basic.py 163 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
10
from functools import wraps
from logging import Logger
11
12
13
from os import SEEK_END
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, is_dtype_sparse, 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
32
33
34
ZERO_THRESHOLD = 1e-35


def _get_sample_count(total_nrow: int, params: str):
    sample_cnt = ctypes.c_int(0)
    _safe_call(_LIB.LGBM_GetSampleCount(
        ctypes.c_int32(total_nrow),
        c_str(params),
        ctypes.byref(sample_cnt),
    ))
    return sample_cnt.value

wxchan's avatar
wxchan committed
35

36
class _DummyLogger:
37
    def info(self, msg: str) -> None:
38
39
        print(msg)

40
    def warning(self, msg: str) -> None:
41
42
43
        warnings.warn(msg, stacklevel=3)


44
_LOGGER: Union[_DummyLogger, Logger] = _DummyLogger()
45
46


47
def register_logger(logger: Logger) -> None:
48
49
50
51
52
53
54
55
56
57
58
59
60
    """Register custom logger.

    Parameters
    ----------
    logger : logging.Logger
        Custom logger.
    """
    if not isinstance(logger, Logger):
        raise TypeError("Logger should inherit logging.Logger class")
    global _LOGGER
    _LOGGER = logger


61
def _normalize_native_string(func: Callable[[str], None]) -> Callable[[str], None]:
62
    """Join log messages from native library which come by chunks."""
63
    msg_normalized: List[str] = []
64
65

    @wraps(func)
66
    def wrapper(msg: str) -> None:
67
68
69
70
71
72
73
74
75
76
77
        nonlocal msg_normalized
        if msg.strip() == '':
            msg = ''.join(msg_normalized)
            msg_normalized = []
            return func(msg)
        else:
            msg_normalized.append(msg)

    return wrapper


78
def _log_info(msg: str) -> None:
79
80
81
    _LOGGER.info(msg)


82
def _log_warning(msg: str) -> None:
83
84
85
86
    _LOGGER.warning(msg)


@_normalize_native_string
87
def _log_native(msg: str) -> None:
88
89
90
    _LOGGER.info(msg)


91
def _log_callback(msg: bytes) -> None:
92
    """Redirect logs from native library into Python."""
93
    _log_native(str(msg.decode('utf-8')))
94
95


wxchan's avatar
wxchan committed
96
def _load_lib():
97
    """Load LightGBM library."""
wxchan's avatar
wxchan committed
98
99
    lib_path = find_lib_path()
    if len(lib_path) == 0:
100
        return None
wxchan's avatar
wxchan committed
101
102
    lib = ctypes.cdll.LoadLibrary(lib_path[0])
    lib.LGBM_GetLastError.restype = ctypes.c_char_p
103
104
105
    callback = ctypes.CFUNCTYPE(None, ctypes.c_char_p)
    lib.callback = callback(_log_callback)
    if lib.LGBM_RegisterLogCallback(lib.callback) != 0:
106
        raise LightGBMError(lib.LGBM_GetLastError().decode('utf-8'))
wxchan's avatar
wxchan committed
107
108
    return lib

wxchan's avatar
wxchan committed
109

wxchan's avatar
wxchan committed
110
111
_LIB = _load_lib()

wxchan's avatar
wxchan committed
112

113
114
115
NUMERIC_TYPES = (int, float, bool)


116
def _safe_call(ret: int) -> None:
117
118
    """Check the return value from C API call.

wxchan's avatar
wxchan committed
119
120
121
    Parameters
    ----------
    ret : int
122
        The return value from C API calls.
wxchan's avatar
wxchan committed
123
124
    """
    if ret != 0:
125
        raise LightGBMError(_LIB.LGBM_GetLastError().decode('utf-8'))
wxchan's avatar
wxchan committed
126

wxchan's avatar
wxchan committed
127

wxchan's avatar
wxchan committed
128
def is_numeric(obj):
129
    """Check whether object is a number or not, include numpy number, etc."""
wxchan's avatar
wxchan committed
130
131
132
    try:
        float(obj)
        return True
wxchan's avatar
wxchan committed
133
134
135
    except (TypeError, ValueError):
        # TypeError: obj is not a string or a number
        # ValueError: invalid literal
wxchan's avatar
wxchan committed
136
137
        return False

wxchan's avatar
wxchan committed
138

wxchan's avatar
wxchan committed
139
def is_numpy_1d_array(data):
140
    """Check whether data is a numpy 1-D array."""
141
    return isinstance(data, np.ndarray) and len(data.shape) == 1
wxchan's avatar
wxchan committed
142

wxchan's avatar
wxchan committed
143

144
145
146
147
148
149
150
151
def is_numpy_column_array(data):
    """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


152
153
def cast_numpy_array_to_dtype(array, dtype):
    """Cast numpy array to given dtype."""
154
155
156
157
158
    if array.dtype == dtype:
        return array
    return array.astype(dtype=dtype, copy=False)


wxchan's avatar
wxchan committed
159
def is_1d_list(data):
160
161
    """Check whether data is a 1-D list."""
    return isinstance(data, list) and (not data or is_numeric(data[0]))
wxchan's avatar
wxchan committed
162

wxchan's avatar
wxchan committed
163

164
165
166
167
168
169
170
171
172
173
def _is_1d_collection(data: Any) -> bool:
    """Check whether data is a 1-D collection."""
    return (
        is_numpy_1d_array(data)
        or is_numpy_column_array(data)
        or is_1d_list(data)
        or isinstance(data, pd_Series)
    )


174
def list_to_1d_numpy(data, dtype=np.float32, name='list'):
175
    """Convert data to numpy 1-D array."""
wxchan's avatar
wxchan committed
176
    if is_numpy_1d_array(data):
177
        return cast_numpy_array_to_dtype(data, dtype)
178
179
180
    elif is_numpy_column_array(data):
        _log_warning('Converting column-vector to 1d array')
        array = data.ravel()
181
        return cast_numpy_array_to_dtype(array, dtype)
wxchan's avatar
wxchan committed
182
183
    elif is_1d_list(data):
        return np.array(data, dtype=dtype, copy=False)
184
    elif isinstance(data, pd_Series):
185
186
        if _get_bad_pandas_dtypes([data.dtypes]):
            raise ValueError('Series.dtypes must be int, float or bool')
187
        return np.array(data, dtype=dtype, copy=False)  # SparseArray should be supported as well
wxchan's avatar
wxchan committed
188
    else:
189
190
        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
191

wxchan's avatar
wxchan committed
192

193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
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."""
    return isinstance(data, list) and len(data) > 0 and is_1d_list(data[0])


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):
        return cast_numpy_array_to_dtype(data, dtype)
    if _is_2d_list(data):
        return np.array(data, dtype=dtype)
    if isinstance(data, pd_DataFrame):
        if _get_bad_pandas_dtypes(data.dtypes):
            raise ValueError('DataFrame.dtypes must be int, float or bool')
        return cast_numpy_array_to_dtype(data.values, dtype)
    raise TypeError(f"Wrong type({type(data).__name__}) for {name}.\n"
                    "It should be list of lists, numpy 2-D array or pandas DataFrame")


wxchan's avatar
wxchan committed
226
def cfloat32_array_to_numpy(cptr, length):
227
    """Convert a ctypes float pointer array to a numpy array."""
wxchan's avatar
wxchan committed
228
    if isinstance(cptr, ctypes.POINTER(ctypes.c_float)):
229
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
wxchan's avatar
wxchan committed
230
    else:
231
        raise RuntimeError('Expected float pointer')
wxchan's avatar
wxchan committed
232

Guolin Ke's avatar
Guolin Ke committed
233

Guolin Ke's avatar
Guolin Ke committed
234
def cfloat64_array_to_numpy(cptr, length):
235
    """Convert a ctypes double pointer array to a numpy array."""
Guolin Ke's avatar
Guolin Ke committed
236
    if isinstance(cptr, ctypes.POINTER(ctypes.c_double)):
237
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
Guolin Ke's avatar
Guolin Ke committed
238
239
240
    else:
        raise RuntimeError('Expected double pointer')

wxchan's avatar
wxchan committed
241

wxchan's avatar
wxchan committed
242
def cint32_array_to_numpy(cptr, length):
243
    """Convert a ctypes int pointer array to a numpy array."""
wxchan's avatar
wxchan committed
244
    if isinstance(cptr, ctypes.POINTER(ctypes.c_int32)):
245
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
wxchan's avatar
wxchan committed
246
    else:
247
248
249
250
251
252
        raise RuntimeError('Expected int32 pointer')


def cint64_array_to_numpy(cptr, length):
    """Convert a ctypes int pointer array to a numpy array."""
    if isinstance(cptr, ctypes.POINTER(ctypes.c_int64)):
253
        return np.ctypeslib.as_array(cptr, shape=(length,)).copy()
254
255
    else:
        raise RuntimeError('Expected int64 pointer')
wxchan's avatar
wxchan committed
256

wxchan's avatar
wxchan committed
257

wxchan's avatar
wxchan committed
258
def c_str(string):
259
    """Convert a Python string to C string."""
wxchan's avatar
wxchan committed
260
261
    return ctypes.c_char_p(string.encode('utf-8'))

wxchan's avatar
wxchan committed
262

wxchan's avatar
wxchan committed
263
def c_array(ctype, values):
264
    """Convert a Python array to C array."""
wxchan's avatar
wxchan committed
265
266
    return (ctype * len(values))(*values)

wxchan's avatar
wxchan committed
267

268
269
270
271
272
273
274
275
276
277
def json_default_with_numpy(obj):
    """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


wxchan's avatar
wxchan committed
278
def param_dict_to_str(data):
279
    """Convert Python dictionary to string, which is passed to C API."""
280
    if data is None or not data:
wxchan's avatar
wxchan committed
281
282
283
        return ""
    pairs = []
    for key, val in data.items():
284
        if isinstance(val, (list, tuple, set)) or is_numpy_1d_array(val):
285
286
            def to_string(x):
                if isinstance(x, list):
287
                    return f"[{','.join(map(str, x))}]"
288
289
                else:
                    return str(x)
290
            pairs.append(f"{key}={','.join(map(to_string, val))}")
291
        elif isinstance(val, (str, Path, NUMERIC_TYPES)) or is_numeric(val):
292
            pairs.append(f"{key}={val}")
293
        elif val is not None:
294
            raise TypeError(f'Unknown type of parameter:{key}, got:{type(val).__name__}')
wxchan's avatar
wxchan committed
295
    return ' '.join(pairs)
296

wxchan's avatar
wxchan committed
297

298
class _TempFile:
299
300
    """Proxy class to workaround errors on Windows."""

301
302
303
    def __enter__(self):
        with NamedTemporaryFile(prefix="lightgbm_tmp_", delete=True) as f:
            self.name = f.name
304
            self.path = Path(self.name)
305
        return self
wxchan's avatar
wxchan committed
306

307
    def __exit__(self, exc_type, exc_val, exc_tb):
308
309
        if self.path.is_file():
            self.path.unlink()
310

wxchan's avatar
wxchan committed
311

312
class LightGBMError(Exception):
313
314
    """Error thrown by LightGBM."""

