csr, params_str, ref_dataset)
1793
            except BaseException:
1794
                raise TypeError(f'Cannot initialize Dataset from {type(data).__name__}')
wxchan's avatar
wxchan committed
1795
1796
1797
        if label is not None:
            self.set_label(label)
        if self.get_label() is None:
1798
            raise ValueError("Label should not be None")
wxchan's avatar
wxchan committed
1799
1800
1801
1802
        if weight is not None:
            self.set_weight(weight)
        if group is not None:
            self.set_group(group)
1803
1804
        if isinstance(predictor, _InnerPredictor):
            if self._predictor is None and init_score is not None:
1805
                _log_warning("The init_score will be overridden by the prediction of init_model.")
1806
1807
1808
1809
1810
            self._set_init_score_by_predictor(
                predictor=predictor,
                data=data,
                used_indices=None
            )
1811
1812
        elif init_score is not None:
            self.set_init_score(init_score)
Guolin Ke's avatar
Guolin Ke committed
1813
        elif predictor is not None:
1814
            raise TypeError(f'Wrong predictor type {type(predictor).__name__}')
Guolin Ke's avatar
Guolin Ke committed
1815
        # set feature names
Nikita Titov's avatar
Nikita Titov committed
1816
        return self.set_feature_name(feature_name)
wxchan's avatar
wxchan committed
1817

1818
1819
    @staticmethod
    def _yield_row_from_seqlist(seqs: List[Sequence], indices: Iterable[int]):
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
        offset = 0
        seq_id = 0
        seq = seqs[seq_id]
        for row_id in indices:
            assert row_id >= offset, "sample indices are expected to be monotonic"
            while row_id >= offset + len(seq):
                offset += len(seq)
                seq_id += 1
                seq = seqs[seq_id]
            id_in_seq = row_id - offset
            row = seq[id_in_seq]
            yield row if row.flags['OWNDATA'] else row.copy()

    def __sample(self, seqs: List[Sequence], total_nrow: int) -> Tuple[List[np.ndarray], List[np.ndarray]]:
        """Sample data from seqs.

        Mimics behavior in c_api.cpp:LGBM_DatasetCreateFromMats()

        Returns
        -------
            sampled_rows, sampled_row_indices
        """
        indices = self._create_sample_indices(total_nrow)

        # Select sampled rows, transpose to column order.
1845
        sampled = np.array([row for row in self._yield_row_from_seqlist(seqs, indices)])
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
        sampled = sampled.T

        filtered = []
        filtered_idx = []
        sampled_row_range = np.arange(len(indices), dtype=np.int32)
        for col in sampled:
            col_predicate = (np.abs(col) > ZERO_THRESHOLD) | np.isnan(col)
            filtered_col = col[col_predicate]
            filtered_row_idx = sampled_row_range[col_predicate]

            filtered.append(filtered_col)
            filtered_idx.append(filtered_row_idx)

        return filtered, filtered_idx

1861
1862
1863
    def __init_from_seqs(
        self,
        seqs: List[Sequence],
1864
        ref_dataset: Optional[_DatasetHandle]
1865
    ) -> "Dataset":
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
        """
        Initialize data from list of Sequence objects.

        Sequence: Generic Data Access Object
            Supports random access and access by batch if properly defined by user

        Data scheme uniformity are trusted, not checked
        """
        total_nrow = sum(len(seq) for seq in seqs)

        # create validation dataset from ref_dataset
        if ref_dataset is not None:
            self._init_from_ref_dataset(total_nrow, ref_dataset)
        else:
1880
            param_str = _param_dict_to_str(self.get_params())
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
            sample_cnt = _get_sample_count(total_nrow, param_str)

            sample_data, col_indices = self.__sample(seqs, total_nrow)
            self._init_from_sample(sample_data, col_indices, sample_cnt, total_nrow)

        for seq in seqs:
            nrow = len(seq)
            batch_size = getattr(seq, 'batch_size', None) or Sequence.batch_size
            for start in range(0, nrow, batch_size):
                end = min(start + batch_size, nrow)
                self._push_rows(seq[start:end])
        return self

1894
1895
1896
1897
1898
1899
    def __init_from_np2d(
        self,
        mat: np.ndarray,
        params_str: str,
        ref_dataset: Optional[_DatasetHandle]
    ) -> "Dataset":
1900
        """Initialize data from a 2-D numpy matrix."""
wxchan's avatar
wxchan committed
1901
1902
1903
1904
1905
1906
        if len(mat.shape) != 2:
            raise ValueError('Input numpy.ndarray must be 2 dimensional')

        self.handle = ctypes.c_void_p()
        if mat.dtype == np.float32 or mat.dtype == np.float64:
            data = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
1907
        else:  # change non-float data to float data, need to copy
wxchan's avatar
wxchan committed
1908
1909
            data = np.array(mat.reshape(mat.size), dtype=np.float32)

1910
        ptr_data, type_ptr_data, _ = _c_float_array(data)
wxchan's avatar
wxchan committed
1911
1912
        _safe_call(_LIB.LGBM_DatasetCreateFromMat(
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1913
            ctypes.c_int(type_ptr_data),
1914
1915
            ctypes.c_int32(mat.shape[0]),
            ctypes.c_int32(mat.shape[1]),
1916
            ctypes.c_int(_C_API_IS_ROW_MAJOR),
1917
            _c_str(params_str),
wxchan's avatar
wxchan committed
1918
1919
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
1920
        return self
wxchan's avatar
wxchan committed
1921

1922
1923
1924
1925
1926
1927
    def __init_from_list_np2d(
        self,
        mats: List[np.ndarray],
        params_str: str,
        ref_dataset: Optional[_DatasetHandle]
    ) -> "Dataset":
1928
        """Initialize data from a list of 2-D numpy matrices."""
1929
        ncol = mats[0].shape[1]
1930
        nrow = np.empty((len(mats),), np.int32)
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
        if mats[0].dtype == np.float64:
            ptr_data = (ctypes.POINTER(ctypes.c_double) * len(mats))()
        else:
            ptr_data = (ctypes.POINTER(ctypes.c_float) * len(mats))()

        holders = []
        type_ptr_data = None

        for i, mat in enumerate(mats):
            if len(mat.shape) != 2:
                raise ValueError('Input numpy.ndarray must be 2 dimensional')

            if mat.shape[1] != ncol:
                raise ValueError('Input arrays must have same number of columns')

            nrow[i] = mat.shape[0]

            if mat.dtype == np.float32 or mat.dtype == np.float64:
                mats[i] = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
1950
            else:  # change non-float data to float data, need to copy
1951
1952
                mats[i] = np.array(mat.reshape(mat.size), dtype=np.float32)

1953
            chunk_ptr_data, chunk_type_ptr_data, holder = _c_float_array(mats[i])
1954
1955
1956
1957
1958
1959
1960
1961
            if type_ptr_data is not None and chunk_type_ptr_data != type_ptr_data:
                raise ValueError('Input chunks must have same type')
            ptr_data[i] = chunk_ptr_data
            type_ptr_data = chunk_type_ptr_data
            holders.append(holder)

        self.handle = ctypes.c_void_p()
        _safe_call(_LIB.LGBM_DatasetCreateFromMats(
1962
            ctypes.c_int32(len(mats)),
1963
1964
1965
            ctypes.cast(ptr_data, ctypes.POINTER(ctypes.POINTER(ctypes.c_double))),
            ctypes.c_int(type_ptr_data),
            nrow.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
1966
            ctypes.c_int32(ncol),
1967
            ctypes.c_int(_C_API_IS_ROW_MAJOR),
1968
            _c_str(params_str),
1969
1970
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
1971
        return self
1972

1973
1974
1975
1976
1977
1978
    def __init_from_csr(
        self,
        csr: scipy.sparse.csr_matrix,
        params_str: str,
        ref_dataset: Optional[_DatasetHandle]
    ) -> "Dataset":
1979
        """Initialize data from a CSR matrix."""
wxchan's avatar
wxchan committed
1980
        if len(csr.indices) != len(csr.data):
1981
            raise ValueError(f'Length mismatch: {len(csr.indices)} vs {len(csr.data)}')
wxchan's avatar
wxchan committed
1982
1983
        self.handle = ctypes.c_void_p()

1984
1985
        ptr_indptr, type_ptr_indptr, __ = _c_int_array(csr.indptr)
        ptr_data, type_ptr_data, _ = _c_float_array(csr.data)
wxchan's avatar
wxchan committed
1986

1987
        assert csr.shape[1] <= _MAX_INT32
1988
        csr_indices = csr.indices.astype(np.int32, copy=False)
1989

wxchan's avatar
wxchan committed
1990
1991
        _safe_call(_LIB.LGBM_DatasetCreateFromCSR(
            ptr_indptr,
Guolin Ke's avatar
Guolin Ke committed
1992
            ctypes.c_int(type_ptr_indptr),
1993
            csr_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
wxchan's avatar
wxchan committed
1994
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1995
1996
1997
1998
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csr.indptr)),
            ctypes.c_int64(len(csr.data)),
            ctypes.c_int64(csr.shape[1]),
1999
            _c_str(params_str),
wxchan's avatar
wxchan committed
2000
2001
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
2002
        return self
wxchan's avatar
wxchan committed
2003

2004
2005
2006
2007
2008
2009
    def __init_from_csc(
        self,
        csc: scipy.sparse.csc_matrix,
        params_str: str,
        ref_dataset: Optional[_DatasetHandle]
    ) -> "Dataset":
2010
        """Initialize data from a CSC matrix."""
Guolin Ke's avatar
Guolin Ke committed
2011
        if len(csc.indices) != len(csc.data):
2012
            raise ValueError(f'Length mismatch: {len(csc.indices)} vs {len(csc.data)}')
Guolin Ke's avatar
Guolin Ke committed
2013
2014
        self.handle = ctypes.c_void_p()

2015
2016
        ptr_indptr, type_ptr_indptr, __ = _c_int_array(csc.indptr)
        ptr_data, type_ptr_data, _ = _c_float_array(csc.data)
Guolin Ke's avatar
Guolin Ke committed
2017

2018
        assert csc.shape[0] <= _MAX_INT32
2019
        csc_indices = csc.indices.astype(np.int32, copy=False)
2020

Guolin Ke's avatar
Guolin Ke committed
2021
2022
        _safe_call(_LIB.LGBM_DatasetCreateFromCSC(
            ptr_indptr,
Guolin Ke's avatar
Guolin Ke committed
2023
            ctypes.c_int(type_ptr_indptr),
2024
            csc_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
Guolin Ke's avatar
Guolin Ke committed
2025
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
2026
2027
2028
2029
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csc.indptr)),
            ctypes.c_int64(len(csc.data)),
            ctypes.c_int64(csc.shape[0]),
2030
            _c_str(params_str),
Guolin Ke's avatar
Guolin Ke committed
2031
2032
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
2033
        return self
Guolin Ke's avatar
Guolin Ke committed
2034

2035
    @staticmethod
2036
2037
2038
2039
2040
2041
    def _compare_params_for_warning(
        params: Optional[Dict[str, Any]],
        other_params: Optional[Dict[str, Any]],
        ignore_keys: Set[str]
    ) -> bool:
        """Compare two dictionaries with params ignoring some keys.
2042

2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
        It is only for the warning purpose.

        Parameters
        ----------
        params : dict or None
            One dictionary with parameters to compare.
        other_params : dict or None
            Another dictionary with parameters to compare.
        ignore_keys : set
            Keys that should be ignored during comparing two dictionaries.
2053
2054
2055

        Returns
        -------
2056
2057
        compare_result : bool
          Returns whether two dictionaries with params are equal.
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
        """
        if params is None:
            params = {}
        if other_params is None:
            other_params = {}
        for k in other_params:
            if k not in ignore_keys:
                if k not in params or params[k] != other_params[k]:
                    return False
        for k in params:
            if k not in ignore_keys:
                if k not in other_params or params[k] != other_params[k]:
                    return False
        return True

2073
    def construct(self) -> "Dataset":
2074
2075
2076
2077
2078
        """Lazy init.

        Returns
        -------
        self : Dataset
Nikita Titov's avatar
Nikita Titov committed
2079
            Constructed Dataset object.
2080
        """
2081
        if self.handle is None:
wxchan's avatar
wxchan committed
2082
            if self.reference is not None:
2083
                reference_params = self.reference.get_params()
2084
2085
                params = self.get_params()
                if params != reference_params:
2086
2087
2088
2089
2090
                    if not self._compare_params_for_warning(
                        params=params,
                        other_params=reference_params,
                        ignore_keys=_ConfigAliases.get("categorical_feature")
                    ):
2091
                        _log_warning('Overriding the parameters from Reference Dataset.')
2092
                    self._update_params(reference_params)
wxchan's avatar
wxchan committed
2093
                if self.used_indices is None:
2094
                    # create valid
2095
                    self._lazy_init(self.data, label=self.label, reference=self.reference,
2096
2097
                                    weight=self.weight, group=self.group,
                                    init_score=self.init_score, predictor=self._predictor,
2098
                                    feature_name=self.feature_name, params=self.params)
wxchan's avatar
wxchan committed
2099
                else:
2100
                    # construct subset
2101
                    used_indices = _list_to_1d_numpy(self.used_indices, np.int32, name='used_indices')
2102
                    assert used_indices.flags.c_contiguous
Guolin Ke's avatar
Guolin Ke committed
2103
                    if self.reference.group is not None:
2104
                        group_info = np.array(self.reference.group).astype(np.int32, copy=False)
2105
                        _, self.group = np.unique(np.repeat(range(len(group_info)), repeats=group_info)[self.used_indices],
2106
                                                  return_counts=True)
2107
                    self.handle = ctypes.c_void_p()
2108
                    params_str = _param_dict_to_str(self.params)
wxchan's avatar
wxchan committed
2109
                    _safe_call(_LIB.LGBM_DatasetGetSubset(
2110
                        self.reference.construct().handle,
wxchan's avatar
wxchan committed
2111
                        used_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
2112
                        ctypes.c_int32(used_indices.shape[0]),
2113
                        _c_str(params_str),
wxchan's avatar
wxchan committed
2114
                        ctypes.byref(self.handle)))
Guolin Ke's avatar
Guolin Ke committed
2115
2116
                    if not self.free_raw_data:
                        self.get_data()
Guolin Ke's avatar
Guolin Ke committed
2117
2118
                    if self.group is not None:
                        self.set_group(self.group)
wxchan's avatar
wxchan committed
2119
2120
                    if self.get_label() is None:
                        raise ValueError("Label should not be None.")
Guolin Ke's avatar
Guolin Ke committed
2121
2122
                    if isinstance(self._predictor, _InnerPredictor) and self._predictor is not self.reference._predictor:
                        self.get_data()
2123
2124
2125
2126
2127
                        self._set_init_score_by_predictor(
                            predictor=self._predictor,
                            data=self.data,
                            used_indices=used_indices
                        )
wxchan's avatar
wxchan committed
2128
            else:
2129
                # create train
2130
                self._lazy_init(self.data, label=self.label,
2131
2132
                                weight=self.weight, group=self.group,
                                init_score=self.init_score, predictor=self._predictor,
2133
                                feature_name=self.feature_name, categorical_feature=self.categorical_feature, params=self.params)
wxchan's avatar
wxchan committed
2134
2135
            if self.free_raw_data:
                self.data = None
2136
            self.feature_name = self.get_feature_name()
wxchan's avatar
wxchan committed
2137
        return self
wxchan's avatar
wxchan committed
2138