315
316
317
    pass


318
319
320
321
322
323
324
325
# DeprecationWarning is not shown by default, so let's create our own with higher level
class LGBMDeprecationWarning(UserWarning):
    """Custom deprecation warning."""

    pass


class _ConfigAliases:
326
327
328
    aliases = {"bin_construct_sample_cnt": {"bin_construct_sample_cnt",
                                            "subsample_for_bin"},
               "boosting": {"boosting",
329
330
331
332
333
                            "boosting_type",
                            "boost"},
               "categorical_feature": {"categorical_feature",
                                       "cat_feature",
                                       "categorical_column",
334
335
                                       "cat_column",
                                       "categorical_features"},
336
337
               "data_random_seed": {"data_random_seed",
                                    "data_seed"},
338
339
340
341
               "early_stopping_round": {"early_stopping_round",
                                        "early_stopping_rounds",
                                        "early_stopping",
                                        "n_iter_no_change"},
342
343
344
               "enable_bundle": {"enable_bundle",
                                 "is_enable_bundle",
                                 "bundle"},
345
346
347
348
349
               "eval_at": {"eval_at",
                           "ndcg_eval_at",
                           "ndcg_at",
                           "map_eval_at",
                           "map_at"},
350
351
352
353
354
355
               "group_column": {"group_column",
                                "group",
                                "group_id",
                                "query_column",
                                "query",
                                "query_id"},
356
357
               "header": {"header",
                          "has_header"},
358
359
360
361
362
363
364
365
366
               "ignore_column": {"ignore_column",
                                 "ignore_feature",
                                 "blacklist"},
               "is_enable_sparse": {"is_enable_sparse",
                                    "is_sparse",
                                    "enable_sparse",
                                    "sparse"},
               "label_column": {"label_column",
                                "label"},
Nikita Titov's avatar
Nikita Titov committed
367
368
               "linear_tree": {"linear_tree",
                               "linear_trees"},
369
370
371
               "local_listen_port": {"local_listen_port",
                                     "local_port",
                                     "port"},
372
373
374
               "machines": {"machines",
                            "workers",
                            "nodes"},
375
376
               "max_bin": {"max_bin",
                           "max_bins"},
377
378
379
380
381
382
383
384
385
386
387
388
               "metric": {"metric",
                          "metrics",
                          "metric_types"},
               "num_class": {"num_class",
                             "num_classes"},
               "num_iterations": {"num_iterations",
                                  "num_iteration",
                                  "n_iter",
                                  "num_tree",
                                  "num_trees",
                                  "num_round",
                                  "num_rounds",
389
                                  "nrounds",
390
                                  "num_boost_round",
391
392
                                  "n_estimators",
                                  "max_iter"},
393
394
               "num_machines": {"num_machines",
                                "num_machine"},
395
396
397
398
399
               "num_threads": {"num_threads",
                               "num_thread",
                               "nthread",
                               "nthreads",
                               "n_jobs"},
400
401
402
               "objective": {"objective",
                             "objective_type",
                             "app",
403
404
                             "application",
                             "loss"},
405
406
               "pre_partition": {"pre_partition",
                                 "is_pre_partition"},
407
408
409
410
               "tree_learner": {"tree_learner",
                                "tree",
                                "tree_type",
                                "tree_learner_type"},
411
412
413
               "two_round": {"two_round",
                             "two_round_loading",
                             "use_two_round_loading"},
414
               "verbosity": {"verbosity",
415
416
417
                             "verbose"},
               "weight_column": {"weight_column",
                                 "weight"}}
418
419
420
421
422

    @classmethod
    def get(cls, *args):
        ret = set()
        for i in args:
423
            ret |= cls.aliases.get(i, {i})
424
425
426
        return ret


427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
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)

    # find a value, and remove other aliases with .pop()
    # prefer the value of 'main_param_name' if it exists, otherwise search the aliases
    found_value = None
    if main_param_name in params.keys():
        found_value = params[main_param_name]

    for param in _ConfigAliases.get(main_param_name):
        val = params.pop(param, None)
        if found_value is None and val is not None:
            found_value = val

    if found_value is not None:
        params[main_param_name] = found_value
    else:
        params[main_param_name] = default_value

    return params


467
468
MAX_INT32 = (1 << 31) - 1

469
"""Macro definition of data type in C API of LightGBM"""
wxchan's avatar
wxchan committed
470
471
472
473
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
474

475
"""Matrix is row major in Python"""
wxchan's avatar
wxchan committed
476
477
C_API_IS_ROW_MAJOR = 1

478
"""Macro definition of prediction type in C API of LightGBM"""
wxchan's avatar
wxchan committed
479
480
481
C_API_PREDICT_NORMAL = 0
C_API_PREDICT_RAW_SCORE = 1
C_API_PREDICT_LEAF_INDEX = 2
482
C_API_PREDICT_CONTRIB = 3
wxchan's avatar
wxchan committed
483

484
485
486
487
"""Macro definition of sparse matrix type"""
C_API_MATRIX_TYPE_CSR = 0
C_API_MATRIX_TYPE_CSC = 1

488
489
490
491
"""Macro definition of feature importance type"""
C_API_FEATURE_IMPORTANCE_SPLIT = 0
C_API_FEATURE_IMPORTANCE_GAIN = 1

492
"""Data type of data field"""
wxchan's avatar
wxchan committed
493
494
FIELD_TYPE_MAPPER = {"label": C_API_DTYPE_FLOAT32,
                     "weight": C_API_DTYPE_FLOAT32,
Guolin Ke's avatar
Guolin Ke committed
495
                     "init_score": C_API_DTYPE_FLOAT64,
496
                     "group": C_API_DTYPE_INT32}
wxchan's avatar
wxchan committed
497

498
499
500
501
"""String name to int feature importance type mapper"""
FEATURE_IMPORTANCE_TYPE_MAPPER = {"split": C_API_FEATURE_IMPORTANCE_SPLIT,
                                  "gain": C_API_FEATURE_IMPORTANCE_GAIN}

wxchan's avatar
wxchan committed
502

503
def convert_from_sliced_object(data):
504
    """Fix the memory of multi-dimensional sliced object."""
505
    if isinstance(data, np.ndarray) and isinstance(data.base, np.ndarray):
506
        if not data.flags.c_contiguous:
507
508
            _log_warning("Usage of np.ndarray subset (sliced data) is not recommended "
                         "due to it will double the peak memory cost in LightGBM.")
509
510
511
512
            return np.copy(data)
    return data


wxchan's avatar
wxchan committed
513
def c_float_array(data):
514
    """Get pointer of float numpy array / list."""
wxchan's avatar
wxchan committed
515
516
517
    if is_1d_list(data):
        data = np.array(data, copy=False)
    if is_numpy_1d_array(data):
518
519
        data = convert_from_sliced_object(data)
        assert data.flags.c_contiguous
wxchan's avatar
wxchan committed
520
521
522
523
524
525
526
        if data.dtype == np.float32:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_float))
            type_data = C_API_DTYPE_FLOAT32
        elif data.dtype == np.float64:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_double))
            type_data = C_API_DTYPE_FLOAT64
        else:
527
            raise TypeError(f"Expected np.float32 or np.float64, met type({data.dtype})")
wxchan's avatar
wxchan committed
528
    else:
529
        raise TypeError(f"Unknown type({type(data).__name__})")
530
    return (ptr_data, type_data, data)  # return `data` to avoid the temporary copy is freed
wxchan's avatar
wxchan committed
531

wxchan's avatar
wxchan committed
532

wxchan's avatar
wxchan committed
533
def c_int_array(data):
534
    """Get pointer of int numpy array / list."""
wxchan's avatar
wxchan committed
535
536
537
    if is_1d_list(data):
        data = np.array(data, copy=False)
    if is_numpy_1d_array(data):
538
539
        data = convert_from_sliced_object(data)
        assert data.flags.c_contiguous
wxchan's avatar
wxchan committed
540
541
542
543
544
545
546
        if data.dtype == np.int32:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_int32))
            type_data = C_API_DTYPE_INT32
        elif data.dtype == np.int64:
            ptr_data = data.ctypes.data_as(ctypes.POINTER(ctypes.c_int64))
            type_data = C_API_DTYPE_INT64
        else:
547
            raise TypeError(f"Expected np.int32 or np.int64, met type({data.dtype})")
wxchan's avatar
wxchan committed
548
    else:
549
        raise TypeError(f"Unknown type({type(data).__name__})")
550
    return (ptr_data, type_data, data)  # return `data` to avoid the temporary copy is freed
wxchan's avatar
wxchan committed
551

wxchan's avatar
wxchan committed
552

553
554
555
556
557
558
559
560
561
562
563
def _get_bad_pandas_dtypes(dtypes):
    pandas_dtype_mapper = {'int8': 'int', 'int16': 'int', 'int32': 'int',
                           'int64': 'int', 'uint8': 'int', 'uint16': 'int',
                           'uint32': 'int', 'uint64': 'int', 'bool': 'int',
                           'float16': 'float', 'float32': 'float', 'float64': 'float'}
    bad_indices = [i for i, dtype in enumerate(dtypes) if (dtype.name not in pandas_dtype_mapper
                                                           and (not is_dtype_sparse(dtype)
                                                                or dtype.subtype.name not in pandas_dtype_mapper))]
    return bad_indices


564
def _data_from_pandas(data, feature_name, categorical_feature, pandas_categorical):
565
    if isinstance(data, pd_DataFrame):
566
567
        if len(data.shape) != 2 or data.shape[0] < 1:
            raise ValueError('Input data must be 2 dimensional and non empty.')
568
569
        if feature_name == 'auto' or feature_name is None:
            data = data.rename(columns=str)
570
571
        cat_cols = list(data.select_dtypes(include=['category']).columns)
        cat_cols_not_ordered = [col for col in cat_cols if not data[col].cat.ordered]
572
573
574
575
576
        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.')
577
            for col, category in zip(cat_cols, pandas_categorical):
578
579
                if list(data[col].cat.categories) != list(category):
                    data[col] = data[col].cat.set_categories(category)
580
        if len(cat_cols):  # cat_cols is list
581
            data = data.copy()  # not alter origin DataFrame
582
            data[cat_cols] = data[cat_cols].apply(lambda x: x.cat.codes).replace({-1: np.nan})
583
584
585
        if categorical_feature is not None:
            if feature_name is None:
                feature_name = list(data.columns)
586
            if categorical_feature == 'auto':  # use cat cols from DataFrame
587
                categorical_feature = cat_cols_not_ordered
588
589
            else:  # use cat cols specified by user
                categorical_feature = list(categorical_feature)
590
591
        if feature_name == 'auto':
            feature_name = list(data.columns)
592
593
        bad_indices = _get_bad_pandas_dtypes(data.dtypes)
        if bad_indices:
594
            bad_index_cols_str = ', '.join(data.columns[bad_indices])
595
            raise ValueError("DataFrame.dtypes for data must be int, float or bool.\n"
596
                             "Did not expect the data types in the following fields: "
597
                             f"{bad_index_cols_str}")
598
599
600
        data = data.values
        if data.dtype != np.float32 and data.dtype != np.float64:
            data = data.astype(np.float32)
601
602
603
604
605
606
    else:
        if feature_name == 'auto':
            feature_name = None
        if categorical_feature == 'auto':
            categorical_feature = None
    return data, feature_name, categorical_feature, pandas_categorical
607
608
609


def _label_from_pandas(label):
610
    if isinstance(label, pd_DataFrame):
611
612
        if len(label.columns) > 1:
            raise ValueError('DataFrame for label cannot have multiple columns')
613
        if _get_bad_pandas_dtypes(label.dtypes):
614
            raise ValueError('DataFrame.dtypes for label must be int, float or bool')
615
        label = np.ravel(label.values.astype(np.float32, copy=False))
616
617
618
    return label


619
def _dump_pandas_categorical(pandas_categorical, file_name=None):
620
621
    categorical_json = json.dumps(pandas_categorical, default=json_default_with_numpy)
    pandas_str = f'\npandas_categorical:{categorical_json}\n'
622
623
624
625
626
627
628
    if file_name is not None:
        with open(file_name, 'a') as f:
            f.write(pandas_str)
    return pandas_str


def _load_pandas_categorical(file_name=None, model_str=None):
629
630
    pandas_key = 'pandas_categorical:'
    offset = -len(pandas_key)
631
    if file_name is not None:
632
        max_offset = -getsize(file_name)
633
634
635
636
        with open(file_name, 'rb') as f:
            while True:
                if offset < max_offset:
                    offset = max_offset
637
                f.seek(offset, SEEK_END)
638
639
640
641
                lines = f.readlines()
                if len(lines) >= 2:
                    break
                offset *= 2
642
        last_line = lines[-1].decode('utf-8').strip()
643
        if not last_line.startswith(pandas_key):
644
            last_line = lines[-2].decode('utf-8').strip()
645
    elif model_str is not None:
646
647
648
649
650
651
        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
652
653


654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
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**.