2139
2140
2141
    def create_valid(
        self,
        data,
2142
        label: Optional[_LGBM_LabelType] = None,
2143
2144
2145
2146
2147
        weight=None,
        group=None,
        init_score=None,
        params: Optional[Dict[str, Any]] = None
    ) -> "Dataset":
2148
        """Create validation data align with current Dataset.
wxchan's avatar
wxchan committed
2149
2150
2151

        Parameters
        ----------
2152
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame, scipy.sparse, Sequence, list of Sequence or list of numpy array
wxchan's avatar
wxchan committed
2153
            Data source of Dataset.
2154
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM) or a LightGBM Dataset binary file.
2155
        label : list, numpy 1-D array, pandas Series / one-column DataFrame or None, optional (default=None)
2156
2157
            Label of the data.
        weight : list, numpy 1-D array, pandas Series or None, optional (default=None)
2158
            Weight for each instance. Weights should be non-negative.
2159
        group : list, numpy 1-D array, pandas Series or None, optional (default=None)
2160
2161
2162
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
2163
2164
            For example, if you have a 100-document dataset with ``group = [10, 20, 40, 10, 10, 10]``, that means that you have 6 groups,
            where the first 10 records are in the first group, records 11-30 are in the second group, records 31-70 are in the third group, etc.
2165
        init_score : list, list of lists (for multi-class task), numpy array, pandas Series, pandas DataFrame (for multi-class task), or None, optional (default=None)
2166
            Init score for Dataset.
Nikita Titov's avatar
Nikita Titov committed
2167
        params : dict or None, optional (default=None)
2168
            Other parameters for validation Dataset.
2169
2170
2171

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2172
2173
        valid : Dataset
            Validation Dataset with reference to self.
wxchan's avatar
wxchan committed
2174
        """
2175
        ret = Dataset(data, label=label, reference=self,
2176
                      weight=weight, group=group, init_score=init_score,
2177
                      params=params, free_raw_data=self.free_raw_data)
wxchan's avatar
wxchan committed
2178
        ret._predictor = self._predictor
2179
        ret.pandas_categorical = self.pandas_categorical
wxchan's avatar
wxchan committed
2180
        return ret
wxchan's avatar
wxchan committed
2181

2182
2183
2184
2185
2186
    def subset(
        self,
        used_indices: List[int],
        params: Optional[Dict[str, Any]] = None
    ) -> "Dataset":
2187
        """Get subset of current Dataset.
wxchan's avatar
wxchan committed
2188
2189
2190
2191

        Parameters
        ----------
        used_indices : list of int
2192
            Indices used to create the subset.
Nikita Titov's avatar
Nikita Titov committed
2193
        params : dict or None, optional (default=None)
2194
            These parameters will be passed to Dataset constructor.
2195
2196
2197
2198
2199

        Returns
        -------
        subset : Dataset
            Subset of the current Dataset.
wxchan's avatar
wxchan committed
2200
        """
wxchan's avatar
wxchan committed
2201
2202
        if params is None:
            params = self.params
wxchan's avatar
wxchan committed
2203
        ret = Dataset(None, reference=self, feature_name=self.feature_name,
2204
2205
                      categorical_feature=self.categorical_feature, params=params,
                      free_raw_data=self.free_raw_data)
wxchan's avatar
wxchan committed
2206
        ret._predictor = self._predictor
2207
        ret.pandas_categorical = self.pandas_categorical
2208
        ret.used_indices = sorted(used_indices)
wxchan's avatar
wxchan committed
2209
2210
        return ret

2211
    def save_binary(self, filename: Union[str, Path]) -> "Dataset":
2212
        """Save Dataset to a binary file.
wxchan's avatar
wxchan committed
2213

2214
2215
2216
2217
2218
        .. note::

            Please note that `init_score` is not saved in binary file.
            If you need it, please set it again after loading Dataset.

wxchan's avatar
wxchan committed
2219
2220
        Parameters
        ----------
2221
        filename : str or pathlib.Path
wxchan's avatar
wxchan committed
2222
            Name of the output file.
Nikita Titov's avatar
Nikita Titov committed
2223
2224
2225
2226
2227

        Returns
        -------
        self : Dataset
            Returns self.
wxchan's avatar
wxchan committed
2228
2229
2230
        """
        _safe_call(_LIB.LGBM_DatasetSaveBinary(
            self.construct().handle,
2231
            _c_str(str(filename))))
Nikita Titov's avatar
Nikita Titov committed
2232
        return self
wxchan's avatar
wxchan committed
2233

2234
    def _update_params(self, params: Optional[Dict[str, Any]]) -> "Dataset":
2235
2236
        if not params:
            return self
2237
        params = deepcopy(params)
2238
2239
2240
2241
2242

        def update():
            if not self.params:
                self.params = params
            else:
2243
                self._params_back_up = deepcopy(self.params)
2244
2245
2246
2247
2248
2249
                self.params.update(params)

        if self.handle is None:
            update()
        elif params is not None:
            ret = _LIB.LGBM_DatasetUpdateParamChecking(
2250
2251
                _c_str(_param_dict_to_str(self.params)),
                _c_str(_param_dict_to_str(params)))
2252
2253
2254
2255
2256
2257
            if ret != 0:
                # could be updated if data is not freed
                if self.data is not None:
                    update()
                    self._free_handle()
                else:
2258
                    raise LightGBMError(_LIB.LGBM_GetLastError().decode('utf-8'))
Nikita Titov's avatar
Nikita Titov committed
2259
        return self
wxchan's avatar
wxchan committed
2260

2261
    def _reverse_update_params(self) -> "Dataset":
2262
        if self.handle is None:
2263
2264
            self.params = deepcopy(self._params_back_up)
            self._params_back_up = None
Nikita Titov's avatar
Nikita Titov committed
2265
        return self
2266

2267
2268
2269
2270
2271
    def set_field(
        self,
        field_name: str,
        data
    ) -> "Dataset":
wxchan's avatar
wxchan committed
2272
        """Set property into the Dataset.
wxchan's avatar
wxchan committed
2273
2274
2275

        Parameters
        ----------
2276
        field_name : str
2277
            The field name of the information.
2278
2279
        data : list, list of lists (for multi-class task), numpy array, pandas Series, pandas DataFrame (for multi-class task), or None
            The data to be set.
Nikita Titov's avatar
Nikita Titov committed
2280
2281
2282
2283
2284

        Returns
        -------
        self : Dataset
            Dataset with set property.
wxchan's avatar
wxchan committed
2285
        """
2286
        if self.handle is None:
2287
            raise Exception(f"Cannot set {field_name} before construct dataset")
wxchan's avatar
wxchan committed
2288
        if data is None:
2289
            # set to None
wxchan's avatar
wxchan committed
2290
2291
            _safe_call(_LIB.LGBM_DatasetSetField(
                self.handle,
2292
                _c_str(field_name),
wxchan's avatar
wxchan committed
2293
                None,
Guolin Ke's avatar
Guolin Ke committed
2294
                ctypes.c_int(0),
2295
                ctypes.c_int(_FIELD_TYPE_MAPPER[field_name])))
Nikita Titov's avatar
Nikita Titov committed
2296
            return self
2297
        if field_name == 'init_score':
Guolin Ke's avatar
Guolin Ke committed
2298
            dtype = np.float64
2299
            if _is_1d_collection(data):
2300
                data = _list_to_1d_numpy(data, dtype, name=field_name)
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
            elif _is_2d_collection(data):
                data = _data_to_2d_numpy(data, dtype, name=field_name)
                data = data.ravel(order='F')
            else:
                raise TypeError(
                    'init_score must be list, numpy 1-D array or pandas Series.\n'
                    'In multiclass classification init_score can also be a list of lists, numpy 2-D array or pandas DataFrame.'
                )
        else:
            dtype = np.int32 if field_name == 'group' else np.float32
2311
            data = _list_to_1d_numpy(data, dtype, name=field_name)
2312

2313
        if data.dtype == np.float32 or data.dtype == np.float64:
2314
            ptr_data, type_data, _ = _c_float_array(data)
wxchan's avatar
wxchan committed
2315
        elif data.dtype == np.int32:
2316
            ptr_data, type_data, _ = _c_int_array(data)
wxchan's avatar
wxchan committed
2317
        else:
2318
            raise TypeError(f"Expected np.float32/64 or np.int32, met type({data.dtype})")
2319
        if type_data != _FIELD_TYPE_MAPPER[field_name]:
2320
            raise TypeError("Input type error for set_field")
wxchan's avatar
wxchan committed
2321
2322
        _safe_call(_LIB.LGBM_DatasetSetField(
            self.handle,
2323
            _c_str(field_name),
wxchan's avatar
wxchan committed
2324
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
2325
2326
            ctypes.c_int(len(data)),
            ctypes.c_int(type_data)))
2327
        self.version += 1
Nikita Titov's avatar
Nikita Titov committed
2328
        return self
wxchan's avatar
wxchan committed
2329

2330
    def get_field(self, field_name: str) -> Optional[np.ndarray]:
wxchan's avatar
wxchan committed
2331
        """Get property from the Dataset.
wxchan's avatar
wxchan committed
2332
2333
2334

        Parameters
        ----------
2335
        field_name : str
2336
            The field name of the information.
wxchan's avatar
wxchan committed
2337
2338
2339

        Returns
        -------
2340
        info : numpy array or None
2341
            A numpy array with information from the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2342
        """
2343
        if self.handle is None:
2344
            raise Exception(f"Cannot get {field_name} before construct Dataset")
2345
2346
        tmp_out_len = ctypes.c_int(0)
        out_type = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2347
2348
2349
        ret = ctypes.POINTER(ctypes.c_void_p)()
        _safe_call(_LIB.LGBM_DatasetGetField(
            self.handle,
2350
            _c_str(field_name),
wxchan's avatar
wxchan committed
2351
2352
2353
            ctypes.byref(tmp_out_len),
            ctypes.byref(ret),
            ctypes.byref(out_type)))
2354
        if out_type.value != _FIELD_TYPE_MAPPER[field_name]:
wxchan's avatar
wxchan committed
2355
2356
2357
            raise TypeError("Return type error for get_field")
        if tmp_out_len.value == 0:
            return None
2358
        if out_type.value == _C_API_DTYPE_INT32:
2359
            arr = _cint32_array_to_numpy(ctypes.cast(ret, ctypes.POINTER(ctypes.c_int32)), tmp_out_len.value)
2360
        elif out_type.value == _C_API_DTYPE_FLOAT32:
2361
            arr = _cfloat32_array_to_numpy(ctypes.cast(ret, ctypes.POINTER(ctypes.c_float)), tmp_out_len.value)
2362
        elif out_type.value == _C_API_DTYPE_FLOAT64:
2363
            arr = _cfloat64_array_to_numpy(ctypes.cast(ret, ctypes.POINTER(ctypes.c_double)), tmp_out_len.value)
2364
        else:
wxchan's avatar
wxchan committed
2365
            raise TypeError("Unknown type")
2366
2367
2368
2369
2370
2371
        if field_name == 'init_score':
            num_data = self.num_data()
            num_classes = arr.size // num_data
            if num_classes > 1:
                arr = arr.reshape((num_data, num_classes), order='F')
        return arr
Guolin Ke's avatar
Guolin Ke committed
2372

2373
2374
    def set_categorical_feature(
        self,
2375
        categorical_feature: _LGBM_CategoricalFeatureConfiguration
2376
    ) -> "Dataset":
2377
        """Set categorical features.
2378
2379
2380

        Parameters
        ----------
2381
        categorical_feature : list of str or int, or 'auto'
2382
            Names or indices of categorical features.
Nikita Titov's avatar
Nikita Titov committed
2383
2384
2385
2386
2387

        Returns
        -------
        self : Dataset
            Dataset with set categorical features.
2388
2389
        """
        if self.categorical_feature == categorical_feature:
Nikita Titov's avatar
Nikita Titov committed
2390
            return self
2391
        if self.data is not None:
2392
2393
            if self.categorical_feature is None:
                self.categorical_feature = categorical_feature
Nikita Titov's avatar
Nikita Titov committed
2394
                return self._free_handle()
2395
            elif categorical_feature == 'auto':
Nikita Titov's avatar
Nikita Titov committed
2396
                return self
2397
            else:
2398
2399
2400
                if self.categorical_feature != 'auto':
                    _log_warning('categorical_feature in Dataset is overridden.\n'
                                 f'New categorical_feature is {sorted(list(categorical_feature))}')
2401
                self.categorical_feature = categorical_feature
Nikita Titov's avatar
Nikita Titov committed
2402
                return self._free_handle()
2403
        else:
2404
2405
            raise LightGBMError("Cannot set categorical feature after freed raw data, "
                                "set free_raw_data=False when construct Dataset to avoid this.")
2406

2407
2408
2409
2410
    def _set_predictor(
        self,
        predictor: Optional[_InnerPredictor]
    ) -> "Dataset":
2411
2412
2413
2414
        """Set predictor for continued training.

        It is not recommended for user to call this function.
        Please use init_model argument in engine.train() or engine.cv() instead.
Guolin Ke's avatar
Guolin Ke committed
2415
        """
2416
        if predictor is None and self._predictor is None:
Nikita Titov's avatar
Nikita Titov committed
2417
            return self
2418
2419
2420
        elif isinstance(predictor, _InnerPredictor) and isinstance(self._predictor, _InnerPredictor):
            if (predictor == self._predictor) and (predictor.current_iteration() == self._predictor.current_iteration()):
                return self
2421
        if self.handle is None:
Guolin Ke's avatar
Guolin Ke committed
2422
            self._predictor = predictor
2423
2424
        elif self.data is not None:
            self._predictor = predictor
2425
2426
2427
2428
2429
            self._set_init_score_by_predictor(
                predictor=self._predictor,
                data=self.data,
                used_indices=None
            )
2430
2431
        elif self.used_indices is not None and self.reference is not None and self.reference.data is not None:
            self._predictor = predictor
2432
2433
2434
2435
2436
            self._set_init_score_by_predictor(
                predictor=self._predictor,
                data=self.reference.data,
                used_indices=self.used_indices
            )
Guolin Ke's avatar
Guolin Ke committed
2437
        else:
2438
2439
            raise LightGBMError("Cannot set predictor after freed raw data, "
                                "set free_raw_data=False when construct Dataset to avoid this.")
2440
        return self
Guolin Ke's avatar
Guolin Ke committed
2441

2442
    def set_reference(self, reference: "Dataset") -> "Dataset":
2443
        """Set reference Dataset.
Guolin Ke's avatar
Guolin Ke committed
2444
2445
2446
2447

        Parameters
        ----------
        reference : Dataset
2448
            Reference that is used as a template to construct the current Dataset.
Nikita Titov's avatar
Nikita Titov committed
2449
2450
2451
2452
2453

        Returns
        -------
        self : Dataset
            Dataset with set reference.
Guolin Ke's avatar
Guolin Ke committed
2454
        """
2455
2456
2457
        self.set_categorical_feature(reference.categorical_feature) \
            .set_feature_name(reference.feature_name) \
            ._set_predictor(reference._predictor)
2458
        # we're done if self and reference share a common upstream reference
2459
        if self.get_ref_chain().intersection(reference.get_ref_chain()):
Nikita Titov's avatar
Nikita Titov committed
2460
            return self
Guolin Ke's avatar
Guolin Ke committed
2461
2462
        if self.data is not None:
            self.reference = reference
Nikita Titov's avatar
Nikita Titov committed
2463
            return self._free_handle()
Guolin Ke's avatar
Guolin Ke committed
2464
        else:
2465
2466
            raise LightGBMError("Cannot set reference after freed raw data, "
                                "set free_raw_data=False when construct Dataset to avoid this.")
Guolin Ke's avatar
Guolin Ke committed
2467

2468
    def set_feature_name(self, feature_name: Union[List[str], str]) -> "Dataset":
2469
        """Set feature name.
Guolin Ke's avatar
Guolin Ke committed
2470
2471
2472

        Parameters
        ----------
2473
        feature_name : list of str
2474
            Feature names.
Nikita Titov's avatar
Nikita Titov committed
2475
2476
2477
2478
2479

        Returns
        -------
        self : Dataset
            Dataset with set feature name.
Guolin Ke's avatar
Guolin Ke committed
2480
        """
2481
2482
        if feature_name != 'auto':
            self.feature_name = feature_name
2483
        if self.handle is not None and feature_name is not None and feature_name != 'auto':
wxchan's avatar
wxchan committed
2484
            if len(feature_name) != self.num_feature():
2485
                raise ValueError(f"Length of feature_name({len(feature_name)}) and num_feature({self.num_feature()}) don't match")
2486
            c_feature_name = [_c_str(name) for name in feature_name]
wxchan's avatar
wxchan committed
2487
2488
            _safe_call(_LIB.LGBM_DatasetSetFeatureNames(
                self.handle,
2489
                _c_array(ctypes.c_char_p, c_feature_name),
Guolin Ke's avatar
Guolin Ke committed
2490
                ctypes.c_int(len(feature_name))))
Nikita Titov's avatar
Nikita Titov committed
2491
        return self
Guolin Ke's avatar
Guolin Ke committed
2492

2493
    def set_label(self, label: Optional[_LGBM_LabelType]) -> "Dataset":
2494
        """Set label of Dataset.
Guolin Ke's avatar
Guolin Ke committed
2495
2496
2497

        Parameters
        ----------
2498
        label : list, numpy 1-D array, pandas Series / one-column DataFrame or None
2499
            The label information to be set into Dataset.
Nikita Titov's avatar
Nikita Titov committed
2500
2501
2502
2503
2504

        Returns
        -------
        self : Dataset
            Dataset with set label.
Guolin Ke's avatar
Guolin Ke committed
2505
2506
        """
        self.label = label
2507
        if self.handle is not None:
2508
2509
2510
2511
            if isinstance(label, pd_DataFrame):
                if len(label.columns) > 1:
                    raise ValueError('DataFrame for label cannot have multiple columns')
                _check_for_bad_pandas_dtypes(label.dtypes)
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
                try:
                    # most common case (no nullable dtypes)
                    label = label.to_numpy(dtype=np.float32, copy=False)
                except TypeError:
                    # 1.0 <= pd version < 1.1 and nullable dtypes, least common case
                    # raises error because array is casted to type(pd.NA) and there's no na_value argument
                    label = label.astype(np.float32, copy=False).values
                except ValueError:
                    # data has nullable dtypes, but we can specify na_value argument and copy will be made
                    label = label.to_numpy(dtype=np.float32, na_value=np.nan)
                label_array = np.ravel(label)
2523
            else:
2524
                label_array = _list_to_1d_numpy(label, name='label')
2525
            self.set_field('label', label_array)
2526
            self.label = self.get_field('label')  # original values can be modified at cpp side
Nikita Titov's avatar
Nikita Titov committed
2527
        return self
Guolin Ke's avatar
Guolin Ke committed
2528

2529
    def set_weight(self, weight) -> "Dataset":
2530
        """Set weight of each instance.
Guolin Ke's avatar
Guolin Ke committed
2531
2532
2533

        Parameters
        ----------
2534
        weight : list, numpy 1-D array, pandas Series or None
2535
            Weight to be set for each data point. Weights should be non-negative.
Nikita Titov's avatar
Nikita Titov committed
2536
2537
2538
2539
2540

        Returns
        -------
        self : Dataset
            Dataset with set weight.
Guolin Ke's avatar
Guolin Ke committed
2541
        """
2542
2543
        if weight is not None and np.all(weight == 1):
            weight = None
Guolin Ke's avatar
Guolin Ke committed
2544
        self.weight = weight
2545
        if self.handle is not None and weight is not None:
2546
            weight = _list_to_1d_numpy(weight, name='weight')
wxchan's avatar
wxchan committed
2547
            self.set_field('weight', weight)
2548
            self.weight = self.get_field('weight')  # original values can be modified at cpp side
Nikita Titov's avatar
Nikita Titov committed
2549
        return self
Guolin Ke's avatar
Guolin Ke committed
2550

2551
    def set_init_score(self, init_score) -> "Dataset":
2552
        """Set init score of Booster to start from.
Guolin Ke's avatar
Guolin Ke committed
2553
2554
2555

        Parameters
        ----------
2556
        init_score : list, list of lists (for multi-class task), numpy array, pandas Series, pandas DataFrame (for multi-class task), or None
2557
            Init score for Booster.
Nikita Titov's avatar
Nikita Titov committed
2558
2559
2560
2561
2562

        Returns
        -------
        self : Dataset
            Dataset with set init score.
Guolin Ke's avatar
Guolin Ke committed
2563
2564
        """
        self.init_score = init_score
2565
        if self.handle is not None and init_score is not None:
wxchan's avatar
wxchan committed
2566
            self.set_field('init_score', init_score)
2567
            self.init_score = self.get_field('init_score')  # original values can be modified at cpp side
Nikita Titov's avatar
Nikita Titov committed
2568
        return self
Guolin Ke's avatar
Guolin Ke committed
2569

2570
    def set_group(self, group) -> "Dataset":
2571
        """Set group size of Dataset (used for ranking).
Guolin Ke's avatar
Guolin Ke committed
2572
2573
2574

        Parameters
        ----------
2575
        group : list, numpy 1-D array, pandas Series or None
2576
2577
2578
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
2579
2580
            For example, if you have a 100-document dataset with ``group = [10, 20, 40, 10, 10, 10]``, that means that you have 6 groups,
            where the first 10 records are in the first group, records 11-30 are in the second group, records 31-70 are in the third group, etc.
Nikita Titov's avatar
Nikita Titov committed
2581
2582
2583
2584
2585

        Returns
        -------
        self : Dataset
            Dataset with set group.
Guolin Ke's avatar
Guolin Ke committed
2586
2587
        """
        self.group = group
2588
        if self.handle is not None and group is not None:
2589
            group = _list_to_1d_numpy(group, np.int32, name='group')
wxchan's avatar
wxchan committed
2590
            self.set_field('group', group)
Nikita Titov's avatar
Nikita Titov committed
2591
        return self
Guolin Ke's avatar
Guolin Ke committed
2592

2593
    def get_feature_name(self) -> List[str]:
2594
2595
2596
2597
        """Get the names of columns (features) in the Dataset.

        Returns
        -------
2598
        feature_names : list of str
2599
2600
2601
2602
2603
2604
2605
2606
            The names of columns (features) in the Dataset.
        """
        if self.handle is None:
            raise LightGBMError("Cannot get feature_name before construct dataset")
        num_feature = self.num_feature()
        tmp_out_len = ctypes.c_int(0)
        reserved_string_buffer_size = 255
        required_string_buffer_size = ctypes.c_size_t(0)
2607
        string_buffers = [ctypes.create_string_buffer(reserved_string_buffer_size) for _ in range(num_feature)]
2608
2609
2610
        ptr_string_buffers = (ctypes.c_char_p * num_feature)(*map(ctypes.addressof, string_buffers))
        _safe_call(_LIB.LGBM_DatasetGetFeatureNames(
            self.handle,
2611
            ctypes.c_int(num_feature),
2612
            ctypes.byref(tmp_out_len),
2613
            ctypes.c_size_t(reserved_string_buffer_size),
2614
2615
2616
2617
            ctypes.byref(required_string_buffer_size),
            ptr_string_buffers))
        if num_feature != tmp_out_len.value:
            raise ValueError("Length of feature names doesn't equal with num_feature")
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
        actual_string_buffer_size = required_string_buffer_size.value
        # if buffer length is not long enough, reallocate buffers
        if reserved_string_buffer_size < actual_string_buffer_size:
            string_buffers = [ctypes.create_string_buffer(actual_string_buffer_size) for _ in range(num_feature)]
            ptr_string_buffers = (ctypes.c_char_p * num_feature)(*map(ctypes.addressof, string_buffers))
            _safe_call(_LIB.LGBM_DatasetGetFeatureNames(
                self.handle,
                ctypes.c_int(num_feature),
                ctypes.byref(tmp_out_len),
                ctypes.c_size_t(actual_string_buffer_size),
                ctypes.byref(required_string_buffer_size),
                ptr_string_buffers))
2630
        return [string_buffers[i].value.decode('utf-8') for i in range(num_feature)]
2631

2632
    def get_label(self) -> Optional[np.ndarray]:
2633
        """Get the label of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2634
2635
2636

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2637
        label : numpy array or None
2638
            The label information from the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2639
        """
2640
        if self.label is None:
wxchan's avatar
wxchan committed
2641
            self.label = self.get_field('label')
Guolin Ke's avatar
Guolin Ke committed
2642
2643
        return self.label

2644
    def get_weight(self) -> Optional[np.ndarray]:
2645
        """Get the weight of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2646
2647
2648

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2649
        weight : numpy array or None
2650
            Weight for each data point from the Dataset. Weights should be non-negative.
Guolin Ke's avatar
Guolin Ke committed
2651
        """
2652
        if self.weight is None:
wxchan's avatar
wxchan committed
2653
            self.weight = self.get_field('weight')
Guolin Ke's avatar
Guolin Ke committed
2654
2655
        return self.weight

2656
    def get_init_score(self) -> Optional[np.ndarray]:
2657
        """Get the initial score of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2658
2659
2660

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2661
        init_score : numpy array or None
2662
            Init score of Booster.
Guolin Ke's avatar
Guolin Ke committed
2663
        """
2664
        if self.init_score is None:
wxchan's avatar
wxchan committed
2665
            self.init_score = self.get_field('init_score')
Guolin Ke's avatar
Guolin Ke committed
2666
2667
        return self.init_score

2668
2669
2670
2671
2672
    def get_data(self):
        """Get the raw data of the Dataset.

        Returns
        -------
2673
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame, scipy.sparse, Sequence, list of Sequence or list of numpy array or None
2674
2675
2676
2677
            Raw data used in the Dataset construction.
        """
        if self.handle is None:
            raise Exception("Cannot get data before construct Dataset")
2678
        if self._need_slice and self.used_indices is not None and self.reference is not None:
Guolin Ke's avatar
Guolin Ke committed
2679
2680
2681
2682
            self.data = self.reference.data
            if self.data is not None:
                if isinstance(self.data, np.ndarray) or scipy.sparse.issparse(self.data):
                    self.data = self.data[self.used_indices, :]
2683
                elif isinstance(self.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2684
                    self.data = self.data.iloc[self.used_indices].copy()
2685
                elif isinstance(self.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2686
                    self.data = self.data[self.used_indices, :]
2687
2688
2689
2690
                elif isinstance(self.data, Sequence):
                    self.data = self.data[self.used_indices]
                elif isinstance(self.data, list) and len(self.data) > 0 and all(isinstance(x, Sequence) for x in self.data):
                    self.data = np.array([row for row in self._yield_row_from_seqlist(self.data, self.used_indices)])
Guolin Ke's avatar
Guolin Ke committed
2691
                else:
2692
2693
                    _log_warning(f"Cannot subset {type(self.data).__name__} type of raw data.\n"
                                 "Returning original raw data")
2694
            self._need_slice = False
Guolin Ke's avatar
Guolin Ke committed
2695
2696
2697
        if self.data is None:
            raise LightGBMError("Cannot call `get_data` after freed raw data, "
                                "set free_raw_data=False when construct Dataset to avoid this.")
2698
2699
        return self.data

Guolin Ke's avatar
Guolin Ke committed
2700
    def get_group(self):
2701
        """Get the group of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2702
2703
2704

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2705
        group : numpy array or None
2706
2707
2708
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
2709
2710
            For example, if you have a 100-document dataset with ``group = [10, 20, 40, 10, 10, 10]``, that means that you have 6 groups,
            where the first 10 records are in the first group, records 11-30 are in the second group, records 31-70 are in the third group, etc.
Guolin Ke's avatar
Guolin Ke committed
2711
        """
2712
        if self.group is None:
wxchan's avatar
wxchan committed
2713
            self.group = self.get_field('group')
Guolin Ke's avatar
Guolin Ke committed
2714
2715
            if self.group is not None:
                # group data from LightGBM is boundaries data, need to convert to group size
Nikita Titov's avatar
Nikita Titov committed
2716
                self.group = np.diff(self.group)
Guolin Ke's avatar
Guolin Ke committed
2717
2718
        return self.group

2719
    def num_data(self) -> int:
2720
        """Get the number of rows in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2721
2722
2723

        Returns
        -------
2724
2725
        number_of_rows : int
            The number of rows in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2726
        """
2727
        if self.handle is not None:
2728
            ret = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2729
2730
2731
            _safe_call(_LIB.LGBM_DatasetGetNumData(self.handle,
                                                   ctypes.byref(ret)))
            return ret.value
Guolin Ke's avatar
Guolin Ke committed
2732
        else:
2733
            raise LightGBMError("Cannot get num_data before construct dataset")
Guolin Ke's avatar
Guolin Ke committed
2734

2735
    def num_feature(self) -> int:
2736
        """Get the number of columns (features) in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2737
2738
2739

        Returns
        -------
2740
2741
        number_of_columns : int
            The number of columns (features) in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2742
        """
2743
        if self.handle is not None:
2744
            ret = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2745
2746
2747
            _safe_call(_LIB.LGBM_DatasetGetNumFeature(self.handle,
                                                      ctypes.byref(ret)))
            return ret.value
Guolin Ke's avatar
Guolin Ke committed
2748
        else:
2749
            raise LightGBMError("Cannot get num_feature before construct dataset")
Guolin Ke's avatar
Guolin Ke committed
2750

2751
    def feature_num_bin(self, feature: Union[int, str]) -> int:
2752
2753
2754
2755
        """Get the number of bins for a feature.

        Parameters
        ----------
2756
2757
        feature : int or str
            Index or name of the feature.
2758
2759
2760
2761
2762
2763
2764

        Returns
        -------
        number_of_bins : int
            The number of constructed bins for the feature in the Dataset.
        """
        if self.handle is not None:
2765
            if isinstance(feature, str):
2766
2767
2768
                feature_index = self.feature_name.index(feature)
            else:
                feature_index = feature
2769
2770
            ret = ctypes.c_int(0)
            _safe_call(_LIB.LGBM_DatasetGetFeatureNumBin(self.handle,
2771
                                                         ctypes.c_int(feature_index),
2772
2773
2774
2775
2776
                                                         ctypes.byref(ret)))
            return ret.value
        else:
            raise LightGBMError("Cannot get feature_num_bin before construct dataset")

2777
    def get_ref_chain(self, ref_limit: int = 100) -> Set["Dataset"]:
2778
2779
2780
2781
2782
        """Get a chain of Dataset objects.