674
675
    .. versionadded:: 3.3.0

676
677
678
679
680
681
682
683
684
    Attributes
    ----------
    batch_size : int
        Default size of a batch.
    """

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

    @abc.abstractmethod
685
    def __getitem__(self, idx: Union[int, slice, List[int]]) -> np.ndarray:
686
687
688
689
690
691
692
        """Return data for given row index.

        A basic implementation should look like this:

        .. code-block:: python

            if isinstance(idx, numbers.Integral):
693
                return self._get_one_line(idx)
694
            elif isinstance(idx, slice):
695
696
                return np.stack([self._get_one_line(i) for i in range(idx.start, idx.stop)])
            elif isinstance(idx, list):
697
                # Only required if using ``Dataset.subset()``.
698
                return np.array([self._get_one_line(i) for i in idx])
699
            else:
700
                raise TypeError(f"Sequence index must be integer, slice or list, got {type(idx).__name__}")
701
702
703

        Parameters
        ----------
704
        idx : int, slice[int], list[int]
705
706
707
708
709
            Item index.

        Returns
        -------
        result : numpy 1-D array, numpy 2-D array
710
            1-D array if idx is int, 2-D array if idx is slice or list.
711
712
713
714
715
716
717
718
719
        """
        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__()")


720
class _InnerPredictor:
721
722
723
724
725
    """_InnerPredictor of LightGBM.

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

Nikita Titov's avatar
Nikita Titov committed
726
727
728
    .. note::

        Can be converted from Booster, but cannot be converted to Booster.
Guolin Ke's avatar
Guolin Ke committed
729
    """
730

731
    def __init__(self, model_file=None, booster_handle=None, pred_parameter=None):
732
        """Initialize the _InnerPredictor.
wxchan's avatar
wxchan committed
733
734
735

        Parameters
        ----------
736
        model_file : str, pathlib.Path or None, optional (default=None)
wxchan's avatar
wxchan committed
737
            Path to the model file.
738
739
740
        booster_handle : object or None, optional (default=None)
            Handle of Booster.
        pred_parameter: dict or None, optional (default=None)
741
            Other parameters for the prediction.
wxchan's avatar
wxchan committed
742
743
744
745
746
        """
        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
747
            out_num_iterations = ctypes.c_int(0)
wxchan's avatar
wxchan committed
748
            _safe_call(_LIB.LGBM_BoosterCreateFromModelfile(
749
                c_str(str(model_file)),
wxchan's avatar
wxchan committed
750
751
                ctypes.byref(out_num_iterations),
                ctypes.byref(self.handle)))
Guolin Ke's avatar
Guolin Ke committed
752
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
753
754
755
756
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.num_class = out_num_class.value
757
            self.num_total_iteration = out_num_iterations.value
758
            self.pandas_categorical = _load_pandas_categorical(file_name=model_file)
wxchan's avatar
wxchan committed
759
        elif booster_handle is not None:
Guolin Ke's avatar
Guolin Ke committed
760
            self.__is_manage_handle = False
wxchan's avatar
wxchan committed
761
            self.handle = booster_handle
Guolin Ke's avatar
Guolin Ke committed
762
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
763
764
765
766
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.num_class = out_num_class.value
767
            self.num_total_iteration = self.current_iteration()
768
            self.pandas_categorical = None
wxchan's avatar
wxchan committed
769
        else:
770
            raise TypeError('Need model_file or booster_handle to create a predictor')
wxchan's avatar
wxchan committed
771

772
773
        pred_parameter = {} if pred_parameter is None else pred_parameter
        self.pred_parameter = param_dict_to_str(pred_parameter)
cbecker's avatar
cbecker committed
774

wxchan's avatar
wxchan committed
775
    def __del__(self):
776
777
778
779
780
        try:
            if self.__is_manage_handle:
                _safe_call(_LIB.LGBM_BoosterFree(self.handle))
        except AttributeError:
            pass
wxchan's avatar
wxchan committed
781

782
783
784
785
786
    def __getstate__(self):
        this = self.__dict__.copy()
        this.pop('handle', None)
        return this

787
    def predict(self, data, start_iteration=0, num_iteration=-1,
788
                raw_score=False, pred_leaf=False, pred_contrib=False, data_has_header=False,
wxchan's avatar
wxchan committed
789
                is_reshape=True):
790
        """Predict logic.
wxchan's avatar
wxchan committed
791
792
793

        Parameters
        ----------
794
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame or scipy.sparse
795
            Data source for prediction.
796
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM).
797
798
        start_iteration : int, optional (default=0)
            Start index of the iteration to predict.
799
800
801
802
803
804
805
806
807
808
809
810
811
        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.
        is_reshape : bool, optional (default=True)
            Whether to reshape to (nrow, ncol).
wxchan's avatar
wxchan committed
812
813
814

        Returns
        -------
815
        result : numpy array, scipy.sparse or list of scipy.sparse
816
            Prediction result.
817
            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
818
        """
wxchan's avatar
wxchan committed
819
        if isinstance(data, Dataset):
820
            raise TypeError("Cannot use Dataset instance for prediction, please use raw data instead")
821
        data = _data_from_pandas(data, None, None, self.pandas_categorical)[0]
wxchan's avatar
wxchan committed
822
823
824
825
826
        predict_type = C_API_PREDICT_NORMAL
        if raw_score:
            predict_type = C_API_PREDICT_RAW_SCORE
        if pred_leaf:
            predict_type = C_API_PREDICT_LEAF_INDEX
827
828
        if pred_contrib:
            predict_type = C_API_PREDICT_CONTRIB
wxchan's avatar
wxchan committed
829
        int_data_has_header = 1 if data_has_header else 0
cbecker's avatar
cbecker committed
830

831
        if isinstance(data, (str, Path)):
832
            with _TempFile() as f:
wxchan's avatar
wxchan committed
833
834
                _safe_call(_LIB.LGBM_BoosterPredictForFile(
                    self.handle,
835
                    c_str(str(data)),
Guolin Ke's avatar
Guolin Ke committed
836
837
                    ctypes.c_int(int_data_has_header),
                    ctypes.c_int(predict_type),
838
                    ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
839
                    ctypes.c_int(num_iteration),
840
                    c_str(self.pred_parameter),
wxchan's avatar
wxchan committed
841
                    c_str(f.name)))
842
843
                preds = np.loadtxt(f.name, dtype=np.float64)
                nrow = preds.shape[0]
wxchan's avatar
wxchan committed
844
        elif isinstance(data, scipy.sparse.csr_matrix):
845
            preds, nrow = self.__pred_for_csr(data, start_iteration, num_iteration, predict_type)
Guolin Ke's avatar
Guolin Ke committed
846
        elif isinstance(data, scipy.sparse.csc_matrix):
847
            preds, nrow = self.__pred_for_csc(data, start_iteration, num_iteration, predict_type)
wxchan's avatar
wxchan committed
848
        elif isinstance(data, np.ndarray):
849
            preds, nrow = self.__pred_for_np2d(data, start_iteration, num_iteration, predict_type)
850
851
852
        elif isinstance(data, list):
            try:
                data = np.array(data)
853
            except BaseException:
854
                raise ValueError('Cannot convert data list to numpy array.')
855
            preds, nrow = self.__pred_for_np2d(data, start_iteration, num_iteration, predict_type)
856
        elif isinstance(data, dt_DataTable):
857
            preds, nrow = self.__pred_for_np2d(data.to_numpy(), start_iteration, num_iteration, predict_type)
wxchan's avatar
wxchan committed
858
859
        else:
            try:
860
                _log_warning('Converting data to scipy sparse matrix.')
wxchan's avatar
wxchan committed
861
                csr = scipy.sparse.csr_matrix(data)
862
            except BaseException:
863
                raise TypeError(f'Cannot predict data for type {type(data).__name__}')
864
            preds, nrow = self.__pred_for_csr(csr, start_iteration, num_iteration, predict_type)
wxchan's avatar
wxchan committed
865
866
        if pred_leaf:
            preds = preds.astype(np.int32)
867
868
        is_sparse = scipy.sparse.issparse(preds) or isinstance(preds, list)
        if is_reshape and not is_sparse and preds.size != nrow:
wxchan's avatar
wxchan committed
869
            if preds.size % nrow == 0:
870
                preds = preds.reshape(nrow, -1)
wxchan's avatar
wxchan committed
871
            else:
872
                raise ValueError(f'Length of predict result ({preds.size}) cannot be divide nrow ({nrow})')
wxchan's avatar
wxchan committed
873
874
        return preds

875
    def __get_num_preds(self, start_iteration, num_iteration, nrow, predict_type):
876
        """Get size of prediction result."""
877
        if nrow > MAX_INT32:
878
            raise LightGBMError('LightGBM cannot perform prediction for data '
879
                                f'with number of rows greater than MAX_INT32 ({MAX_INT32}).\n'
880
                                'You can split your data into chunks '
881
                                'and then concatenate predictions for them')
Guolin Ke's avatar
Guolin Ke committed
882
883
884
        n_preds = ctypes.c_int64(0)
        _safe_call(_LIB.LGBM_BoosterCalcNumPredict(
            self.handle,
Guolin Ke's avatar
Guolin Ke committed
885
886
            ctypes.c_int(nrow),
            ctypes.c_int(predict_type),
887
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
888
            ctypes.c_int(num_iteration),
Guolin Ke's avatar
Guolin Ke committed
889
890
            ctypes.byref(n_preds)))
        return n_preds.value
wxchan's avatar
wxchan committed
891

892
    def __pred_for_np2d(self, mat, start_iteration, num_iteration, predict_type):
893
        """Predict for a 2-D numpy matrix."""
wxchan's avatar
wxchan committed
894
        if len(mat.shape) != 2:
895
            raise ValueError('Input numpy.ndarray or list must be 2 dimensional')
wxchan's avatar
wxchan committed
896