        Starts with r, then goes to r.reference (if exists),
        then to r.reference.reference, etc.
        until we hit ``ref_limit`` or a reference loop.
2783
2784
2785
2786
2787

        Parameters
        ----------
        ref_limit : int, optional (default=100)
            The limit number of references.
2788
2789
2790

        Returns
        -------
2791
2792
2793
        ref_chain : set of Dataset
            Chain of references of the Datasets.
        """
2794
        head = self
2795
        ref_chain: Set[Dataset] = set()
2796
2797
        while len(ref_chain) < ref_limit:
            if isinstance(head, Dataset):
2798
                ref_chain.add(head)
2799
2800
2801
2802
2803
2804
                if (head.reference is not None) and (head.reference not in ref_chain):
                    head = head.reference
                else:
                    break
            else:
                break
Nikita Titov's avatar
Nikita Titov committed
2805
        return ref_chain
2806

2807
    def add_features_from(self, other: "Dataset") -> "Dataset":
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
        """Add features from other Dataset to the current Dataset.

        Both Datasets must be constructed before calling this method.

        Parameters
        ----------
        other : Dataset
            The Dataset to take features from.

        Returns
        -------
        self : Dataset
            Dataset with the new features added.
        """
        if self.handle is None or other.handle is None:
            raise ValueError('Both source and target Datasets must be constructed before adding features')
        _safe_call(_LIB.LGBM_DatasetAddFeaturesFrom(self.handle, other.handle))
Guolin Ke's avatar
Guolin Ke committed
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
        was_none = self.data is None
        old_self_data_type = type(self.data).__name__
        if other.data is None:
            self.data = None
        elif self.data is not None:
            if isinstance(self.data, np.ndarray):
                if isinstance(other.data, np.ndarray):
                    self.data = np.hstack((self.data, other.data))
                elif scipy.sparse.issparse(other.data):
                    self.data = np.hstack((self.data, other.data.toarray()))
2835
                elif isinstance(other.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2836
                    self.data = np.hstack((self.data, other.data.values))
2837
                elif isinstance(other.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2838
2839
2840
2841
2842
2843
2844
                    self.data = np.hstack((self.data, other.data.to_numpy()))
                else:
                    self.data = None
            elif scipy.sparse.issparse(self.data):
                sparse_format = self.data.getformat()
                if isinstance(other.data, np.ndarray) or scipy.sparse.issparse(other.data):
                    self.data = scipy.sparse.hstack((self.data, other.data), format=sparse_format)
2845
                elif isinstance(other.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2846
                    self.data = scipy.sparse.hstack((self.data, other.data.values), format=sparse_format)
2847
                elif isinstance(other.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2848
2849
2850
                    self.data = scipy.sparse.hstack((self.data, other.data.to_numpy()), format=sparse_format)
                else:
                    self.data = None
2851
            elif isinstance(self.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2852
2853
                if not PANDAS_INSTALLED:
                    raise LightGBMError("Cannot add features to DataFrame type of raw data "
2854
2855
                                        "without pandas installed. "
                                        "Install pandas and restart your session.")
Guolin Ke's avatar
Guolin Ke committed
2856
                if isinstance(other.data, np.ndarray):
2857
                    self.data = concat((self.data, pd_DataFrame(other.data)),
Guolin Ke's avatar
Guolin Ke committed
2858
2859
                                       axis=1, ignore_index=True)
                elif scipy.sparse.issparse(other.data):
2860
                    self.data = concat((self.data, pd_DataFrame(other.data.toarray())),
Guolin Ke's avatar
Guolin Ke committed
2861
                                       axis=1, ignore_index=True)
2862
                elif isinstance(other.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2863
2864
                    self.data = concat((self.data, other.data),
                                       axis=1, ignore_index=True)
2865
2866
                elif isinstance(other.data, dt_DataTable):
                    self.data = concat((self.data, pd_DataFrame(other.data.to_numpy())),
Guolin Ke's avatar
Guolin Ke committed
2867
2868
2869
                                       axis=1, ignore_index=True)
                else:
                    self.data = None
2870
            elif isinstance(self.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2871
                if isinstance(other.data, np.ndarray):
2872
                    self.data = dt_DataTable(np.hstack((self.data.to_numpy(), other.data)))
Guolin Ke's avatar
Guolin Ke committed
2873
                elif scipy.sparse.issparse(other.data):
2874
2875
2876
2877
2878
                    self.data = dt_DataTable(np.hstack((self.data.to_numpy(), other.data.toarray())))
                elif isinstance(other.data, pd_DataFrame):
                    self.data = dt_DataTable(np.hstack((self.data.to_numpy(), other.data.values)))
                elif isinstance(other.data, dt_DataTable):
                    self.data = dt_DataTable(np.hstack((self.data.to_numpy(), other.data.to_numpy())))
Guolin Ke's avatar
Guolin Ke committed
2879
2880
2881
2882
2883
                else:
                    self.data = None
            else:
                self.data = None
        if self.data is None:
2884
2885
            err_msg = (f"Cannot add features from {type(other.data).__name__} type of raw data to "
                       f"{old_self_data_type} type of raw data.\n")
Guolin Ke's avatar
Guolin Ke committed
2886
2887
            err_msg += ("Set free_raw_data=False when construct Dataset to avoid this"
                        if was_none else "Freeing raw data")
2888
            _log_warning(err_msg)
Guolin Ke's avatar
Guolin Ke committed
2889
        self.feature_name = self.get_feature_name()
2890
2891
        _log_warning("Reseting categorical features.\n"
                     "You can set new categorical features via ``set_categorical_feature`` method")
Guolin Ke's avatar
Guolin Ke committed
2892
2893
        self.categorical_feature = "auto"
        self.pandas_categorical = None
2894
2895
        return self

2896
    def _dump_text(self, filename: Union[str, Path]) -> "Dataset":
2897
2898
2899
2900
2901
2902
        """Save Dataset to a text file.

        This format cannot be loaded back in by LightGBM, but is useful for debugging purposes.

        Parameters
        ----------
2903
        filename : str or pathlib.Path
2904
2905
2906
2907
2908
2909
2910
2911
2912
            Name of the output file.

        Returns
        -------
        self : Dataset
            Returns self.
        """
        _safe_call(_LIB.LGBM_DatasetDumpText(
            self.construct().handle,
2913
            _c_str(str(filename))))
2914
2915
        return self

wxchan's avatar
wxchan committed
2916

2917
2918
2919
2920
_LGBM_CustomObjectiveFunction = Callable[
    [np.ndarray, Dataset],
    Tuple[np.ndarray, np.ndarray]
]
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
_LGBM_CustomEvalFunction = Union[
    Callable[
        [np.ndarray, Dataset],
        _LGBM_EvalFunctionResultType
    ],
    Callable[
        [np.ndarray, Dataset],
        List[_LGBM_EvalFunctionResultType]
    ]
]
2931
2932


2933
class Booster:
2934
    """Booster in LightGBM."""
2935

2936
2937
2938
2939
2940
2941
2942
    def __init__(
        self,
        params: Optional[Dict[str, Any]] = None,
        train_set: Optional[Dataset] = None,
        model_file: Optional[Union[str, Path]] = None,
        model_str: Optional[str] = None
    ):
2943
        """Initialize the Booster.
wxchan's avatar
wxchan committed
2944
2945
2946

        Parameters
        ----------
Nikita Titov's avatar
Nikita Titov committed
2947
        params : dict or None, optional (default=None)
2948
2949
2950
            Parameters for Booster.
        train_set : Dataset or None, optional (default=None)
            Training dataset.
2951
        model_file : str, pathlib.Path or None, optional (default=None)
wxchan's avatar
wxchan committed
2952
            Path to the model file.
2953
        model_str : str or None, optional (default=None)
2954
            Model will be loaded from this string.
wxchan's avatar
wxchan committed
2955
        """
2956
        self.handle = None
2957
        self._network = False
wxchan's avatar
wxchan committed
2958
        self.__need_reload_eval_info = True
2959
        self._train_data_name = "training"
2960
        self.__set_objective_to_none = False
wxchan's avatar
wxchan committed
2961
        self.best_iteration = -1
2962
        self.best_score: _LGBM_BoosterBestScoreType = {}
2963
        params = {} if params is None else deepcopy(params)
wxchan's avatar
wxchan committed
2964
        if train_set is not None:
2965
            # Training task
wxchan's avatar
wxchan committed
2966
            if not isinstance(train_set, Dataset):
2967
                raise TypeError(f'Training data should be Dataset instance, met {type(train_set).__name__}')
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
            params = _choose_param_value(
                main_param_name="machines",
                params=params,
                default_value=None
            )
            # if "machines" is given, assume user wants to do distributed learning, and set up network
            if params["machines"] is None:
                params.pop("machines", None)
            else:
                machines = params["machines"]
                if isinstance(machines, str):
                    num_machines_from_machine_list = len(machines.split(','))
                elif isinstance(machines, (list, set)):
                    num_machines_from_machine_list = len(machines)
                    machines = ','.join(machines)
                else:
                    raise ValueError("Invalid machines in params.")

                params = _choose_param_value(
                    main_param_name="num_machines",
                    params=params,
                    default_value=num_machines_from_machine_list
                )
                params = _choose_param_value(
                    main_param_name="local_listen_port",
                    params=params,
                    default_value=12400
                )
                self.set_network(
                    machines=machines,
                    local_listen_port=params["local_listen_port"],
                    listen_time_out=params.get("time_out", 120),
                    num_machines=params["num_machines"]
                )
3002
            # construct booster object
3003
3004
3005
            train_set.construct()
            # copy the parameters from train_set
            params.update(train_set.get_params())
3006
            params_str = _param_dict_to_str(params)
3007
            self.handle = ctypes.c_void_p()
wxchan's avatar
wxchan committed
3008
            _safe_call(_LIB.LGBM_BoosterCreate(
3009
                train_set.handle,
3010
                _c_str(params_str),
wxchan's avatar
wxchan committed
3011
                ctypes.byref(self.handle)))
3012
            # save reference to data
wxchan's avatar
wxchan committed
3013
            self.train_set = train_set
3014
3015
            self.valid_sets: List[Dataset] = []
            self.name_valid_sets: List[str] = []
wxchan's avatar
wxchan committed
3016
            self.__num_dataset = 1
Guolin Ke's avatar
Guolin Ke committed
3017
3018
            self.__init_predictor = train_set._predictor
            if self.__init_predictor is not None:
wxchan's avatar
wxchan committed
3019
3020
                _safe_call(_LIB.LGBM_BoosterMerge(
                    self.handle,
Guolin Ke's avatar
Guolin Ke committed
3021
                    self.__init_predictor.handle))
Guolin Ke's avatar
Guolin Ke committed
3022
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
3023
3024
3025
3026
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.__num_class = out_num_class.value
3027
            # buffer for inner predict
wxchan's avatar
wxchan committed
3028
3029
3030
            self.__inner_predict_buffer = [None]
            self.__is_predicted_cur_iter = [False]
            self.__get_eval_info()
3031
            self.pandas_categorical = train_set.pandas_categorical
3032
            self.train_set_version = train_set.version
wxchan's avatar
wxchan committed
3033
        elif model_file is not None:
3034
            # Prediction task
Guolin Ke's avatar
Guolin Ke committed
3035
            out_num_iterations = ctypes.c_int(0)
3036
            self.handle = ctypes.c_void_p()
wxchan's avatar
wxchan committed
3037
            _safe_call(_LIB.LGBM_BoosterCreateFromModelfile(
3038
                _c_str(str(model_file)),
wxchan's avatar
wxchan committed
3039
3040
                ctypes.byref(out_num_iterations),
                ctypes.byref(self.handle)))
Guolin Ke's avatar
Guolin Ke committed
3041
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
3042
3043
3044
3045
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.__num_class = out_num_class.value
3046
            self.pandas_categorical = _load_pandas_categorical(file_name=model_file)
3047
3048
3049
            if params:
                _log_warning('Ignoring params argument, using parameters from model file.')
            params = self._get_loaded_param()
3050
        elif model_str is not None:
3051
            self.model_from_string(model_str)
wxchan's avatar
wxchan committed
3052
        else:
3053
3054
            raise TypeError('Need at least one training dataset or model file or model string '
                            'to create Booster instance')
3055
        self.params = params
wxchan's avatar
wxchan committed
3056

3057
    def __del__(self) -> None:
3058
        try:
3059
            if self._network:
3060
3061
3062
3063
3064
3065
3066
3067
                self.free_network()
        except AttributeError:
            pass
        try:
            if self.handle is not None:
                _safe_call(_LIB.LGBM_BoosterFree(self.handle))
        except AttributeError:
            pass
wxchan's avatar
wxchan committed
3068

3069
    def __copy__(self) -> "Booster":
wxchan's avatar
wxchan committed
3070
3071
        return self.__deepcopy__(None)

3072
    def __deepcopy__(self, _) -> "Booster":
3073
        model_str = self.model_to_string(num_iteration=-1)
3074
        booster = Booster(model_str=model_str)
3075
        return booster
wxchan's avatar
wxchan committed
3076

3077
    def __getstate__(self) -> Dict[str, Any]:
wxchan's avatar
wxchan committed
3078
3079
3080
3081
3082
        this = self.__dict__.copy()
        handle = this['handle']
        this.pop('train_set', None)
        this.pop('valid_sets', None)
        if handle is not None:
3083
            this["handle"] = self.model_to_string(num_iteration=-1)
wxchan's avatar
wxchan committed
3084
3085
        return this

3086
    def __setstate__(self, state: Dict[str, Any]) -> None:
3087
3088
        model_str = state.get('handle', None)
        if model_str is not None:
wxchan's avatar
wxchan committed
3089
            handle = ctypes.c_void_p()
Guolin Ke's avatar
Guolin Ke committed
3090
            out_num_iterations = ctypes.c_int(0)
3091
            _safe_call(_LIB.LGBM_BoosterLoadModelFromString(
3092
                _c_str(model_str),
3093
3094
                ctypes.byref(out_num_iterations),
                ctypes.byref(handle)))
wxchan's avatar
wxchan committed
3095
3096
3097
            state['handle'] = handle
        self.__dict__.update(state)

3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
    def _get_loaded_param(self) -> Dict[str, Any]:
        buffer_len = 1 << 20
        tmp_out_len = ctypes.c_int64(0)
        string_buffer = ctypes.create_string_buffer(buffer_len)
        ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
        _safe_call(_LIB.LGBM_BoosterGetLoadedParam(
            self.handle,
            ctypes.c_int64(buffer_len),
            ctypes.byref(tmp_out_len),
            ptr_string_buffer))
        actual_len = tmp_out_len.value
        # if buffer length is not long enough, re-allocate a buffer
        if actual_len > buffer_len:
            string_buffer = ctypes.create_string_buffer(actual_len)
            ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
            _safe_call(_LIB.LGBM_BoosterGetLoadedParam(
                self.handle,
                ctypes.c_int64(actual_len),
                ctypes.byref(tmp_out_len),
                ptr_string_buffer))
        return json.loads(string_buffer.value.decode('utf-8'))

3120
    def free_dataset(self) -> "Booster":
Nikita Titov's avatar
Nikita Titov committed
3121
3122
3123
3124
3125
3126
3127
        """Free Booster's Datasets.

        Returns
        -------
        self : Booster
            Booster without Datasets.
        """
wxchan's avatar
wxchan committed
3128
3129
        self.__dict__.pop('train_set', None)
        self.__dict__.pop('valid_sets', None)
3130
        self.__num_dataset = 0
Nikita Titov's avatar
Nikita Titov committed
3131
        return self
wxchan's avatar
wxchan committed
3132

3133
    def _free_buffer(self) -> "Booster":
3134
3135
        self.__inner_predict_buffer = []
        self.__is_predicted_cur_iter = []
Nikita Titov's avatar
Nikita Titov committed
3136
        return self
3137

3138
3139
3140
3141
3142
3143
3144
    def set_network(
        self,
        machines: Union[List[str], Set[str], str],
        local_listen_port: int = 12400,
        listen_time_out: int = 120,
        num_machines: int = 1
    ) -> "Booster":
3145
3146
3147
3148
        """Set the network configuration.

        Parameters
        ----------
3149
        machines : list, set or str
3150
            Names of machines.
Nikita Titov's avatar
Nikita Titov committed
3151
        local_listen_port : int, optional (default=12400)
3152
            TCP listen port for local machines.
Nikita Titov's avatar
Nikita Titov committed
3153
        listen_time_out : int, optional (default=120)
3154
            Socket time-out in minutes.
Nikita Titov's avatar
Nikita Titov committed
3155
        num_machines : int, optional (default=1)
3156
            The number of machines for distributed learning application.
Nikita Titov's avatar
Nikita Titov committed
3157
3158
3159
3160
3161

        Returns
        -------
        self : Booster
            Booster with set network.
3162
        """
3163
3164
        if isinstance(machines, (list, set)):
            machines = ','.join(machines)
3165
        _safe_call(_LIB.LGBM_NetworkInit(_c_str(machines),
3166
3167
3168
                                         ctypes.c_int(local_listen_port),
                                         ctypes.c_int(listen_time_out),
                                         ctypes.c_int(num_machines)))
3169
        self._network = True
Nikita Titov's avatar
Nikita Titov committed
3170
        return self
3171

3172
    def free_network(self) -> "Booster":
Nikita Titov's avatar
Nikita Titov committed
3173
3174
3175
3176
3177
3178
3179
        """Free Booster's network.

        Returns
        -------
        self : Booster
            Booster with freed network.
        """
3180
        _safe_call(_LIB.LGBM_NetworkFree())
3181
        self._network = False
Nikita Titov's avatar
Nikita Titov committed
3182
        return self
3183

3184
    def trees_to_dataframe(self) -> pd_DataFrame:
3185
3186
        """Parse the fitted model and return in an easy-to-read pandas DataFrame.

3187
3188
3189
3190
        The returned DataFrame has the following columns.

            - ``tree_index`` : int64, which tree a node belongs to. 0-based, so a value of ``6``, for example, means "this node is in the 7th tree".
            - ``node_depth`` : int64, how far a node is from the root of the tree. The root node has a value of ``1``, its direct children are ``2``, etc.
3191
3192
3193
3194
3195
            - ``node_index`` : str, unique identifier for a node.
            - ``left_child`` : str, ``node_index`` of the child node to the left of a split. ``None`` for leaf nodes.
            - ``right_child`` : str, ``node_index`` of the child node to the right of a split. ``None`` for leaf nodes.
            - ``parent_index`` : str, ``node_index`` of this node's parent. ``None`` for the root node.
            - ``split_feature`` : str, name of the feature used for splitting. ``None`` for leaf nodes.
3196
3197
            - ``split_gain`` : float64, gain from adding this split to the tree. ``NaN`` for leaf nodes.
            - ``threshold`` : float64, value of the feature used to decide which side of the split a record will go down. ``NaN`` for leaf nodes.
3198
            - ``decision_type`` : str, logical operator describing how to compare a value to ``threshold``.
3199
3200
              For example, ``split_feature = "Column_10", threshold = 15, decision_type = "<="`` means that
              records where ``Column_10 <= 15`` follow the left side of the split, otherwise follows the right side of the split. ``None`` for leaf nodes.
3201
3202
            - ``missing_direction`` : str, split direction that missing values should go to. ``None`` for leaf nodes.
            - ``missing_type`` : str, describes what types of values are treated as missing.
3203
            - ``value`` : float64, predicted value for this leaf node, multiplied by the learning rate.
3204
            - ``weight`` : float64 or int64, sum of Hessian (second-order derivative of objective), summed over observations that fall in this node.
3205
3206
            - ``count`` : int64, number of records in the training data that fall into this node.

3207
3208
3209
3210
3211
3212
        Returns
        -------
        result : pandas DataFrame
            Returns a pandas DataFrame of the parsed model.
        """
        if not PANDAS_INSTALLED:
3213
3214
            raise LightGBMError('This method cannot be run without pandas installed. '
                                'You must install pandas and restart your session to use this method.')
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225

        if self.num_trees() == 0:
            raise LightGBMError('There are no trees in this Booster and thus nothing to parse')

        def _is_split_node(tree):
            return 'split_index' in tree.keys()

        def create_node_record(tree, node_depth=1, tree_index=None,
                               feature_names=None, parent_node=None):

            def _get_node_index(tree, tree_index):
3226
                tree_num = f'{tree_index}-' if tree_index is not None else ''
3227
3228
3229
                is_split = _is_split_node(tree)
                node_type = 'S' if is_split else 'L'
                # if a single node tree it won't have `leaf_index` so return 0
3230
3231
                node_num = tree.get('split_index' if is_split else 'leaf_index', 0)
                return f"{tree_num}{node_type}{node_num}"
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243

            def _get_split_feature(tree, feature_names):
                if _is_split_node(tree):
                    if feature_names is not None:
                        feature_name = feature_names[tree['split_feature']]
                    else:
                        feature_name = tree['split_feature']
                else:
                    feature_name = None
                return feature_name

            def _is_single_node_tree(tree):
3244
                return set(tree.keys()) == {'leaf_value'}
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317

            # Create the node record, and populate universal data members
            node = OrderedDict()
            node['tree_index'] = tree_index
            node['node_depth'] = node_depth
            node['node_index'] = _get_node_index(tree, tree_index)
            node['left_child'] = None
            node['right_child'] = None
            node['parent_index'] = parent_node
            node['split_feature'] = _get_split_feature(tree, feature_names)
            node['split_gain'] = None
            node['threshold'] = None
            node['decision_type'] = None
            node['missing_direction'] = None
            node['missing_type'] = None
            node['value'] = None
            node['weight'] = None
            node['count'] = None

            # Update values to reflect node type (leaf or split)
            if _is_split_node(tree):
                node['left_child'] = _get_node_index(tree['left_child'], tree_index)
                node['right_child'] = _get_node_index(tree['right_child'], tree_index)
                node['split_gain'] = tree['split_gain']
                node['threshold'] = tree['threshold']
                node['decision_type'] = tree['decision_type']
                node['missing_direction'] = 'left' if tree['default_left'] else 'right'
                node['missing_type'] = tree['missing_type']
                node['value'] = tree['internal_value']
                node['weight'] = tree['internal_weight']
                node['count'] = tree['internal_count']
            else:
                node['value'] = tree['leaf_value']
                if not _is_single_node_tree(tree):
                    node['weight'] = tree['leaf_weight']
                    node['count'] = tree['leaf_count']

            return node

        def tree_dict_to_node_list(tree, node_depth=1, tree_index=None,
                                   feature_names=None, parent_node=None):

            node = create_node_record(tree,
                                      node_depth=node_depth,
                                      tree_index=tree_index,
                                      feature_names=feature_names,
                                      parent_node=parent_node)

            res = [node]

            if _is_split_node(tree):
                # traverse the next level of the tree
                children = ['left_child', 'right_child']
                for child in children:
                    subtree_list = tree_dict_to_node_list(
                        tree[child],
                        node_depth=node_depth + 1,
                        tree_index=tree_index,
                        feature_names=feature_names,
                        parent_node=node['node_index'])
                    # In tree format, "subtree_list" is a list of node records (dicts),
                    # and we add node to the list.
                    res.extend(subtree_list)
            return res

        model_dict = self.dump_model()
        feature_names = model_dict['feature_names']
        model_list = []
        for tree in model_dict['tree_info']:
            model_list.extend(tree_dict_to_node_list(tree['tree_structure'],
                                                     tree_index=tree['tree_index'],
                                                     feature_names=feature_names))

3318
        return pd_DataFrame(model_list, columns=model_list[0].keys())
3319

3320
    def set_train_data_name(self, name: str) -> "Booster":
3321
3322
3323
3324
        """Set the name to the training Dataset.

        Parameters
        ----------
3325
        name : str
Nikita Titov's avatar
Nikita Titov committed
3326
3327
3328
3329
3330
3331
            Name for the training Dataset.

        Returns
        -------
        self : Booster
            Booster with set training Dataset name.
3332
        """
3333
        self._train_data_name = name
Nikita Titov's avatar
Nikita Titov committed
3334
        return self
wxchan's avatar
wxchan committed
3335

3336
    def add_valid(self, data: Dataset, name: str) -> "Booster":
3337
        """Add validation data.
wxchan's avatar
wxchan committed
3338
3339
3340
3341

        Parameters
        ----------
        data : Dataset
3342
            Validation data.
3343
        name : str
3344
            Name of validation data.
Nikita Titov's avatar
Nikita Titov committed
3345
3346
3347
3348
3349

        Returns
        -------
        self : Booster
            Booster with set validation data.
wxchan's avatar
wxchan committed
3350
        """
Guolin Ke's avatar
Guolin Ke committed
3351
        if not isinstance(data, Dataset):
3352
            raise TypeError(f'Validation data should be Dataset instance, met {type(data).__name__}')
Guolin Ke's avatar
Guolin Ke committed
3353
        if data._predictor is not self.__init_predictor:
3354
3355
            raise LightGBMError("Add validation data failed, "
                                "you should use same predictor for these data")
wxchan's avatar
wxchan committed
3356
3357
        _safe_call(_LIB.LGBM_BoosterAddValidData(
            self.handle,
wxchan's avatar
wxchan committed
3358
            data.construct().handle))
wxchan's avatar
wxchan committed
3359
3360
3361
3362
3363
        self.valid_sets.append(data)
        self.name_valid_sets.append(name)
        self.__num_dataset += 1
        self.__inner_predict_buffer.append(None)
        self.__is_predicted_cur_iter.append(False)
Nikita Titov's avatar
Nikita Titov committed
3364
        return self
wxchan's avatar
wxchan committed
3365

3366
    def reset_parameter(self, params: Dict[str, Any]) -> "Booster":
3367
        """Reset parameters of Booster.
wxchan's avatar
wxchan committed
3368
3369
3370
3371

        Parameters
        ----------
        params : dict
3372
            New parameters for Booster.
Nikita Titov's avatar
Nikita Titov committed
3373
3374
3375
3376
3377

        Returns
        -------
        self : Booster
            Booster with new parameters.
wxchan's avatar
wxchan committed
3378
        """
3379
        params_str = _param_dict_to_str(params)
wxchan's avatar
wxchan committed
3380
3381
3382
        if params_str:
            _safe_call(_LIB.LGBM_BoosterResetParameter(
                self.handle,
3383
                _c_str(params_str)))
Guolin Ke's avatar
Guolin Ke committed
3384
        self.params.update(params)
Nikita Titov's avatar
Nikita Titov committed
3385
        return self
wxchan's avatar
wxchan committed
3386

3387
3388
3389
3390
3391
    def update(
        self,
        train_set: Optional[Dataset] = None,
        fobj: Optional[_LGBM_CustomObjectiveFunction] = None
    ) -> bool:
Nikita Titov's avatar
Nikita Titov committed
3392
        """Update Booster for one iteration.
3393

wxchan's avatar
wxchan committed
3394
3395
        Parameters
        ----------
3396
3397
3398
3399
        train_set : Dataset or None, optional (default=None)
            Training data.
            If None, last training data is used.
        fobj : callable or None, optional (default=None)
wxchan's avatar
wxchan committed
3400
            Customized objective function.
3401
3402
3403
            Should accept two parameters: preds, train_data,
            and return (grad, hess).

3404
                preds : numpy 1-D array or numpy 2-D array (for multi-class task)
3405
                    The predicted values.
3406
3407
                    Predicted values are returned before any transformation,
                    e.g. they are raw margin instead of probability of positive class for binary task.
3408
3409
                train_data : Dataset
                    The training dataset.
3410
                grad : numpy 1-D array or numpy 2-D array (for multi-class task)
3411
3412
                    The value of the first order derivative (gradient) of the loss
                    with respect to the elements of preds for each sample point.
3413
                hess : numpy 1-D array or numpy 2-D array (for multi-class task)
3414
3415
                    The value of the second order derivative (Hessian) of the loss
                    with respect to the elements of preds for each sample point.
wxchan's avatar
wxchan committed
3416

3417
            For multi-class task, preds are numpy 2-D array of shape = [n_samples, n_classes],
3418
            and grad and hess should be returned in the same format.
3419

wxchan's avatar
wxchan committed
3420
3421
        Returns
        -------
3422
3423
        is_finished : bool
            Whether the update was successfully finished.
wxchan's avatar
wxchan committed
3424
        """
3425
        # need reset training data
3426
3427
3428
3429
3430
3431
        if train_set is None and self.train_set_version != self.train_set.version:
            train_set = self.train_set
            is_the_same_train_set = False
        else:
            is_the_same_train_set = train_set is self.train_set and self.train_set_version == train_set.version
        if train_set is not None and not is_the_same_train_set:
Guolin Ke's avatar
Guolin Ke committed
3432
            if not isinstance(train_set, Dataset):
3433
                raise TypeError(f'Training data should be Dataset instance, met {type(train_set).__name__}')
Guolin Ke's avatar
Guolin Ke committed
3434
            if train_set._predictor is not self.__init_predictor:
3435
3436
                raise LightGBMError("Replace training data failed, "
                                    "you should use same predictor for these data")
wxchan's avatar
wxchan committed
3437
3438
3439
            self.train_set = train_set
            _safe_call(_LIB.LGBM_BoosterResetTrainingData(
                self.handle,
wxchan's avatar
wxchan committed
3440
                self.train_set.construct().handle))
wxchan's avatar
wxchan committed
3441
            self.__inner_predict_buffer[0] = None
3442
            self.train_set_version = self.train_set.version
wxchan's avatar
wxchan committed
3443
3444
        is_finished = ctypes.c_int(0)
        if fobj is None:
3445
            if self.__set_objective_to_none:
3446
                raise LightGBMError('Cannot update due to null objective function.')
wxchan's avatar
wxchan committed
3447
3448
3449
            _safe_call(_LIB.LGBM_BoosterUpdateOneIter(
                self.handle,
                ctypes.byref(is_finished)))
3450
            self.__is_predicted_cur_iter = [False for _ in range(self.__num_dataset)]
wxchan's avatar
wxchan committed
3451
3452
            return is_finished.value == 1
        else:
3453
            if not self.__set_objective_to_none:
Nikita Titov's avatar
Nikita Titov committed
3454
                self.reset_parameter({"objective": "none"}).__set_objective_to_none = True
wxchan's avatar
wxchan committed
3455
3456
3457
            grad, hess = fobj(self.__inner_predict(0), self.train_set)
            return self.__boost(grad, hess)

3458
3459
3460
3461
3462
    def __boost(
        self,
        grad: np.ndarray,
        hess: np.ndarray
    ) -> bool:
3463
        """Boost Booster for one iteration with customized gradient statistics.
Nikita Titov's avatar
Nikita Titov committed
3464

Nikita Titov's avatar
Nikita Titov committed
3465
3466
        .. note::