897
        def inner_predict(mat, start_iteration, num_iteration, predict_type, preds=None):
898
899
            if mat.dtype == np.float32 or mat.dtype == np.float64:
                data = np.array(mat.reshape(mat.size), dtype=mat.dtype, copy=False)
900
            else:  # change non-float data to float data, need to copy
901
902
                data = np.array(mat.reshape(mat.size), dtype=np.float32)
            ptr_data, type_ptr_data, _ = c_float_array(data)
903
            n_preds = self.__get_num_preds(start_iteration, num_iteration, mat.shape[0], predict_type)
904
            if preds is None:
905
                preds = np.empty(n_preds, dtype=np.float64)
906
907
908
909
910
911
912
            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),
913
914
                ctypes.c_int32(mat.shape[0]),
                ctypes.c_int32(mat.shape[1]),
915
916
                ctypes.c_int(C_API_IS_ROW_MAJOR),
                ctypes.c_int(predict_type),
917
                ctypes.c_int(start_iteration),
918
919
920
921
922
923
924
925
926
927
928
929
                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]

        nrow = mat.shape[0]
        if nrow > MAX_INT32:
            sections = np.arange(start=MAX_INT32, stop=nrow, step=MAX_INT32)
            # __get_num_preds() cannot work with nrow > MAX_INT32, so calculate overall number of predictions piecemeal
930
            n_preds = [self.__get_num_preds(start_iteration, num_iteration, i, predict_type) for i in np.diff([0] + list(sections) + [nrow])]
931
            n_preds_sections = np.array([0] + n_preds, dtype=np.intp).cumsum()
932
            preds = np.empty(sum(n_preds), dtype=np.float64)
933
934
            for chunk, (start_idx_pred, end_idx_pred) in zip(np.array_split(mat, sections),
                                                             zip(n_preds_sections, n_preds_sections[1:])):
935
                # avoid memory consumption by arrays concatenation operations
936
                inner_predict(chunk, start_iteration, num_iteration, predict_type, preds[start_idx_pred:end_idx_pred])
937
            return preds, nrow
wxchan's avatar
wxchan committed
938
        else:
939
            return inner_predict(mat, start_iteration, num_iteration, predict_type)
wxchan's avatar
wxchan committed
940

941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
    def __create_sparse_native(self, cs, out_shape, out_ptr_indptr, out_ptr_indices, out_ptr_data,
                               indptr_type, data_type, is_csr=True):
        # create numpy array from output arrays
        data_indices_len = out_shape[0]
        indptr_len = out_shape[1]
        if indptr_type == C_API_DTYPE_INT32:
            out_indptr = cint32_array_to_numpy(out_ptr_indptr, indptr_len)
        elif indptr_type == C_API_DTYPE_INT64:
            out_indptr = cint64_array_to_numpy(out_ptr_indptr, indptr_len)
        else:
            raise TypeError("Expected int32 or int64 type for indptr")
        if data_type == C_API_DTYPE_FLOAT32:
            out_data = cfloat32_array_to_numpy(out_ptr_data, data_indices_len)
        elif data_type == C_API_DTYPE_FLOAT64:
            out_data = cfloat64_array_to_numpy(out_ptr_data, data_indices_len)
        else:
            raise TypeError("Expected float32 or float64 type for data")
        out_indices = cint32_array_to_numpy(out_ptr_indices, data_indices_len)
        # 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

987
    def __pred_for_csr(self, csr, start_iteration, num_iteration, predict_type):
988
        """Predict for a CSR data."""
989
        def inner_predict(csr, start_iteration, num_iteration, predict_type, preds=None):
990
            nrow = len(csr.indptr) - 1
991
            n_preds = self.__get_num_preds(start_iteration, num_iteration, nrow, predict_type)
992
            if preds is None:
993
                preds = np.empty(n_preds, dtype=np.float64)
994
995
996
997
998
999
1000
            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)

            ptr_indptr, type_ptr_indptr, __ = c_int_array(csr.indptr)
            ptr_data, type_ptr_data, _ = c_float_array(csr.data)

1001
            assert csr.shape[1] <= MAX_INT32
1002
            csr_indices = csr.indices.astype(np.int32, copy=False)
1003

1004
1005
1006
            _safe_call(_LIB.LGBM_BoosterPredictForCSR(
                self.handle,
                ptr_indptr,
1007
                ctypes.c_int(type_ptr_indptr),
1008
                csr_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
1009
1010
1011
1012
1013
1014
                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),
1015
                ctypes.c_int(start_iteration),
1016
1017
1018
1019
1020
1021
1022
                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
wxchan's avatar
wxchan committed
1023

1024
        def inner_predict_sparse(csr, start_iteration, num_iteration, predict_type):
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
            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)()
1038
            out_shape = np.empty(2, dtype=np.int64)
1039
1040
1041
            _safe_call(_LIB.LGBM_BoosterPredictSparseOutput(
                self.handle,
                ptr_indptr,
1042
                ctypes.c_int(type_ptr_indptr),
1043
1044
1045
1046
1047
1048
1049
                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),
1050
                ctypes.c_int(start_iteration),
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
                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(csr, out_shape, out_ptr_indptr, out_ptr_indices, out_ptr_data,
                                                   type_ptr_indptr, type_ptr_data, is_csr=True)
            nrow = len(csr.indptr) - 1
            return matrices, nrow

        if predict_type == C_API_PREDICT_CONTRIB:
1064
            return inner_predict_sparse(csr, start_iteration, num_iteration, predict_type)
1065
1066
1067
1068
        nrow = len(csr.indptr) - 1
        if nrow > MAX_INT32:
            sections = [0] + list(np.arange(start=MAX_INT32, stop=nrow, step=MAX_INT32)) + [nrow]
            # __get_num_preds() cannot work with nrow > MAX_INT32, so calculate overall number of predictions piecemeal
1069
            n_preds = [self.__get_num_preds(start_iteration, num_iteration, i, predict_type) for i in np.diff(sections)]
1070
            n_preds_sections = np.array([0] + n_preds, dtype=np.intp).cumsum()
1071
            preds = np.empty(sum(n_preds), dtype=np.float64)
1072
1073
            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:])):
1074
                # avoid memory consumption by arrays concatenation operations
1075
                inner_predict(csr[start_idx:end_idx], start_iteration, num_iteration, predict_type, preds[start_idx_pred:end_idx_pred])
1076
1077
            return preds, nrow
        else:
1078
            return inner_predict(csr, start_iteration, num_iteration, predict_type)
Guolin Ke's avatar
Guolin Ke committed
1079

1080
    def __pred_for_csc(self, csc, start_iteration, num_iteration, predict_type):
1081
        """Predict for a CSC data."""
1082
        def inner_predict_sparse(csc, start_iteration, num_iteration, predict_type):
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
            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)()
1096
            out_shape = np.empty(2, dtype=np.int64)
1097
1098
1099
            _safe_call(_LIB.LGBM_BoosterPredictSparseOutput(
                self.handle,
                ptr_indptr,
1100
                ctypes.c_int(type_ptr_indptr),
1101
1102
1103
1104
1105
1106
1107
                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),
1108
                ctypes.c_int(start_iteration),
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
                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(csc, out_shape, out_ptr_indptr, out_ptr_indices, out_ptr_data,
                                                   type_ptr_indptr, type_ptr_data, is_csr=False)
            nrow = csc.shape[0]
            return matrices, nrow

Guolin Ke's avatar
Guolin Ke committed
1121
        nrow = csc.shape[0]
1122
        if nrow > MAX_INT32:
1123
            return self.__pred_for_csr(csc.tocsr(), start_iteration, num_iteration, predict_type)
1124
        if predict_type == C_API_PREDICT_CONTRIB:
1125
1126
            return inner_predict_sparse(csc, start_iteration, num_iteration, predict_type)
        n_preds = self.__get_num_preds(start_iteration, num_iteration, nrow, predict_type)
1127
        preds = np.empty(n_preds, dtype=np.float64)
Guolin Ke's avatar
Guolin Ke committed
1128
1129
        out_num_preds = ctypes.c_int64(0)

1130
1131
        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
1132

1133
        assert csc.shape[0] <= MAX_INT32
1134
        csc_indices = csc.indices.astype(np.int32, copy=False)
1135

Guolin Ke's avatar
Guolin Ke committed
1136
1137
1138
        _safe_call(_LIB.LGBM_BoosterPredictForCSC(
            self.handle,
            ptr_indptr,
1139
            ctypes.c_int(type_ptr_indptr),
1140
            csc_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
Guolin Ke's avatar
Guolin Ke committed
1141
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1142
1143
1144
1145
1146
            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),
1147
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
1148
            ctypes.c_int(num_iteration),
1149
            c_str(self.pred_parameter),
Guolin Ke's avatar
Guolin Ke committed
1150
            ctypes.byref(out_num_preds),
wxchan's avatar
wxchan committed
1151
            preds.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
wxchan's avatar
wxchan committed
1152
        if n_preds != out_num_preds.value:
1153
            raise ValueError("Wrong length for predict results")
wxchan's avatar
wxchan committed
1154
1155
        return preds, nrow

1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
    def current_iteration(self):
        """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
1170

1171
class Dataset:
wxchan's avatar
wxchan committed
1172
    """Dataset in LightGBM."""
1173

1174
    def __init__(self, data, label=None, reference=None,
1175
                 weight=None, group=None, init_score=None, silent='warn',
1176
                 feature_name='auto', categorical_feature='auto', params=None,
wxchan's avatar
wxchan committed
1177
                 free_raw_data=True):
1178
        """Initialize Dataset.
1179

wxchan's avatar
wxchan committed
1180
1181
        Parameters
        ----------
1182
        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
1183
            Data source of Dataset.
1184
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM) or a LightGBM Dataset binary file.
1185
        label : list, numpy 1-D array, pandas Series / one-column DataFrame or None, optional (default=None)
1186
1187
1188
            Label of the data.
        reference : Dataset or None, optional (default=None)
            If this is Dataset for validation, training data should be used as reference.
1189
        weight : list, numpy 1-D array, pandas Series or None, optional (default=None)
wxchan's avatar
wxchan committed
1190
            Weight for each instance.
1191
        group : list, numpy 1-D array, pandas Series or None, optional (default=None)
1192
1193
1194
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
1195
1196
            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.
1197
        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)
1198
            Init score for Dataset.
1199
1200
        silent : bool, optional (default=False)
            Whether to print messages during construction.
1201
        feature_name : list of str, or 'auto', optional (default="auto")
1202
1203
            Feature names.
            If 'auto' and data is pandas DataFrame, data columns names are used.
1204
        categorical_feature : list of str or int, or 'auto', optional (default="auto")
1205
1206
            Categorical features.
            If list of int, interpreted as indices.
1207
            If list of str, interpreted as feature names (need to specify ``feature_name`` as well).
1208
            If 'auto' and data is pandas DataFrame, pandas unordered categorical columns are used.
1209
            All values in categorical features should be less than int32 max value (2147483647).
1210
            Large values could be memory consuming. Consider using consecutive integers starting from zero.
1211
            All negative values in categorical features will be treated as missing values.
1212
            The output cannot be monotonically constrained with respect to a categorical feature.
Nikita Titov's avatar
Nikita Titov committed
1213
        params : dict or None, optional (default=None)
1214
            Other parameters for Dataset.
Nikita Titov's avatar
Nikita Titov committed
1215
        free_raw_data : bool, optional (default=True)
1216
            If True, raw data is freed after constructing inner Dataset.
wxchan's avatar
wxchan committed
1217
        """
wxchan's avatar
wxchan committed
1218
1219
1220
1221
1222
1223
        self.handle = None
        self.data = data
        self.label = label
        self.reference = reference
        self.weight = weight
        self.group = group
1224
        self.init_score = init_score
wxchan's avatar
wxchan committed
1225
1226
        self.silent = silent
        self.feature_name = feature_name
1227
        self.categorical_feature = categorical_feature
1228
        self.params = deepcopy(params)
wxchan's avatar
wxchan committed
1229
1230
        self.free_raw_data = free_raw_data
        self.used_indices = None
1231
        self.need_slice = True
wxchan's avatar
wxchan committed
1232
        self._predictor = None
1233
        self.pandas_categorical = None
1234
        self.params_back_up = None
1235
1236
        self.feature_penalty = None
        self.monotone_constraints = None
1237
        self.version = 0
1238
        self._start_row = 0  # Used when pushing rows one by one.
wxchan's avatar
wxchan committed
1239
1240

    def __del__(self):
1241
1242
1243
1244
        try:
            self._free_handle()
        except AttributeError:
            pass
1245