3467
3468
            Score is returned before any transformation,
            e.g. it is raw margin instead of probability of positive class for binary task.
3469
            For multi-class task, score are numpy 2-D array of shape = [n_samples, n_classes],
3470
            and grad and hess should be returned in the same format.
3471

wxchan's avatar
wxchan committed
3472
3473
        Parameters
        ----------
3474
        grad : numpy 1-D array or numpy 2-D array (for multi-class task)
3475
3476
            The value of the first order derivative (gradient) of the loss
            with respect to the elements of score for each sample point.
3477
        hess : numpy 1-D array or numpy 2-D array (for multi-class task)
3478
3479
            The value of the second order derivative (Hessian) of the loss
            with respect to the elements of score for each sample point.
wxchan's avatar
wxchan committed
3480
3481
3482

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3483
3484
        is_finished : bool
            Whether the boost was successfully finished.
wxchan's avatar
wxchan committed
3485
        """
3486
3487
3488
        if self.__num_class > 1:
            grad = grad.ravel(order='F')
            hess = hess.ravel(order='F')
3489
3490
        grad = _list_to_1d_numpy(grad, name='gradient')
        hess = _list_to_1d_numpy(hess, name='hessian')
3491
3492
        assert grad.flags.c_contiguous
        assert hess.flags.c_contiguous
wxchan's avatar
wxchan committed
3493
        if len(grad) != len(hess):
3494
3495
            raise ValueError(f"Lengths of gradient ({len(grad)}) and Hessian ({len(hess)}) don't match")
        num_train_data = self.train_set.num_data()
3496
        if len(grad) != num_train_data * self.__num_class:
3497
3498
3499
            raise ValueError(
                f"Lengths of gradient ({len(grad)}) and Hessian ({len(hess)}) "
                f"don't match training data length ({num_train_data}) * "
3500
                f"number of models per one iteration ({self.__num_class})"
3501
            )
wxchan's avatar
wxchan committed
3502
3503
3504
3505
3506
3507
        is_finished = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterUpdateOneIterCustom(
            self.handle,
            grad.ctypes.data_as(ctypes.POINTER(ctypes.c_float)),
            hess.ctypes.data_as(ctypes.POINTER(ctypes.c_float)),
            ctypes.byref(is_finished)))
3508
        self.__is_predicted_cur_iter = [False for _ in range(self.__num_dataset)]
wxchan's avatar
wxchan committed
3509
3510
        return is_finished.value == 1

3511
    def rollback_one_iter(self) -> "Booster":
Nikita Titov's avatar
Nikita Titov committed
3512
3513
3514
3515
3516
3517
3518
        """Rollback one iteration.

        Returns
        -------
        self : Booster
            Booster with rolled back one iteration.
        """
wxchan's avatar
wxchan committed
3519
3520
        _safe_call(_LIB.LGBM_BoosterRollbackOneIter(
            self.handle))
3521
        self.__is_predicted_cur_iter = [False for _ in range(self.__num_dataset)]
Nikita Titov's avatar
Nikita Titov committed
3522
        return self
wxchan's avatar
wxchan committed
3523

3524
    def current_iteration(self) -> int:
3525
3526
3527
3528
3529
3530
3531
        """Get the index of the current iteration.

        Returns
        -------
        cur_iter : int
            The index of the current iteration.
        """
Guolin Ke's avatar
Guolin Ke committed
3532
        out_cur_iter = ctypes.c_int(0)
wxchan's avatar
wxchan committed
3533
3534
3535
3536
3537
        _safe_call(_LIB.LGBM_BoosterGetCurrentIteration(
            self.handle,
            ctypes.byref(out_cur_iter)))
        return out_cur_iter.value

3538
    def num_model_per_iteration(self) -> int:
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
        """Get number of models per iteration.

        Returns
        -------
        model_per_iter : int
            The number of models per iteration.
        """
        model_per_iter = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterNumModelPerIteration(
            self.handle,
            ctypes.byref(model_per_iter)))
        return model_per_iter.value

3552
    def num_trees(self) -> int:
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
        """Get number of weak sub-models.

        Returns
        -------
        num_trees : int
            The number of weak sub-models.
        """
        num_trees = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterNumberOfTotalModel(
            self.handle,
            ctypes.byref(num_trees)))
        return num_trees.value

3566
    def upper_bound(self) -> float:
3567
3568
3569
3570
        """Get upper bound value of a model.

        Returns
        -------
3571
        upper_bound : float
3572
3573
3574
3575
3576
3577
3578
3579
            Upper bound value of the model.
        """
        ret = ctypes.c_double(0)
        _safe_call(_LIB.LGBM_BoosterGetUpperBoundValue(
            self.handle,
            ctypes.byref(ret)))
        return ret.value

3580
    def lower_bound(self) -> float:
3581
3582
3583
3584
        """Get lower bound value of a model.

        Returns
        -------
3585
        lower_bound : float
3586
3587
3588
3589
3590
3591
3592
3593
            Lower bound value of the model.
        """
        ret = ctypes.c_double(0)
        _safe_call(_LIB.LGBM_BoosterGetLowerBoundValue(
            self.handle,
            ctypes.byref(ret)))
        return ret.value

3594
3595
3596
3597
3598
3599
    def eval(
        self,
        data: Dataset,
        name: str,
        feval: Optional[Union[_LGBM_CustomEvalFunction, List[_LGBM_CustomEvalFunction]]] = None
    ) -> List[_LGBM_BoosterEvalMethodResultType]:
3600
        """Evaluate for data.
wxchan's avatar
wxchan committed
3601
3602
3603

        Parameters
        ----------
3604
3605
        data : Dataset
            Data for the evaluating.
3606
        name : str
3607
            Name of the data.
3608
        feval : callable, list of callable, or None, optional (default=None)
3609
            Customized evaluation function.
3610
            Each evaluation function should accept two parameters: preds, eval_data,
3611
            and return (eval_name, eval_result, is_higher_better) or list of such tuples.
3612

3613
                preds : numpy 1-D array or numpy 2-D array (for multi-class task)
3614
                    The predicted values.
3615
                    For multi-class task, preds are numpy 2-D array of shape = [n_samples, n_classes].
3616
                    If custom objective function is used, predicted values are returned before any transformation,
3617
                    e.g. they are raw margin instead of probability of positive class for binary task in this case.
3618
                eval_data : Dataset
3619
                    A ``Dataset`` to evaluate.
3620
                eval_name : str
Andrew Ziem's avatar
Andrew Ziem committed
3621
                    The name of evaluation function (without whitespace).
3622
3623
3624
3625
3626
                eval_result : float
                    The eval result.
                is_higher_better : bool
                    Is eval result higher better, e.g. AUC is ``is_higher_better``.

wxchan's avatar
wxchan committed
3627
3628
        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3629
        result : list
3630
            List with (dataset_name, eval_name, eval_result, is_higher_better) tuples.
wxchan's avatar
wxchan committed
3631
        """
Guolin Ke's avatar
Guolin Ke committed
3632
3633
        if not isinstance(data, Dataset):
            raise TypeError("Can only eval for Dataset instance")
wxchan's avatar
wxchan committed
3634
3635
3636
3637
        data_idx = -1
        if data is self.train_set:
            data_idx = 0
        else:
3638
            for i in range(len(self.valid_sets)):
wxchan's avatar
wxchan committed
3639
3640
3641
                if data is self.valid_sets[i]:
                    data_idx = i + 1
                    break
3642
        # need to push new valid data
wxchan's avatar
wxchan committed
3643
3644
3645
3646
3647
3648
        if data_idx == -1:
            self.add_valid(data, name)
            data_idx = self.__num_dataset - 1

        return self.__inner_eval(name, data_idx, feval)

3649
3650
3651
3652
    def eval_train(
        self,
        feval: Optional[Union[_LGBM_CustomEvalFunction, List[_LGBM_CustomEvalFunction]]] = None
    ) -> List[_LGBM_BoosterEvalMethodResultType]:
3653
        """Evaluate for training data.
wxchan's avatar
wxchan committed
3654
3655
3656

        Parameters
        ----------
3657
        feval : callable, list of callable, or None, optional (default=None)
3658
            Customized evaluation function.
3659
            Each evaluation function should accept two parameters: preds, eval_data,
3660
            and return (eval_name, eval_result, is_higher_better) or list of such tuples.
3661

3662
                preds : numpy 1-D array or numpy 2-D array (for multi-class task)
3663
                    The predicted values.
3664
                    For multi-class task, preds are numpy 2-D array of shape = [n_samples, n_classes].
3665
                    If custom objective function is used, predicted values are returned before any transformation,
3666
                    e.g. they are raw margin instead of probability of positive class for binary task in this case.
Akshita Dixit's avatar
Akshita Dixit committed
3667
                eval_data : Dataset
3668
                    The training dataset.
3669
                eval_name : str
Andrew Ziem's avatar
Andrew Ziem committed
3670
                    The name of evaluation function (without whitespace).
3671
3672
3673
3674
3675
                eval_result : float
                    The eval result.
                is_higher_better : bool
                    Is eval result higher better, e.g. AUC is ``is_higher_better``.

wxchan's avatar
wxchan committed
3676
3677
        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3678
        result : list
3679
            List with (train_dataset_name, eval_name, eval_result, is_higher_better) tuples.
wxchan's avatar
wxchan committed
3680
        """
3681
        return self.__inner_eval(self._train_data_name, 0, feval)
wxchan's avatar
wxchan committed
3682

3683
3684
3685
3686
    def eval_valid(
        self,
        feval: Optional[Union[_LGBM_CustomEvalFunction, List[_LGBM_CustomEvalFunction]]] = None
    ) -> List[_LGBM_BoosterEvalMethodResultType]:
3687
        """Evaluate for validation data.
wxchan's avatar
wxchan committed
3688
3689
3690

        Parameters
        ----------
3691
        feval : callable, list of callable, or None, optional (default=None)
3692
            Customized evaluation function.
3693
            Each evaluation function should accept two parameters: preds, eval_data,
3694
            and return (eval_name, eval_result, is_higher_better) or list of such tuples.
3695

3696
                preds : numpy 1-D array or numpy 2-D array (for multi-class task)
3697
                    The predicted values.
3698
                    For multi-class task, preds are numpy 2-D array of shape = [n_samples, n_classes].
3699
                    If custom objective function is used, predicted values are returned before any transformation,
3700
                    e.g. they are raw margin instead of probability of positive class for binary task in this case.
Akshita Dixit's avatar
Akshita Dixit committed
3701
                eval_data : Dataset
3702
                    The validation dataset.
3703
                eval_name : str
Andrew Ziem's avatar
Andrew Ziem committed
3704
                    The name of evaluation function (without whitespace).
3705
3706
3707
3708
3709
                eval_result : float
                    The eval result.
                is_higher_better : bool
                    Is eval result higher better, e.g. AUC is ``is_higher_better``.

wxchan's avatar
wxchan committed
3710
3711
        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3712
        result : list
3713
            List with (validation_dataset_name, eval_name, eval_result, is_higher_better) tuples.
wxchan's avatar
wxchan committed
3714
        """
3715
        return [item for i in range(1, self.__num_dataset)
wxchan's avatar
wxchan committed
3716
                for item in self.__inner_eval(self.name_valid_sets[i - 1], i, feval)]
wxchan's avatar
wxchan committed
3717

3718
3719
3720
3721
3722
3723
3724
    def save_model(
        self,
        filename: Union[str, Path],
        num_iteration: Optional[int] = None,
        start_iteration: int = 0,
        importance_type: str = 'split'
    ) -> "Booster":
3725
        """Save Booster to file.
wxchan's avatar
wxchan committed
3726
3727
3728

        Parameters
        ----------
3729
        filename : str or pathlib.Path
3730
            Filename to save Booster.
3731
3732
3733
3734
        num_iteration : int or None, optional (default=None)
            Index of the iteration that should be saved.
            If None, if the best iteration exists, it is saved; otherwise, all iterations are saved.
            If <= 0, all iterations are saved.
Nikita Titov's avatar
Nikita Titov committed
3735
        start_iteration : int, optional (default=0)
3736
            Start index of the iteration that should be saved.
3737
        importance_type : str, optional (default="split")
3738
3739
3740
            What type of feature importance should be saved.
            If "split", result contains numbers of times the feature is used in a model.
            If "gain", result contains total gains of splits which use the feature.
Nikita Titov's avatar
Nikita Titov committed
3741
3742
3743
3744
3745

        Returns
        -------
        self : Booster
            Returns self.
wxchan's avatar
wxchan committed
3746
        """
3747
        if num_iteration is None:
3748
            num_iteration = self.best_iteration
3749
        importance_type_int = _FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
wxchan's avatar
wxchan committed
3750
3751
        _safe_call(_LIB.LGBM_BoosterSaveModel(
            self.handle,
3752
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
3753
            ctypes.c_int(num_iteration),
3754
            ctypes.c_int(importance_type_int),
3755
            _c_str(str(filename))))
3756
        _dump_pandas_categorical(self.pandas_categorical, filename)
Nikita Titov's avatar
Nikita Titov committed
3757
        return self
wxchan's avatar
wxchan committed
3758

3759
3760
3761
3762
3763
    def shuffle_models(
        self,
        start_iteration: int = 0,
        end_iteration: int = -1
    ) -> "Booster":
3764
        """Shuffle models.
Nikita Titov's avatar
Nikita Titov committed
3765

3766
3767
3768
        Parameters
        ----------
        start_iteration : int, optional (default=0)
3769
            The first iteration that will be shuffled.
3770
3771
        end_iteration : int, optional (default=-1)
            The last iteration that will be shuffled.
3772
            If <= 0, means the last available iteration.
3773

Nikita Titov's avatar
Nikita Titov committed
3774
3775
3776
3777
        Returns
        -------
        self : Booster
            Booster with shuffled models.
3778
        """
3779
3780
        _safe_call(_LIB.LGBM_BoosterShuffleModels(
            self.handle,
Guolin Ke's avatar
Guolin Ke committed
3781
3782
            ctypes.c_int(start_iteration),
            ctypes.c_int(end_iteration)))
Nikita Titov's avatar
Nikita Titov committed
3783
        return self
3784

3785
    def model_from_string(self, model_str: str) -> "Booster":
3786
3787
3788
3789
        """Load Booster from a string.

        Parameters
        ----------
3790
        model_str : str
3791
3792
3793
3794
            Model will be loaded from this string.

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3795
        self : Booster
3796
3797
            Loaded Booster object.
        """
3798
3799
3800
3801
        if self.handle is not None:
            _safe_call(_LIB.LGBM_BoosterFree(self.handle))
        self._free_buffer()
        self.handle = ctypes.c_void_p()
3802
3803
        out_num_iterations = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterLoadModelFromString(
3804
            _c_str(model_str),
3805
3806
3807
3808
3809
3810
3811
            ctypes.byref(out_num_iterations),
            ctypes.byref(self.handle)))
        out_num_class = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterGetNumClasses(
            self.handle,
            ctypes.byref(out_num_class)))
        self.__num_class = out_num_class.value
3812
        self.pandas_categorical = _load_pandas_categorical(model_str=model_str)
3813
3814
        return self

3815
3816
3817
3818
3819
3820
    def model_to_string(
        self,
        num_iteration: Optional[int] = None,
        start_iteration: int = 0,
        importance_type: str = 'split'
    ) -> str:
3821
        """Save Booster to string.
3822

3823
3824
3825
3826
3827
3828
        Parameters
        ----------
        num_iteration : int or None, optional (default=None)
            Index of the iteration that should be saved.
            If None, if the best iteration exists, it is saved; otherwise, all iterations are saved.
            If <= 0, all iterations are saved.
Nikita Titov's avatar
Nikita Titov committed
3829
        start_iteration : int, optional (default=0)
3830
            Start index of the iteration that should be saved.
3831
        importance_type : str, optional (default="split")
3832
3833
3834
            What type of feature importance should be saved.
            If "split", result contains numbers of times the feature is used in a model.
            If "gain", result contains total gains of splits which use the feature.
3835
3836
3837

        Returns
        -------
3838
        str_repr : str
3839
3840
            String representation of Booster.
        """
3841
        if num_iteration is None:
3842
            num_iteration = self.best_iteration
3843
        importance_type_int = _FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
3844
        buffer_len = 1 << 20
3845
        tmp_out_len = ctypes.c_int64(0)
3846
3847
3848
3849
        string_buffer = ctypes.create_string_buffer(buffer_len)
        ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
        _safe_call(_LIB.LGBM_BoosterSaveModelToString(
            self.handle,
3850
            ctypes.c_int(start_iteration),
3851
            ctypes.c_int(num_iteration),
3852
            ctypes.c_int(importance_type_int),
3853
            ctypes.c_int64(buffer_len),
3854
3855
3856
            ctypes.byref(tmp_out_len),
            ptr_string_buffer))
        actual_len = tmp_out_len.value
3857
        # if buffer length is not long enough, re-allocate a buffer
3858
3859
3860
3861
3862
        if actual_len > buffer_len:
            string_buffer = ctypes.create_string_buffer(actual_len)
            ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
            _safe_call(_LIB.LGBM_BoosterSaveModelToString(
                self.handle,
3863
                ctypes.c_int(start_iteration),
3864
                ctypes.c_int(num_iteration),
3865
                ctypes.c_int(importance_type_int),
3866
                ctypes.c_int64(actual_len),
3867
3868
                ctypes.byref(tmp_out_len),
                ptr_string_buffer))
3869
        ret = string_buffer.value.decode('utf-8')
3870
3871
        ret += _dump_pandas_categorical(self.pandas_categorical)
        return ret
3872

3873
3874
3875
3876
3877
3878
3879
    def dump_model(
        self,
        num_iteration: Optional[int] = None,
        start_iteration: int = 0,
        importance_type: str = 'split',
        object_hook: Optional[Callable[[Dict[str, Any]], Dict[str, Any]]] = None
    ) -> Dict[str, Any]:
Nikita Titov's avatar
Nikita Titov committed
3880
        """Dump Booster to JSON format.
wxchan's avatar
wxchan committed
3881

3882
3883
        Parameters
        ----------
3884
3885
3886
3887
        num_iteration : int or None, optional (default=None)
            Index of the iteration that should be dumped.
            If None, if the best iteration exists, it is dumped; otherwise, all iterations are dumped.
            If <= 0, all iterations are dumped.
Nikita Titov's avatar
Nikita Titov committed
3888
        start_iteration : int, optional (default=0)
3889
            Start index of the iteration that should be dumped.
3890
        importance_type : str, optional (default="split")
3891
3892
3893
            What type of feature importance should be dumped.
            If "split", result contains numbers of times the feature is used in a model.
            If "gain", result contains total gains of splits which use the feature.
3894
3895
3896
3897
3898
3899
3900
3901
3902
        object_hook : callable or None, optional (default=None)
            If not None, ``object_hook`` is a function called while parsing the json
            string returned by the C API. It may be used to alter the json, to store
            specific values while building the json structure. It avoids
            walking through the structure again. It saves a significant amount
            of time if the number of trees is huge.
            Signature is ``def object_hook(node: dict) -> dict``.
            None is equivalent to ``lambda node: node``.
            See documentation of ``json.loads()`` for further details.
3903

wxchan's avatar
wxchan committed
3904
3905
        Returns
        -------
3906
        json_repr : dict
Nikita Titov's avatar
Nikita Titov committed
3907
            JSON format of Booster.
wxchan's avatar
wxchan committed
3908
        """
3909
        if num_iteration is None:
3910
            num_iteration = self.best_iteration
3911
        importance_type_int = _FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
wxchan's avatar
wxchan committed
3912
        buffer_len = 1 << 20
3913
        tmp_out_len = ctypes.c_int64(0)
wxchan's avatar
wxchan committed
3914
3915
3916
3917
        string_buffer = ctypes.create_string_buffer(buffer_len)
        ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
        _safe_call(_LIB.LGBM_BoosterDumpModel(
            self.handle,
3918
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
3919
            ctypes.c_int(num_iteration),
3920
            ctypes.c_int(importance_type_int),
3921
            ctypes.c_int64(buffer_len),
wxchan's avatar
wxchan committed
3922
            ctypes.byref(tmp_out_len),
Guolin Ke's avatar
Guolin Ke committed
3923
            ptr_string_buffer))
wxchan's avatar
wxchan committed
3924
        actual_len = tmp_out_len.value
3925
        # if buffer length is not long enough, reallocate a buffer
wxchan's avatar
wxchan committed
3926
3927
3928
3929
3930
        if actual_len > buffer_len:
            string_buffer = ctypes.create_string_buffer(actual_len)
            ptr_string_buffer = ctypes.c_char_p(*[ctypes.addressof(string_buffer)])
            _safe_call(_LIB.LGBM_BoosterDumpModel(
                self.handle,
3931
                ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
3932
                ctypes.c_int(num_iteration),
3933
                ctypes.c_int(importance_type_int),
3934
                ctypes.c_int64(actual_len),
wxchan's avatar
wxchan committed
3935
                ctypes.byref(tmp_out_len),
Guolin Ke's avatar
Guolin Ke committed
3936
                ptr_string_buffer))
3937
        ret = json.loads(string_buffer.value.decode('utf-8'), object_hook=object_hook)
3938
        ret['pandas_categorical'] = json.loads(json.dumps(self.pandas_categorical,
3939
                                                          default=_json_default_with_numpy))
3940
        return ret
wxchan's avatar
wxchan committed
3941

3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
    def predict(
        self,
        data,
        start_iteration: int = 0,
        num_iteration: Optional[int] = None,
        raw_score: bool = False,
        pred_leaf: bool = False,
        pred_contrib: bool = False,
        data_has_header: bool = False,
        validate_features: bool = False,
        **kwargs: Any
    ):
3954
        """Make a prediction.
wxchan's avatar
wxchan committed
3955
3956
3957

        Parameters
        ----------
3958
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame or scipy.sparse
3959
            Data source for prediction.
3960
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM).
3961
        start_iteration : int, optional (default=0)
3962
            Start index of the iteration to predict.
3963
            If <= 0, starts from the first iteration.
3964
        num_iteration : int or None, optional (default=None)
3965
3966
3967
3968
            Total number of iterations used in the prediction.
            If None, if the best iteration exists and start_iteration <= 0, the best iteration is used;
            otherwise, all iterations from ``start_iteration`` are used (no limits).
            If <= 0, all iterations from ``start_iteration`` are used (no limits).
3969
3970
3971
3972
        raw_score : bool, optional (default=False)
            Whether to predict raw scores.
        pred_leaf : bool, optional (default=False)
            Whether to predict leaf index.
3973
3974
        pred_contrib : bool, optional (default=False)
            Whether to predict feature contributions.
3975

Nikita Titov's avatar
Nikita Titov committed
3976
3977
3978
3979
3980
3981
3982
            .. note::

                If you want to get more explanations for your model's predictions using SHAP values,
                like SHAP interaction values,
                you can install the shap package (https://github.com/slundberg/shap).
                Note that unlike the shap package, with ``pred_contrib`` we return a matrix with an extra
                column, where the last column is the expected value.
3983

3984
3985
        data_has_header : bool, optional (default=False)
            Whether the data has header.
3986
            Used only if data is str.
3987
3988
3989
        validate_features : bool, optional (default=False)
            If True, ensure that the features used to predict match the ones used to train.
            Used only if data is pandas DataFrame.
3990
3991
        **kwargs
            Other parameters for the prediction.
wxchan's avatar
wxchan committed
3992
3993
3994

        Returns
        -------
3995
        result : numpy array, scipy.sparse or list of scipy.sparse
3996
            Prediction result.
3997
            Can be sparse or a list of sparse objects (each element represents predictions for one class) for feature contributions (when ``pred_contrib=True``).
wxchan's avatar
wxchan committed
3998
        """
3999
        predictor = self._to_predictor(deepcopy(kwargs))
4000
        if num_iteration is None:
4001
            if start_iteration <= 0:
4002
4003
4004
4005
                num_iteration = self.best_iteration
            else:
                num_iteration = -1
        return predictor.predict(data, start_iteration, num_iteration,
4006
                                 raw_score, pred_leaf, pred_contrib,
4007
                                 data_has_header, validate_features)
wxchan's avatar
wxchan committed
4008

4009
4010
4011
4012
    def refit(
        self,
        data,
        label,
4013
4014
        decay_rate: float = 0.9,
        reference: Optional[Dataset] = None,
4015
4016
4017
        weight=None,
        group=None,
        init_score=None,
4018
4019
        feature_name: _LGBM_FeatureNameConfiguration = 'auto',
        categorical_feature: _LGBM_CategoricalFeatureConfiguration = 'auto',
4020
4021
4022
        dataset_params: Optional[Dict[str, Any]] = None,
        free_raw_data: bool = True,
        validate_features: bool = False,
4023
4024
        **kwargs
    ):
Guolin Ke's avatar
Guolin Ke committed
4025
4026
4027
4028
        """Refit the existing Booster by new data.

        Parameters
        ----------
4029
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame or scipy.sparse
Guolin Ke's avatar
Guolin Ke committed
4030
            Data source for refit.
4031
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM).
4032
        label : list, numpy 1-D array or pandas Series / one-column DataFrame
Guolin Ke's avatar
Guolin Ke committed
4033
4034
            Label for refit.
        decay_rate : float, optional (default=0.9)
4035
4036
            Decay rate of refit,
            will use ``leaf_output = decay_rate * old_leaf_output + (1.0 - decay_rate) * new_leaf_output`` to refit trees.
4037
4038
4039
        reference : Dataset or None, optional (default=None)
            Reference for ``data``.
        weight : list, numpy 1-D array, pandas Series or None, optional (default=None)
4040
            Weight for each ``data`` instance. Weights should be non-negative.
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
        group : list, numpy 1-D array, pandas Series or None, optional (default=None)
            Group/query size for ``data``.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
            For example, if you have a 100-document dataset with ``group = [10, 20, 40, 10, 10, 10]``, that means that you have 6 groups,
            where the first 10 records are in the first group, records 11-30 are in the second group, records 31-70 are in the third group, etc.
        init_score : list, list of lists (for multi-class task), numpy array, pandas Series, pandas DataFrame (for multi-class task), or None, optional (default=None)
            Init score for ``data``.
        feature_name : list of str, or 'auto', optional (default="auto")
            Feature names for ``data``.
            If 'auto' and data is pandas DataFrame, data columns names are used.
        categorical_feature : list of str or int, or 'auto', optional (default="auto")
            Categorical features for ``data``.
            If list of int, interpreted as indices.
            If list of str, interpreted as feature names (need to specify ``feature_name`` as well).
            If 'auto' and data is pandas DataFrame, pandas unordered categorical columns are used.
4057
            All values in categorical features will be cast to int32 and thus should be less than int32 max value (2147483647).
4058
4059
4060
            Large values could be memory consuming. Consider using consecutive integers starting from zero.
            All negative values in categorical features will be treated as missing values.
            The output cannot be monotonically constrained with respect to a categorical feature.
4061
            Floating point numbers in categorical features will be rounded towards 0.
4062
4063
4064
4065
        dataset_params : dict or None, optional (default=None)
            Other parameters for Dataset ``data``.
        free_raw_data : bool, optional (default=True)
            If True, raw data is freed after constructing inner Dataset for ``data``.
4066
4067
4068
        validate_features : bool, optional (default=False)
            If True, ensure that the features used to refit the model match the original ones.
            Used only if data is pandas DataFrame.
4069
4070
        **kwargs
            Other parameters for refit.
4071
            These parameters will be passed to ``predict`` method.
Guolin Ke's avatar
Guolin Ke committed
4072
4073
4074
4075
4076
4077

        Returns
        -------
        result : Booster
            Refitted Booster.
        """
4078
4079
        if self.__set_objective_to_none:
            raise LightGBMError('Cannot refit due to null objective function.')
4080
4081
        if dataset_params is None:
            dataset_params = {}
4082
        predictor = self._to_predictor(deepcopy(kwargs))
4083
        leaf_preds = predictor.predict(data, -1, pred_leaf=True, validate_features=validate_features)
4084
        nrow, ncol = leaf_preds.shape
4085
        out_is_linear = ctypes.c_int(0)
4086
4087
4088
        _safe_call(_LIB.LGBM_BoosterGetLinear(
            self.handle,
            ctypes.byref(out_is_linear)))
Nikita Titov's avatar
Nikita Titov committed
4089
4090
4091
4092
4093
        new_params = _choose_param_value(
            main_param_name="linear_tree",
            params=self.params,
            default_value=None
        )
4094
        new_params["linear_tree"] = bool(out_is_linear.value)
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
        new_params.update(dataset_params)
        train_set = Dataset(
            data=data,
            label=label,
            reference=reference,
            weight=weight,
            group=group,
            init_score=init_score,
            feature_name=feature_name,
            categorical_feature=categorical_feature,
            params=new_params,
            free_raw_data=free_raw_data,
        )
4108
        new_params['refit_decay_rate'] = decay_rate
4109
        new_booster = Booster(new_params, train_set)
Guolin Ke's avatar
Guolin Ke committed
4110
4111
4112
4113
4114
        # Copy models
        _safe_call(_LIB.LGBM_BoosterMerge(
            new_booster.handle,
            predictor.handle))
        leaf_preds = leaf_preds.reshape(-1)
4115
        ptr_data, _, _ = _c_int_array(leaf_preds)
Guolin Ke's avatar
Guolin Ke committed
4116
4117
4118
        _safe_call(_LIB.LGBM_BoosterRefit(
            new_booster.handle,
            ptr_data,
4119
4120
            ctypes.c_int32(nrow),
            ctypes.c_int32(ncol)))
4121
        new_booster._network = self._network
Guolin Ke's avatar
Guolin Ke committed
4122
4123
        return new_booster

4124
    def get_leaf_output(self, tree_id: int, leaf_id: int) -> float:
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
        """Get the output of a leaf.

        Parameters
        ----------
        tree_id : int
            The index of the tree.
        leaf_id : int
            The index of the leaf in the tree.

        Returns
        -------
        result : float
            The output of the leaf.
        """
4139
4140
4141
4142
4143
4144
4145
4146
        ret = ctypes.c_double(0)
        _safe_call(_LIB.LGBM_BoosterGetLeafValue(
            self.handle,
            ctypes.c_int(tree_id),
            ctypes.c_int(leaf_id),
            ctypes.byref(ret)))
        return ret.value

4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
    def set_leaf_output(
        self,
        tree_id: int,
        leaf_id: int,
        value: float,
    ) -> 'Booster':
        """Set the output of a leaf.

        Parameters
        ----------
        tree_id : int
            The index of the tree.
        leaf_id : int
            The index of the leaf in the tree.
        value : float
            Value to set as the output of the leaf.