1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
    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.
        """
        param_str = param_dict_to_str(self.get_params())
        sample_cnt = _get_sample_count(total_nrow, param_str)
        indices = np.empty(sample_cnt, dtype=np.int32)
        ptr_data, _, _ = c_int_array(indices)
        actual_sample_cnt = ctypes.c_int32(0)

        _safe_call(_LIB.LGBM_SampleIndices(
            ctypes.c_int32(total_nrow),
            c_str(param_str),
            ptr_data,
            ctypes.byref(actual_sample_cnt),
        ))
1275
1276
        assert sample_cnt == actual_sample_cnt.value
        return indices
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

    def _init_from_ref_dataset(self, total_nrow: int, ref_dataset: 'Dataset') -> 'Dataset':
        """Create dataset from a reference dataset.

        Parameters
        ----------
        total_nrow : int
            Number of rows expected to add to dataset.
        ref_dataset : Dataset
            Reference dataset to extract meta from.

        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
        ----------
1312
        sample_data : list of numpy array
1313
            Sample data for each column.
1314
        sample_indices : list of numpy array
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
            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):
            sample_col_ptr[i] = c_float_array(sample_data[i])[0]
            indices_col_ptr[i] = c_int_array(sample_indices[i])[0]

        num_per_col = np.array([len(d) for d in sample_indices], dtype=np.int32)
        num_per_col_ptr, _, _ = c_int_array(num_per_col)

        self.handle = ctypes.c_void_p()
        params_str = param_dict_to_str(self.get_params())
        _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),
            c_str(params_str),
            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)
        data_ptr, data_type, _ = c_float_array(data)

        _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

1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
    def get_params(self):
        """Get the used parameters in the Dataset.

        Returns
        -------
        params : dict or None
            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",
1411
                                                "linear_tree",
1412
1413
1414
1415
                                                "max_bin",
                                                "max_bin_by_feature",
                                                "min_data_in_bin",
                                                "pre_partition",
Nikita Titov's avatar
Nikita Titov committed
1416
                                                "precise_float_parser",
1417
1418
1419
1420
1421
1422
                                                "two_round",
                                                "use_missing",
                                                "weight_column",
                                                "zero_as_missing")
            return {k: v for k, v in self.params.items() if k in dataset_params}

1423
    def _free_handle(self):
1424
        if self.handle is not None:
1425
            _safe_call(_LIB.LGBM_DatasetFree(self.handle))
1426
            self.handle = None
Guolin Ke's avatar
Guolin Ke committed
1427
1428
1429
        self.need_slice = True
        if self.used_indices is not None:
            self.data = None
Nikita Titov's avatar
Nikita Titov committed
1430
        return self
wxchan's avatar
wxchan committed
1431

Guolin Ke's avatar
Guolin Ke committed
1432
1433
    def _set_init_score_by_predictor(self, predictor, data, used_indices=None):
        data_has_header = False
1434
        if isinstance(data, (str, Path)):
Guolin Ke's avatar
Guolin Ke committed
1435
            # check data has header or not
1436
            data_has_header = any(self.params.get(alias, False) for alias in _ConfigAliases.get("header"))
Guolin Ke's avatar
Guolin Ke committed
1437
        num_data = self.num_data()
1438
1439
1440
1441
1442
1443
1444
        if predictor is not None:
            init_score = predictor.predict(data,
                                           raw_score=True,
                                           data_has_header=data_has_header,
                                           is_reshape=False)
            if used_indices is not None:
                assert not self.need_slice
1445
                if isinstance(data, (str, Path)):
1446
                    sub_init_score = np.empty(num_data * predictor.num_class, dtype=np.float64)
1447
                    assert num_data == len(used_indices)
1448
1449
                    for i in range(len(used_indices)):
                        for j in range(predictor.num_class):
1450
1451
1452
1453
                            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
1454
                new_init_score = np.empty(init_score.size, dtype=np.float64)
1455
1456
                for i in range(num_data):
                    for j in range(predictor.num_class):
1457
1458
1459
                        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:
1460
            init_score = np.zeros(self.init_score.shape, dtype=np.float64)
1461
1462
        else:
            return self
Guolin Ke's avatar
Guolin Ke committed
1463
1464
        self.set_init_score(init_score)

1465
    def _lazy_init(self, data, label=None, reference=None,
1466
                   weight=None, group=None, init_score=None, predictor=None,
wxchan's avatar
wxchan committed
1467
                   silent=False, feature_name='auto',
1468
                   categorical_feature='auto', params=None):
wxchan's avatar
wxchan committed
1469
1470
        if data is None:
            self.handle = None
Nikita Titov's avatar
Nikita Titov committed
1471
            return self
Guolin Ke's avatar
Guolin Ke committed
1472
1473
1474
        if reference is not None:
            self.pandas_categorical = reference.pandas_categorical
            categorical_feature = reference.categorical_feature
1475
1476
1477
1478
        data, feature_name, categorical_feature, self.pandas_categorical = _data_from_pandas(data,
                                                                                             feature_name,
                                                                                             categorical_feature,
                                                                                             self.pandas_categorical)
wxchan's avatar
wxchan committed
1479
        label = _label_from_pandas(label)
Guolin Ke's avatar
Guolin Ke committed
1480

1481
        # process for args
wxchan's avatar
wxchan committed
1482
        params = {} if params is None else params
1483
1484
1485
        args_names = (getattr(self.__class__, '_lazy_init')
                      .__code__
                      .co_varnames[:getattr(self.__class__, '_lazy_init').__code__.co_argcount])
1486
        for key in params.keys():
1487
            if key in args_names:
1488
1489
                _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.')
1490
        # user can set verbose with params, it has higher priority
1491
1492
1493
1494
1495
        if silent != "warn":
            _log_warning("'silent' argument is deprecated and will be removed in a future release of LightGBM. "
                         "Pass 'verbose' parameter via 'params' instead.")
        else:
            silent = False
1496
        if not any(verbose_alias in params for verbose_alias in _ConfigAliases.get("verbosity")) and silent:
1497
            params["verbose"] = -1
1498
        # get categorical features
1499
1500
1501
1502
1503
1504
        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:
1505
                if isinstance(name, str) and name in feature_dict:
1506
                    categorical_indices.add(feature_dict[name])
1507
                elif isinstance(name, int):
1508
1509
                    categorical_indices.add(name)
                else:
1510
                    raise TypeError(f"Wrong type({type(name).__name__}) or unknown name({name}) in categorical_feature")
1511
            if categorical_indices:
1512
1513
                for cat_alias in _ConfigAliases.get("categorical_feature"):
                    if cat_alias in params:
1514
1515
1516
                        # If the params[cat_alias] is equal to categorical_indices, do not report the warning.
                        if not(isinstance(params[cat_alias], list) and set(params[cat_alias]) == categorical_indices):
                            _log_warning(f'{cat_alias} in param dict is overridden.')
1517
                        params.pop(cat_alias, None)
1518
                params['categorical_column'] = sorted(categorical_indices)
1519

wxchan's avatar
wxchan committed
1520
        params_str = param_dict_to_str(params)
1521
        self.params = params
1522
        # process for reference dataset
wxchan's avatar
wxchan committed
1523
        ref_dataset = None
wxchan's avatar
wxchan committed
1524
        if isinstance(reference, Dataset):
1525
            ref_dataset = reference.construct().handle
wxchan's avatar
wxchan committed
1526
1527
        elif reference is not None:
            raise TypeError('Reference dataset should be None or dataset instance')
1528
        # start construct data
1529
        if isinstance(data, (str, Path)):
wxchan's avatar
wxchan committed
1530
1531
            self.handle = ctypes.c_void_p()
            _safe_call(_LIB.LGBM_DatasetCreateFromFile(
1532
                c_str(str(data)),
wxchan's avatar
wxchan committed
1533
1534
1535
1536
1537
                c_str(params_str),
                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
1538
1539
        elif isinstance(data, scipy.sparse.csc_matrix):
            self.__init_from_csc(data, params_str, ref_dataset)
wxchan's avatar
wxchan committed
1540
1541
        elif isinstance(data, np.ndarray):
            self.__init_from_np2d(data, params_str, ref_dataset)
1542
1543
1544
1545
1546
1547
1548
1549
1550
        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)
1551
        elif isinstance(data, dt_DataTable):
1552
            self.__init_from_np2d(data.to_numpy(), params_str, ref_dataset)
wxchan's avatar
wxchan committed
1553
1554
1555
1556
        else:
            try:
                csr = scipy.sparse.csr_matrix(data)
                self.__init_from_csr(csr, params_str, ref_dataset)
1557
            except BaseException:
1558
                raise TypeError(f'Cannot initialize Dataset from {type(data).__name__}')
wxchan's avatar
wxchan committed
1559
1560
1561
        if label is not None:
            self.set_label(label)
        if self.get_label() is None:
1562
            raise ValueError("Label should not be None")
wxchan's avatar
wxchan committed
1563
1564
1565
1566
        if weight is not None:
            self.set_weight(weight)
        if group is not None:
            self.set_group(group)
1567
1568
        if isinstance(predictor, _InnerPredictor):
            if self._predictor is None and init_score is not None:
1569
                _log_warning("The init_score will be overridden by the prediction of init_model.")
Guolin Ke's avatar
Guolin Ke committed
1570
            self._set_init_score_by_predictor(predictor, data)
1571
1572
        elif init_score is not None:
            self.set_init_score(init_score)
Guolin Ke's avatar
Guolin Ke committed
1573
        elif predictor is not None:
1574
            raise TypeError(f'Wrong predictor type {type(predictor).__name__}')
Guolin Ke's avatar
Guolin Ke committed
1575
        # set feature names
Nikita Titov's avatar
Nikita Titov committed
1576
        return self.set_feature_name(feature_name)
wxchan's avatar
wxchan committed
1577