        Returns
        -------
        self : Booster
            Booster with the leaf output set.
        """
        _safe_call(
            _LIB.LGBM_BoosterSetLeafValue(
                self.handle,
                ctypes.c_int(tree_id),
                ctypes.c_int(leaf_id),
                ctypes.c_double(value)
            )
        )
        return self

4179
4180
4181
4182
    def _to_predictor(
        self,
        pred_parameter: Optional[Dict[str, Any]] = None
    ) -> _InnerPredictor:
4183
        """Convert to predictor."""
4184
        predictor = _InnerPredictor(booster_handle=self.handle, pred_parameter=pred_parameter)
4185
        predictor.pandas_categorical = self.pandas_categorical
wxchan's avatar
wxchan committed
4186
4187
        return predictor

4188
    def num_feature(self) -> int:
4189
4190
4191
4192
4193
4194
4195
        """Get number of features.

        Returns
        -------
        num_feature : int
            The number of features.
        """
4196
4197
4198
4199
4200
4201
        out_num_feature = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterGetNumFeature(
            self.handle,
            ctypes.byref(out_num_feature)))
        return out_num_feature.value

4202
    def feature_name(self) -> List[str]:
4203
        """Get names of features.
wxchan's avatar
wxchan committed
4204
4205
4206

        Returns
        -------
4207
        result : list of str
4208
            List with names of features.
wxchan's avatar
wxchan committed
4209
        """
4210
        num_feature = self.num_feature()
4211
        # Get name of features
wxchan's avatar
wxchan committed
4212
        tmp_out_len = ctypes.c_int(0)
4213
4214
        reserved_string_buffer_size = 255
        required_string_buffer_size = ctypes.c_size_t(0)
4215
        string_buffers = [ctypes.create_string_buffer(reserved_string_buffer_size) for _ in range(num_feature)]
wxchan's avatar
wxchan committed
4216
4217
4218
        ptr_string_buffers = (ctypes.c_char_p * num_feature)(*map(ctypes.addressof, string_buffers))
        _safe_call(_LIB.LGBM_BoosterGetFeatureNames(
            self.handle,
4219
            ctypes.c_int(num_feature),
wxchan's avatar
wxchan committed
4220
            ctypes.byref(tmp_out_len),
4221
            ctypes.c_size_t(reserved_string_buffer_size),
4222
            ctypes.byref(required_string_buffer_size),
wxchan's avatar
wxchan committed
4223
4224
4225
            ptr_string_buffers))
        if num_feature != tmp_out_len.value:
            raise ValueError("Length of feature names doesn't equal with num_feature")
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
        actual_string_buffer_size = required_string_buffer_size.value
        # if buffer length is not long enough, reallocate buffers
        if reserved_string_buffer_size < actual_string_buffer_size:
            string_buffers = [ctypes.create_string_buffer(actual_string_buffer_size) for _ in range(num_feature)]
            ptr_string_buffers = (ctypes.c_char_p * num_feature)(*map(ctypes.addressof, string_buffers))
            _safe_call(_LIB.LGBM_BoosterGetFeatureNames(
                self.handle,
                ctypes.c_int(num_feature),
                ctypes.byref(tmp_out_len),
                ctypes.c_size_t(actual_string_buffer_size),
                ctypes.byref(required_string_buffer_size),
                ptr_string_buffers))
4238
        return [string_buffers[i].value.decode('utf-8') for i in range(num_feature)]
wxchan's avatar
wxchan committed
4239

4240
4241
4242
4243
4244
    def feature_importance(
        self,
        importance_type: str = 'split',
        iteration: Optional[int] = None
    ) -> np.ndarray:
4245
        """Get feature importances.
4246

4247
4248
        Parameters
        ----------
4249
        importance_type : str, optional (default="split")
4250
4251
4252
            How the importance is calculated.
            If "split", result contains numbers of times the feature is used in a model.
            If "gain", result contains total gains of splits which use the feature.
4253
4254
4255
4256
        iteration : int or None, optional (default=None)
            Limit number of iterations in the feature importance calculation.
            If None, if the best iteration exists, it is used; otherwise, all trees are used.
            If <= 0, all trees are used (no limits).
4257

4258
4259
        Returns
        -------
4260
4261
        result : numpy array
            Array with feature importances.
4262
        """
4263
4264
        if iteration is None:
            iteration = self.best_iteration
4265
        importance_type_int = _FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
4266
        result = np.empty(self.num_feature(), dtype=np.float64)
4267
4268
4269
4270
4271
        _safe_call(_LIB.LGBM_BoosterFeatureImportance(
            self.handle,
            ctypes.c_int(iteration),
            ctypes.c_int(importance_type_int),
            result.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
4272
        if importance_type_int == _C_API_FEATURE_IMPORTANCE_SPLIT:
4273
            return result.astype(np.int32)
4274
4275
        else:
            return result
4276

4277
4278
4279
4280
4281
4282
    def get_split_value_histogram(
        self,
        feature: Union[int, str],
        bins: Optional[Union[int, str]] = None,
        xgboost_style: bool = False
    ) -> Union[Tuple[np.ndarray, np.ndarray], np.ndarray, pd_DataFrame]:
4283
4284
4285
4286
        """Get split value histogram for the specified feature.

        Parameters
        ----------
4287
        feature : int or str
4288
4289
            The feature name or index the histogram is calculated for.
            If int, interpreted as index.
4290
            If str, interpreted as name.
4291

Nikita Titov's avatar
Nikita Titov committed
4292
4293
4294
            .. warning::

                Categorical features are not supported.
4295

4296
        bins : int, str or None, optional (default=None)
4297
4298
4299
            The maximum number of bins.
            If None, or int and > number of unique split values and ``xgboost_style=True``,
            the number of bins equals number of unique split values.
4300
            If str, it should be one from the list of the supported values by ``numpy.histogram()`` function.
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
        xgboost_style : bool, optional (default=False)
            Whether the returned result should be in the same form as it is in XGBoost.
            If False, the returned value is tuple of 2 numpy arrays as it is in ``numpy.histogram()`` function.
            If True, the returned value is matrix, in which the first column is the right edges of non-empty bins
            and the second one is the histogram values.

        Returns
        -------
        result_tuple : tuple of 2 numpy arrays
            If ``xgboost_style=False``, the values of the histogram of used splitting values for the specified feature
            and the bin edges.
        result_array_like : numpy array or pandas DataFrame (if pandas is installed)
            If ``xgboost_style=True``, the histogram of used splitting values for the specified feature.
        """
4315
        def add(root: Dict[str, Any]) -> None:
4316
4317
            """Recursively add thresholds."""
            if 'split_index' in root:  # non-leaf
4318
                if feature_names is not None and isinstance(feature, str):
4319
4320
4321
4322
                    split_feature = feature_names[root['split_feature']]
                else:
                    split_feature = root['split_feature']
                if split_feature == feature:
4323
                    if isinstance(root['threshold'], str):
4324
4325
4326
                        raise LightGBMError('Cannot compute split value histogram for the categorical feature')
                    else:
                        values.append(root['threshold'])
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
                add(root['left_child'])
                add(root['right_child'])

        model = self.dump_model()
        feature_names = model.get('feature_names')
        tree_infos = model['tree_info']
        values = []
        for tree_info in tree_infos:
            add(tree_info['tree_structure'])

4337
        if bins is None or isinstance(bins, int) and xgboost_style:
4338
4339
4340
4341
4342
4343
4344
            n_unique = len(np.unique(values))
            bins = max(min(n_unique, bins) if bins is not None else n_unique, 1)
        hist, bin_edges = np.histogram(values, bins=bins)
        if xgboost_style:
            ret = np.column_stack((bin_edges[1:], hist))
            ret = ret[ret[:, 1] > 0]
            if PANDAS_INSTALLED:
4345
                return pd_DataFrame(ret, columns=['SplitValue', 'Count'])
4346
4347
4348
4349
4350
            else:
                return ret
        else:
            return hist, bin_edges

4351
4352
4353
4354
    def __inner_eval(
        self,
        data_name: str,
        data_idx: int,
4355
        feval: Optional[Union[_LGBM_CustomEvalFunction, List[_LGBM_CustomEvalFunction]]]
4356
    ) -> List[_LGBM_BoosterEvalMethodResultType]:
4357
        """Evaluate training or validation data."""
wxchan's avatar
wxchan committed
4358
        if data_idx >= self.__num_dataset:
4359
            raise ValueError("Data_idx should be smaller than number of dataset")
wxchan's avatar
wxchan committed
4360
4361
4362
        self.__get_eval_info()
        ret = []
        if self.__num_inner_eval > 0:
4363
            result = np.empty(self.__num_inner_eval, dtype=np.float64)
Guolin Ke's avatar
Guolin Ke committed
4364
            tmp_out_len = ctypes.c_int(0)
wxchan's avatar
wxchan committed
4365
4366
            _safe_call(_LIB.LGBM_BoosterGetEval(
                self.handle,
Guolin Ke's avatar
Guolin Ke committed
4367
                ctypes.c_int(data_idx),
wxchan's avatar
wxchan committed
4368
                ctypes.byref(tmp_out_len),
Guolin Ke's avatar
Guolin Ke committed
4369
                result.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
wxchan's avatar
wxchan committed
4370
            if tmp_out_len.value != self.__num_inner_eval:
4371
                raise ValueError("Wrong length of eval results")
4372
            for i in range(self.__num_inner_eval):
4373
4374
                ret.append((data_name, self.__name_inner_eval[i],
                            result[i], self.__higher_better_inner_eval[i]))
4375
4376
        if callable(feval):
            feval = [feval]
wxchan's avatar
wxchan committed
4377
4378
4379
4380
4381
        if feval is not None:
            if data_idx == 0:
                cur_data = self.train_set
            else:
                cur_data = self.valid_sets[data_idx - 1]
4382
4383
4384
4385
4386
4387
4388
4389
4390
            for eval_function in feval:
                if eval_function is None:
                    continue
                feval_ret = eval_function(self.__inner_predict(data_idx), cur_data)
                if isinstance(feval_ret, list):
                    for eval_name, val, is_higher_better in feval_ret:
                        ret.append((data_name, eval_name, val, is_higher_better))
                else:
                    eval_name, val, is_higher_better = feval_ret
wxchan's avatar
wxchan committed
4391
4392
4393
                    ret.append((data_name, eval_name, val, is_higher_better))
        return ret

4394
    def __inner_predict(self, data_idx: int) -> np.ndarray:
4395
        """Predict for training and validation dataset."""
wxchan's avatar
wxchan committed
4396
        if data_idx >= self.__num_dataset:
4397
            raise ValueError("Data_idx should be smaller than number of dataset")
wxchan's avatar
wxchan committed
4398
4399
4400
4401
4402
        if self.__inner_predict_buffer[data_idx] is None:
            if data_idx == 0:
                n_preds = self.train_set.num_data() * self.__num_class
            else:
                n_preds = self.valid_sets[data_idx - 1].num_data() * self.__num_class
4403
            self.__inner_predict_buffer[data_idx] = np.empty(n_preds, dtype=np.float64)
4404
        # avoid to predict many time in one iteration
wxchan's avatar
wxchan committed
4405
4406
        if not self.__is_predicted_cur_iter[data_idx]:
            tmp_out_len = ctypes.c_int64(0)
Guolin Ke's avatar
Guolin Ke committed
4407
            data_ptr = self.__inner_predict_buffer[data_idx].ctypes.data_as(ctypes.POINTER(ctypes.c_double))
wxchan's avatar
wxchan committed
4408
4409
            _safe_call(_LIB.LGBM_BoosterGetPredict(
                self.handle,
Guolin Ke's avatar
Guolin Ke committed
4410
                ctypes.c_int(data_idx),
wxchan's avatar
wxchan committed
4411
4412
4413
                ctypes.byref(tmp_out_len),
                data_ptr))
            if tmp_out_len.value != len(self.__inner_predict_buffer[data_idx]):
4414
                raise ValueError(f"Wrong length of predict results for data {data_idx}")
wxchan's avatar
wxchan committed
4415
            self.__is_predicted_cur_iter[data_idx] = True
4416
4417
4418
4419
4420
        result = self.__inner_predict_buffer[data_idx]
        if self.__num_class > 1:
            num_data = result.size // self.__num_class
            result = result.reshape(num_data, self.__num_class, order='F')
        return result
wxchan's avatar
wxchan committed
4421

4422
    def __get_eval_info(self) -> None:
4423
        """Get inner evaluation count and names."""
wxchan's avatar
wxchan committed
4424
4425
        if self.__need_reload_eval_info:
            self.__need_reload_eval_info = False
Guolin Ke's avatar
Guolin Ke committed
4426
            out_num_eval = ctypes.c_int(0)
4427
            # Get num of inner evals
wxchan's avatar
wxchan committed
4428
4429
4430
4431
4432
            _safe_call(_LIB.LGBM_BoosterGetEvalCounts(
                self.handle,
                ctypes.byref(out_num_eval)))
            self.__num_inner_eval = out_num_eval.value
            if self.__num_inner_eval > 0:
4433
                # Get name of eval metrics
Guolin Ke's avatar
Guolin Ke committed
4434
                tmp_out_len = ctypes.c_int(0)
4435
4436
4437
                reserved_string_buffer_size = 255
                required_string_buffer_size = ctypes.c_size_t(0)
                string_buffers = [
4438
                    ctypes.create_string_buffer(reserved_string_buffer_size) for _ in range(self.__num_inner_eval)
4439
                ]
wxchan's avatar
wxchan committed
4440
                ptr_string_buffers = (ctypes.c_char_p * self.__num_inner_eval)(*map(ctypes.addressof, string_buffers))
wxchan's avatar
wxchan committed
4441
4442
                _safe_call(_LIB.LGBM_BoosterGetEvalNames(
                    self.handle,
4443
                    ctypes.c_int(self.__num_inner_eval),
wxchan's avatar
wxchan committed
4444
                    ctypes.byref(tmp_out_len),
4445
                    ctypes.c_size_t(reserved_string_buffer_size),
4446
                    ctypes.byref(required_string_buffer_size),
wxchan's avatar
wxchan committed
4447
4448
                    ptr_string_buffers))
                if self.__num_inner_eval != tmp_out_len.value:
4449
                    raise ValueError("Length of eval names doesn't equal with num_evals")
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
                actual_string_buffer_size = required_string_buffer_size.value
                # if buffer length is not long enough, reallocate buffers
                if reserved_string_buffer_size < actual_string_buffer_size:
                    string_buffers = [
                        ctypes.create_string_buffer(actual_string_buffer_size) for _ in range(self.__num_inner_eval)
                    ]
                    ptr_string_buffers = (ctypes.c_char_p * self.__num_inner_eval)(*map(ctypes.addressof, string_buffers))
                    _safe_call(_LIB.LGBM_BoosterGetEvalNames(
                        self.handle,
                        ctypes.c_int(self.__num_inner_eval),
                        ctypes.byref(tmp_out_len),
                        ctypes.c_size_t(actual_string_buffer_size),
                        ctypes.byref(required_string_buffer_size),
                        ptr_string_buffers))
                self.__name_inner_eval = [
                    string_buffers[i].value.decode('utf-8') for i in range(self.__num_inner_eval)
                ]
                self.__higher_better_inner_eval = [
                    name.startswith(('auc', 'ndcg@', 'map@', 'average_precision')) for name in self.__name_inner_eval
                ]