1578
1579
    @staticmethod
    def _yield_row_from_seqlist(seqs: List[Sequence], indices: Iterable[int]):
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
        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.
1605
        sampled = np.array([row for row in self._yield_row_from_seqlist(seqs, indices)])
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
        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

    def __init_from_seqs(self, seqs: List[Sequence], ref_dataset: Optional['Dataset'] = None):
        """
        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:
            param_str = param_dict_to_str(self.get_params())
            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

wxchan's avatar
wxchan committed
1650
    def __init_from_np2d(self, mat, params_str, ref_dataset):
1651
        """Initialize data from a 2-D numpy matrix."""
wxchan's avatar
wxchan committed
1652
1653
1654
1655
1656
1657
        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)
1658
        else:  # change non-float data to float data, need to copy
wxchan's avatar
wxchan committed
1659
1660
            data = np.array(mat.reshape(mat.size), dtype=np.float32)

1661
        ptr_data, type_ptr_data, _ = c_float_array(data)
wxchan's avatar
wxchan committed
1662
1663
        _safe_call(_LIB.LGBM_DatasetCreateFromMat(
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1664
            ctypes.c_int(type_ptr_data),
1665
1666
            ctypes.c_int32(mat.shape[0]),
            ctypes.c_int32(mat.shape[1]),
Guolin Ke's avatar
Guolin Ke committed
1667
            ctypes.c_int(C_API_IS_ROW_MAJOR),
wxchan's avatar
wxchan committed
1668
1669
1670
            c_str(params_str),
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
1671
        return self
wxchan's avatar
wxchan committed
1672

1673
    def __init_from_list_np2d(self, mats, params_str, ref_dataset):
1674
        """Initialize data from a list of 2-D numpy matrices."""
1675
        ncol = mats[0].shape[1]
1676
        nrow = np.empty((len(mats),), np.int32)
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
        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)
1696
            else:  # change non-float data to float data, need to copy
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
                mats[i] = np.array(mat.reshape(mat.size), dtype=np.float32)

            chunk_ptr_data, chunk_type_ptr_data, holder = c_float_array(mats[i])
            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(
1708
            ctypes.c_int32(len(mats)),
1709
1710
1711
            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)),
1712
            ctypes.c_int32(ncol),
1713
1714
1715
1716
            ctypes.c_int(C_API_IS_ROW_MAJOR),
            c_str(params_str),
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
1717
        return self
1718

wxchan's avatar
wxchan committed
1719
    def __init_from_csr(self, csr, params_str, ref_dataset):
1720
        """Initialize data from a CSR matrix."""
wxchan's avatar
wxchan committed
1721
        if len(csr.indices) != len(csr.data):
1722
            raise ValueError(f'Length mismatch: {len(csr.indices)} vs {len(csr.data)}')
wxchan's avatar
wxchan committed
1723
1724
        self.handle = ctypes.c_void_p()

1725
1726
        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
1727

1728
        assert csr.shape[1] <= MAX_INT32
1729
        csr_indices = csr.indices.astype(np.int32, copy=False)
1730

wxchan's avatar
wxchan committed
1731
1732
        _safe_call(_LIB.LGBM_DatasetCreateFromCSR(
            ptr_indptr,
Guolin Ke's avatar
Guolin Ke committed
1733
            ctypes.c_int(type_ptr_indptr),
1734
            csr_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
wxchan's avatar
wxchan committed
1735
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1736
1737
1738
1739
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csr.indptr)),
            ctypes.c_int64(len(csr.data)),
            ctypes.c_int64(csr.shape[1]),
wxchan's avatar
wxchan committed
1740
1741
1742
            c_str(params_str),
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
1743
        return self
wxchan's avatar
wxchan committed
1744

Guolin Ke's avatar
Guolin Ke committed
1745
    def __init_from_csc(self, csc, params_str, ref_dataset):
1746
        """Initialize data from a CSC matrix."""
Guolin Ke's avatar
Guolin Ke committed
1747
        if len(csc.indices) != len(csc.data):
1748
            raise ValueError(f'Length mismatch: {len(csc.indices)} vs {len(csc.data)}')
Guolin Ke's avatar
Guolin Ke committed
1749
1750
        self.handle = ctypes.c_void_p()

1751
1752
        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
1753

1754
        assert csc.shape[0] <= MAX_INT32
1755
        csc_indices = csc.indices.astype(np.int32, copy=False)
1756

Guolin Ke's avatar
Guolin Ke committed
1757
1758
        _safe_call(_LIB.LGBM_DatasetCreateFromCSC(
            ptr_indptr,
Guolin Ke's avatar
Guolin Ke committed
1759
            ctypes.c_int(type_ptr_indptr),
1760
            csc_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
Guolin Ke's avatar
Guolin Ke committed
1761
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
1762
1763
1764
1765
            ctypes.c_int(type_ptr_data),
            ctypes.c_int64(len(csc.indptr)),
            ctypes.c_int64(len(csc.data)),
            ctypes.c_int64(csc.shape[0]),
Guolin Ke's avatar
Guolin Ke committed
1766
1767
1768
            c_str(params_str),
            ref_dataset,
            ctypes.byref(self.handle)))
Nikita Titov's avatar
Nikita Titov committed
1769
        return self
Guolin Ke's avatar
Guolin Ke committed
1770

1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
    @staticmethod
    def _compare_params_for_warning(params, other_params):
        """Compare params.

        It is only for the warning purpose. Thus some keys are ignored.

        Returns
        -------
        compare_result: bool
          If they are equal, return True; Otherwise, return False.
        """
        ignore_keys = _ConfigAliases.get("categorical_feature")
        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

wxchan's avatar
wxchan committed
1797
    def construct(self):
1798
1799
1800
1801
1802
        """Lazy init.

        Returns
        -------
        self : Dataset
Nikita Titov's avatar
Nikita Titov committed
1803
            Constructed Dataset object.
1804
        """
1805
        if self.handle is None:
wxchan's avatar
wxchan committed
1806
            if self.reference is not None:
1807
                reference_params = self.reference.get_params()
1808
1809
1810
1811
                params = self.get_params()
                if params != reference_params:
                    if self._compare_params_for_warning(params, reference_params) is False:
                        _log_warning('Overriding the parameters from Reference Dataset.')
1812
                    self._update_params(reference_params)
wxchan's avatar
wxchan committed
1813
                if self.used_indices is None:
1814
                    # create valid
1815
                    self._lazy_init(self.data, label=self.label, reference=self.reference,
1816
1817
                                    weight=self.weight, group=self.group,
                                    init_score=self.init_score, predictor=self._predictor,
1818
                                    silent=self.silent, feature_name=self.feature_name, params=self.params)
wxchan's avatar
wxchan committed
1819
                else:
1820
                    # construct subset
wxchan's avatar
wxchan committed
1821
                    used_indices = list_to_1d_numpy(self.used_indices, np.int32, name='used_indices')
1822
                    assert used_indices.flags.c_contiguous
Guolin Ke's avatar
Guolin Ke committed
1823
                    if self.reference.group is not None:
1824
                        group_info = np.array(self.reference.group).astype(np.int32, copy=False)
1825
                        _, self.group = np.unique(np.repeat(range(len(group_info)), repeats=group_info)[self.used_indices],
1826
                                                  return_counts=True)
1827
                    self.handle = ctypes.c_void_p()
wxchan's avatar
wxchan committed
1828
1829
                    params_str = param_dict_to_str(self.params)
                    _safe_call(_LIB.LGBM_DatasetGetSubset(
1830
                        self.reference.construct().handle,
wxchan's avatar
wxchan committed
1831
                        used_indices.ctypes.data_as(ctypes.POINTER(ctypes.c_int32)),
1832
                        ctypes.c_int32(used_indices.shape[0]),
wxchan's avatar
wxchan committed
1833
1834
                        c_str(params_str),
                        ctypes.byref(self.handle)))
Guolin Ke's avatar
Guolin Ke committed
1835
1836
                    if not self.free_raw_data:
                        self.get_data()
Guolin Ke's avatar
Guolin Ke committed
1837
1838
                    if self.group is not None:
                        self.set_group(self.group)
wxchan's avatar
wxchan committed
1839
1840
                    if self.get_label() is None:
                        raise ValueError("Label should not be None.")
Guolin Ke's avatar
Guolin Ke committed
1841
1842
1843
                    if isinstance(self._predictor, _InnerPredictor) and self._predictor is not self.reference._predictor:
                        self.get_data()
                        self._set_init_score_by_predictor(self._predictor, self.data, used_indices)
wxchan's avatar
wxchan committed
1844
            else:
1845
                # create train
1846
                self._lazy_init(self.data, label=self.label,
1847
1848
1849
                                weight=self.weight, group=self.group,
                                init_score=self.init_score, predictor=self._predictor,
                                silent=self.silent, feature_name=self.feature_name,
1850
                                categorical_feature=self.categorical_feature, params=self.params)
wxchan's avatar
wxchan committed
1851
1852
1853
            if self.free_raw_data:
                self.data = None
        return self
wxchan's avatar
wxchan committed
1854

wxchan's avatar
wxchan committed
1855
    def create_valid(self, data, label=None, weight=None, group=None,
1856
                     init_score=None, silent='warn', params=None):
1857
        """Create validation data align with current Dataset.
wxchan's avatar
wxchan committed
1858
1859
1860

        Parameters
        ----------
1861
        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
1862
            Data source of Dataset.
1863
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM) or a LightGBM Dataset binary file.
1864
        label : list, numpy 1-D array, pandas Series / one-column DataFrame or None, optional (default=None)
1865
1866
            Label of the data.
        weight : list, numpy 1-D array, pandas Series or None, optional (default=None)
wxchan's avatar
wxchan committed
1867
            Weight for each instance.
1868
        group : list, numpy 1-D array, pandas Series or None, optional (default=None)
1869
1870
1871
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
1872
1873
            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.
1874
        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)
1875
            Init score for Dataset.
1876
1877
        silent : bool, optional (default=False)
            Whether to print messages during construction.
Nikita Titov's avatar
Nikita Titov committed
1878
        params : dict or None, optional (default=None)
1879
            Other parameters for validation Dataset.
1880
1881
1882

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
1883
1884
        valid : Dataset
            Validation Dataset with reference to self.
wxchan's avatar
wxchan committed
1885
        """
1886
        ret = Dataset(data, label=label, reference=self,
1887
1888
                      weight=weight, group=group, init_score=init_score,
                      silent=silent, params=params, free_raw_data=self.free_raw_data)
wxchan's avatar
wxchan committed
1889
        ret._predictor = self._predictor
1890
        ret.pandas_categorical = self.pandas_categorical
wxchan's avatar
wxchan committed
1891
        return ret
wxchan's avatar
wxchan committed
1892

wxchan's avatar
wxchan committed
1893
    def subset(self, used_indices, params=None):
1894
        """Get subset of current Dataset.
wxchan's avatar
wxchan committed
1895
1896
1897
1898

        Parameters
        ----------
        used_indices : list of int
1899
            Indices used to create the subset.
Nikita Titov's avatar
Nikita Titov committed
1900
        params : dict or None, optional (default=None)
1901
            These parameters will be passed to Dataset constructor.
1902
1903
1904
1905
1906

        Returns
        -------
        subset : Dataset
            Subset of the current Dataset.
wxchan's avatar
wxchan committed
1907
        """
wxchan's avatar
wxchan committed
1908
1909
        if params is None:
            params = self.params
wxchan's avatar
wxchan committed
1910
        ret = Dataset(None, reference=self, feature_name=self.feature_name,
1911
1912
                      categorical_feature=self.categorical_feature, params=params,
                      free_raw_data=self.free_raw_data)
wxchan's avatar
wxchan committed
1913
        ret._predictor = self._predictor
1914
        ret.pandas_categorical = self.pandas_categorical
1915
        ret.used_indices = sorted(used_indices)
wxchan's avatar
wxchan committed
1916
1917
1918
        return ret

    def save_binary(self, filename):
1919
        """Save Dataset to a binary file.
wxchan's avatar
wxchan committed
1920

1921
1922
1923
1924
1925
        .. 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
1926
1927
        Parameters
        ----------
1928
        filename : str or pathlib.Path
wxchan's avatar
wxchan committed
1929
            Name of the output file.
Nikita Titov's avatar
Nikita Titov committed
1930
1931
1932
1933
1934

        Returns
        -------
        self : Dataset
            Returns self.
wxchan's avatar
wxchan committed
1935
1936
1937
        """
        _safe_call(_LIB.LGBM_DatasetSaveBinary(
            self.construct().handle,
1938
            c_str(str(filename))))
Nikita Titov's avatar
Nikita Titov committed
1939
        return self
wxchan's avatar
wxchan committed
1940
1941

    def _update_params(self, params):
1942
1943
        if not params:
            return self
1944
        params = deepcopy(params)
1945
1946
1947
1948
1949

        def update():
            if not self.params:
                self.params = params
            else:
1950
                self.params_back_up = deepcopy(self.params)
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
                self.params.update(params)

        if self.handle is None:
            update()
        elif params is not None:
            ret = _LIB.LGBM_DatasetUpdateParamChecking(
                c_str(param_dict_to_str(self.params)),
                c_str(param_dict_to_str(params)))
            if ret != 0:
                # could be updated if data is not freed
                if self.data is not None:
                    update()
                    self._free_handle()
                else:
1965
                    raise LightGBMError(_LIB.LGBM_GetLastError().decode('utf-8'))
Nikita Titov's avatar
Nikita Titov committed
1966
        return self
wxchan's avatar
wxchan committed
1967

1968
    def _reverse_update_params(self):
1969
        if self.handle is None:
1970
            self.params = deepcopy(self.params_back_up)
1971
            self.params_back_up = None
Nikita Titov's avatar
Nikita Titov committed
1972
        return self
1973

wxchan's avatar
wxchan committed
1974
    def set_field(self, field_name, data):
wxchan's avatar
wxchan committed
1975
        """Set property into the Dataset.
wxchan's avatar
wxchan committed
1976
1977
1978

        Parameters
        ----------
1979
        field_name : str
1980
            The field name of the information.
1981
1982
        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
1983
1984
1985
1986
1987

        Returns
        -------
        self : Dataset
            Dataset with set property.
wxchan's avatar
wxchan committed
1988
        """
1989
        if self.handle is None:
1990
            raise Exception(f"Cannot set {field_name} before construct dataset")
wxchan's avatar
wxchan committed
1991
        if data is None:
1992
            # set to None
wxchan's avatar
wxchan committed
1993
1994
1995
1996
            _safe_call(_LIB.LGBM_DatasetSetField(
                self.handle,
                c_str(field_name),
                None,
Guolin Ke's avatar
Guolin Ke committed
1997
1998
                ctypes.c_int(0),
                ctypes.c_int(FIELD_TYPE_MAPPER[field_name])))
Nikita Titov's avatar
Nikita Titov committed
1999
            return self
2000
        if field_name == 'init_score':
Guolin Ke's avatar
Guolin Ke committed
2001
            dtype = np.float64
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
            if _is_1d_collection(data):
                data = list_to_1d_numpy(data, dtype, name=field_name)
            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
            data = list_to_1d_numpy(data, dtype, name=field_name)

2016
2017
        if data.dtype == np.float32 or data.dtype == np.float64:
            ptr_data, type_data, _ = c_float_array(data)
wxchan's avatar
wxchan committed
2018
        elif data.dtype == np.int32:
2019
            ptr_data, type_data, _ = c_int_array(data)
wxchan's avatar
wxchan committed
2020
        else:
2021
            raise TypeError(f"Expected np.float32/64 or np.int32, met type({data.dtype})")
wxchan's avatar
wxchan committed
2022
        if type_data != FIELD_TYPE_MAPPER[field_name]:
2023
            raise TypeError("Input type error for set_field")
wxchan's avatar
wxchan committed
2024
2025
2026
2027
        _safe_call(_LIB.LGBM_DatasetSetField(
            self.handle,
            c_str(field_name),
            ptr_data,
Guolin Ke's avatar
Guolin Ke committed
2028
2029
            ctypes.c_int(len(data)),
            ctypes.c_int(type_data)))
2030
        self.version += 1
Nikita Titov's avatar
Nikita Titov committed
2031
        return self
wxchan's avatar
wxchan committed
2032

wxchan's avatar
wxchan committed
2033
2034
    def get_field(self, field_name):
        """Get property from the Dataset.
wxchan's avatar
wxchan committed
2035
2036
2037

        Parameters
        ----------
2038
        field_name : str
2039
            The field name of the information.
wxchan's avatar
wxchan committed
2040
2041
2042

        Returns
        -------
2043
        info : numpy array or None
2044
            A numpy array with information from the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2045
        """
2046
        if self.handle is None:
2047
            raise Exception(f"Cannot get {field_name} before construct Dataset")
2048
2049
        tmp_out_len = ctypes.c_int(0)
        out_type = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
        ret = ctypes.POINTER(ctypes.c_void_p)()
        _safe_call(_LIB.LGBM_DatasetGetField(
            self.handle,
            c_str(field_name),
            ctypes.byref(tmp_out_len),
            ctypes.byref(ret),
            ctypes.byref(out_type)))
        if out_type.value != FIELD_TYPE_MAPPER[field_name]:
            raise TypeError("Return type error for get_field")
        if tmp_out_len.value == 0:
            return None
        if out_type.value == C_API_DTYPE_INT32:
2062
            arr = cint32_array_to_numpy(ctypes.cast(ret, ctypes.POINTER(ctypes.c_int32)), tmp_out_len.value)
wxchan's avatar
wxchan committed
2063
        elif out_type.value == C_API_DTYPE_FLOAT32:
2064
            arr = cfloat32_array_to_numpy(ctypes.cast(ret, ctypes.POINTER(ctypes.c_float)), tmp_out_len.value)
Guolin Ke's avatar
Guolin Ke committed
2065
        elif out_type.value == C_API_DTYPE_FLOAT64:
2066
            arr = cfloat64_array_to_numpy(ctypes.cast(ret, ctypes.POINTER(ctypes.c_double)), tmp_out_len.value)
2067
        else:
wxchan's avatar
wxchan committed
2068
            raise TypeError("Unknown type")
2069
2070
2071
2072
2073
2074
        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
2075

2076
    def set_categorical_feature(self, categorical_feature):
2077
        """Set categorical features.
2078
2079
2080

        Parameters
        ----------
2081
        categorical_feature : list of int or str
2082
            Names or indices of categorical features.
Nikita Titov's avatar
Nikita Titov committed
2083
2084
2085
2086
2087

        Returns
        -------
        self : Dataset
            Dataset with set categorical features.
2088
2089
        """
        if self.categorical_feature == categorical_feature:
Nikita Titov's avatar
Nikita Titov committed
2090
            return self
2091
        if self.data is not None:
2092
2093
            if self.categorical_feature is None:
                self.categorical_feature = categorical_feature
Nikita Titov's avatar
Nikita Titov committed
2094
                return self._free_handle()
2095
            elif categorical_feature == 'auto':
Nikita Titov's avatar
Nikita Titov committed
2096
                return self
2097
            else:
2098
2099
2100
                if self.categorical_feature != 'auto':
                    _log_warning('categorical_feature in Dataset is overridden.\n'
                                 f'New categorical_feature is {sorted(list(categorical_feature))}')
2101
                self.categorical_feature = categorical_feature
Nikita Titov's avatar
Nikita Titov committed
2102
                return self._free_handle()
2103
        else:
2104
2105
            raise LightGBMError("Cannot set categorical feature after freed raw data, "
                                "set free_raw_data=False when construct Dataset to avoid this.")
2106

Guolin Ke's avatar
Guolin Ke committed
2107
    def _set_predictor(self, predictor):
2108
2109
2110
2111
        """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
2112
        """
2113
        if predictor is self._predictor and (predictor is None or predictor.current_iteration() == self._predictor.current_iteration()):
Nikita Titov's avatar
Nikita Titov committed
2114
            return self
2115
        if self.handle is None:
Guolin Ke's avatar
Guolin Ke committed
2116
            self._predictor = predictor
2117
2118
2119
2120
2121
2122
        elif self.data is not None:
            self._predictor = predictor
            self._set_init_score_by_predictor(self._predictor, self.data)
        elif self.used_indices is not None and self.reference is not None and self.reference.data is not None:
            self._predictor = predictor
            self._set_init_score_by_predictor(self._predictor, self.reference.data, self.used_indices)
Guolin Ke's avatar
Guolin Ke committed
2123
        else:
2124
2125
            raise LightGBMError("Cannot set predictor after freed raw data, "
                                "set free_raw_data=False when construct Dataset to avoid this.")
2126
        return self
Guolin Ke's avatar
Guolin Ke committed
2127
2128

    def set_reference(self, reference):
2129
        """Set reference Dataset.
Guolin Ke's avatar
Guolin Ke committed
2130
2131
2132
2133

        Parameters
        ----------
        reference : Dataset
2134
            Reference that is used as a template to construct the current Dataset.
Nikita Titov's avatar
Nikita Titov committed
2135
2136
2137
2138
2139

        Returns
        -------
        self : Dataset
            Dataset with set reference.
Guolin Ke's avatar
Guolin Ke committed
2140
        """
2141
2142
2143
        self.set_categorical_feature(reference.categorical_feature) \
            .set_feature_name(reference.feature_name) \
            ._set_predictor(reference._predictor)
2144
        # we're done if self and reference share a common upstream reference
2145
        if self.get_ref_chain().intersection(reference.get_ref_chain()):
Nikita Titov's avatar
Nikita Titov committed
2146
            return self
Guolin Ke's avatar
Guolin Ke committed
2147
2148
        if self.data is not None:
            self.reference = reference
Nikita Titov's avatar
Nikita Titov committed
2149
            return self._free_handle()
Guolin Ke's avatar
Guolin Ke committed
2150
        else:
2151
2152
            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
2153
2154

    def set_feature_name(self, feature_name):
2155
        """Set feature name.
Guolin Ke's avatar
Guolin Ke committed
2156
2157
2158

        Parameters
        ----------
2159
        feature_name : list of str
2160
            Feature names.
Nikita Titov's avatar
Nikita Titov committed
2161
2162
2163
2164
2165

        Returns
        -------
        self : Dataset
            Dataset with set feature name.
Guolin Ke's avatar
Guolin Ke committed
2166
        """
2167
2168
        if feature_name != 'auto':
            self.feature_name = feature_name
2169
        if self.handle is not None and feature_name is not None and feature_name != 'auto':
wxchan's avatar
wxchan committed
2170
            if len(feature_name) != self.num_feature():
2171
                raise ValueError(f"Length of feature_name({len(feature_name)}) and num_feature({self.num_feature()}) don't match")
2172
            c_feature_name = [c_str(name) for name in feature_name]
wxchan's avatar
wxchan committed
2173
2174
2175
            _safe_call(_LIB.LGBM_DatasetSetFeatureNames(
                self.handle,
                c_array(ctypes.c_char_p, c_feature_name),
Guolin Ke's avatar
Guolin Ke committed
2176
                ctypes.c_int(len(feature_name))))
Nikita Titov's avatar
Nikita Titov committed
2177
        return self
Guolin Ke's avatar
Guolin Ke committed
2178
2179

    def set_label(self, label):
2180
        """Set label of Dataset.
Guolin Ke's avatar
Guolin Ke committed
2181
2182
2183

        Parameters
        ----------
2184
        label : list, numpy 1-D array, pandas Series / one-column DataFrame or None
2185
            The label information to be set into Dataset.
Nikita Titov's avatar
Nikita Titov committed
2186
2187
2188
2189
2190

        Returns
        -------
        self : Dataset
            Dataset with set label.
Guolin Ke's avatar
Guolin Ke committed
2191
2192
        """
        self.label = label
2193
        if self.handle is not None:
2194
            label = list_to_1d_numpy(_label_from_pandas(label), name='label')
wxchan's avatar
wxchan committed
2195
            self.set_field('label', label)
2196
            self.label = self.get_field('label')  # original values can be modified at cpp side
Nikita Titov's avatar
Nikita Titov committed
2197
        return self
Guolin Ke's avatar
Guolin Ke committed
2198
2199

    def set_weight(self, weight):
2200
        """Set weight of each instance.
Guolin Ke's avatar
Guolin Ke committed
2201
2202
2203

        Parameters
        ----------
2204
        weight : list, numpy 1-D array, pandas Series or None
2205
            Weight to be set for each data point.
Nikita Titov's avatar
Nikita Titov committed
2206
2207
2208
2209
2210

        Returns
        -------
        self : Dataset
            Dataset with set weight.
Guolin Ke's avatar
Guolin Ke committed
2211
        """
2212
2213
        if weight is not None and np.all(weight == 1):
            weight = None
Guolin Ke's avatar
Guolin Ke committed
2214
        self.weight = weight
2215
        if self.handle is not None and weight is not None:
wxchan's avatar
wxchan committed
2216
2217
            weight = list_to_1d_numpy(weight, name='weight')
            self.set_field('weight', weight)
2218
            self.weight = self.get_field('weight')  # original values can be modified at cpp side
Nikita Titov's avatar
Nikita Titov committed
2219
        return self
Guolin Ke's avatar
Guolin Ke committed
2220
2221

    def set_init_score(self, init_score):
2222
        """Set init score of Booster to start from.
Guolin Ke's avatar
Guolin Ke committed
2223
2224
2225

        Parameters
        ----------
2226
        init_score : list, list of lists (for multi-class task), numpy array, pandas Series, pandas DataFrame (for multi-class task), or None
2227
            Init score for Booster.
Nikita Titov's avatar
Nikita Titov committed
2228
2229
2230
2231
2232

        Returns
        -------
        self : Dataset
            Dataset with set init score.
Guolin Ke's avatar
Guolin Ke committed
2233
2234
        """
        self.init_score = init_score
2235
        if self.handle is not None and init_score is not None:
wxchan's avatar
wxchan committed
2236
            self.set_field('init_score', init_score)
2237
            self.init_score = self.get_field('init_score')  # original values can be modified at cpp side
Nikita Titov's avatar
Nikita Titov committed
2238
        return self
Guolin Ke's avatar
Guolin Ke committed
2239
2240

    def set_group(self, group):
2241
        """Set group size of Dataset (used for ranking).
Guolin Ke's avatar
Guolin Ke committed
2242
2243
2244

        Parameters
        ----------
2245
        group : list, numpy 1-D array, pandas Series or None
2246
2247
2248
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
2249
2250
            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
2251
2252
2253
2254
2255

        Returns
        -------
        self : Dataset
            Dataset with set group.
Guolin Ke's avatar
Guolin Ke committed
2256
2257
        """
        self.group = group
2258
        if self.handle is not None and group is not None:
wxchan's avatar
wxchan committed
2259
2260
            group = list_to_1d_numpy(group, np.int32, name='group')
            self.set_field('group', group)
Nikita Titov's avatar
Nikita Titov committed
2261
        return self
Guolin Ke's avatar
Guolin Ke committed
2262

2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
    def get_feature_name(self):
        """Get the names of columns (features) in the Dataset.

        Returns
        -------
        feature_names : list
            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)
2277
        string_buffers = [ctypes.create_string_buffer(reserved_string_buffer_size) for _ in range(num_feature)]
2278
2279
2280
        ptr_string_buffers = (ctypes.c_char_p * num_feature)(*map(ctypes.addressof, string_buffers))
        _safe_call(_LIB.LGBM_DatasetGetFeatureNames(
            self.handle,
2281
            ctypes.c_int(num_feature),
2282
            ctypes.byref(tmp_out_len),
2283
            ctypes.c_size_t(reserved_string_buffer_size),
2284
2285
2286
2287
            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")
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
        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))
2300
        return [string_buffers[i].value.decode('utf-8') for i in range(num_feature)]
2301

Guolin Ke's avatar
Guolin Ke committed
2302
    def get_label(self):
2303
        """Get the label of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2304
2305
2306

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2307
        label : numpy array or None
2308
            The label information from the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2309
        """
2310
        if self.label is None:
wxchan's avatar
wxchan committed
2311
            self.label = self.get_field('label')
Guolin Ke's avatar
Guolin Ke committed
2312
2313
2314
        return self.label

    def get_weight(self):
2315
        """Get the weight of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2316
2317
2318

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2319
        weight : numpy array or None
2320
            Weight for each data point from the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2321
        """
2322
        if self.weight is None:
wxchan's avatar
wxchan committed
2323
            self.weight = self.get_field('weight')
Guolin Ke's avatar
Guolin Ke committed
2324
2325
2326
        return self.weight

    def get_init_score(self):
2327
        """Get the initial score of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2328
2329
2330

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2331
        init_score : numpy array or None
2332
            Init score of Booster.
Guolin Ke's avatar
Guolin Ke committed
2333
        """
2334
        if self.init_score is None:
wxchan's avatar
wxchan committed
2335
            self.init_score = self.get_field('init_score')
Guolin Ke's avatar
Guolin Ke committed
2336
2337
        return self.init_score

2338
2339
2340
2341
2342
    def get_data(self):
        """Get the raw data of the Dataset.

        Returns
        -------
2343
        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
2344
2345
2346
2347
            Raw data used in the Dataset construction.
        """
        if self.handle is None:
            raise Exception("Cannot get data before construct Dataset")
Guolin Ke's avatar
Guolin Ke committed
2348
2349
2350
2351
2352
        if self.need_slice and self.used_indices is not None and self.reference is not None:
            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, :]
2353
                elif isinstance(self.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2354
                    self.data = self.data.iloc[self.used_indices].copy()
2355
                elif isinstance(self.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2356
                    self.data = self.data[self.used_indices, :]
2357
2358
2359
2360
                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
2361
                else:
2362
2363
                    _log_warning(f"Cannot subset {type(self.data).__name__} type of raw data.\n"
                                 "Returning original raw data")
2364
            self.need_slice = False
Guolin Ke's avatar
Guolin Ke committed
2365
2366
2367
        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.")
2368
2369
        return self.data

Guolin Ke's avatar
Guolin Ke committed
2370
    def get_group(self):
2371
        """Get the group of the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2372
2373
2374

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
2375
        group : numpy array or None
2376
2377
2378
            Group/query data.
            Only used in the learning-to-rank task.
            sum(group) = n_samples.
2379
2380
            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
2381
        """
2382
        if self.group is None:
wxchan's avatar
wxchan committed
2383
            self.group = self.get_field('group')
Guolin Ke's avatar
Guolin Ke committed
2384
2385
            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
2386
                self.group = np.diff(self.group)
Guolin Ke's avatar
Guolin Ke committed
2387
2388
2389
        return self.group

    def num_data(self):
2390
        """Get the number of rows in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2391
2392
2393

        Returns
        -------
2394
2395
        number_of_rows : int
            The number of rows in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2396
        """
2397
        if self.handle is not None:
2398
            ret = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2399
2400
2401
            _safe_call(_LIB.LGBM_DatasetGetNumData(self.handle,
                                                   ctypes.byref(ret)))
            return ret.value
Guolin Ke's avatar
Guolin Ke committed
2402
        else:
2403
            raise LightGBMError("Cannot get num_data before construct dataset")
Guolin Ke's avatar
Guolin Ke committed
2404
2405

    def num_feature(self):
2406
        """Get the number of columns (features) in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2407
2408
2409

        Returns
        -------
2410
2411
        number_of_columns : int
            The number of columns (features) in the Dataset.
Guolin Ke's avatar
Guolin Ke committed
2412
        """
2413
        if self.handle is not None:
2414
            ret = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2415
2416
2417
            _safe_call(_LIB.LGBM_DatasetGetNumFeature(self.handle,
                                                      ctypes.byref(ret)))
            return ret.value
Guolin Ke's avatar
Guolin Ke committed
2418
        else:
2419
            raise LightGBMError("Cannot get num_feature before construct dataset")
Guolin Ke's avatar
Guolin Ke committed
2420

2421
    def get_ref_chain(self, ref_limit=100):
2422
2423
2424
2425
2426
        """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.
2427
2428
2429
2430
2431

        Parameters
        ----------
        ref_limit : int, optional (default=100)
            The limit number of references.
2432
2433
2434

        Returns
        -------
2435
2436
2437
        ref_chain : set of Dataset
            Chain of references of the Datasets.
        """
2438
        head = self
2439
        ref_chain = set()
2440
2441
        while len(ref_chain) < ref_limit:
            if isinstance(head, Dataset):
2442
                ref_chain.add(head)
2443
2444
2445
2446
2447
2448
                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
2449
        return ref_chain
2450

2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
    def add_features_from(self, other):
        """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
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
        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()))
2479
                elif isinstance(other.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2480
                    self.data = np.hstack((self.data, other.data.values))
2481
                elif isinstance(other.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2482
2483
2484
2485
2486
2487
2488
                    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)
2489
                elif isinstance(other.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2490
                    self.data = scipy.sparse.hstack((self.data, other.data.values), format=sparse_format)
2491
                elif isinstance(other.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2492
2493
2494
                    self.data = scipy.sparse.hstack((self.data, other.data.to_numpy()), format=sparse_format)
                else:
                    self.data = None
2495
            elif isinstance(self.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2496
2497
                if not PANDAS_INSTALLED:
                    raise LightGBMError("Cannot add features to DataFrame type of raw data "
2498
2499
                                        "without pandas installed. "
                                        "Install pandas and restart your session.")
Guolin Ke's avatar
Guolin Ke committed
2500
                if isinstance(other.data, np.ndarray):
2501
                    self.data = concat((self.data, pd_DataFrame(other.data)),
Guolin Ke's avatar
Guolin Ke committed
2502
2503
                                       axis=1, ignore_index=True)
                elif scipy.sparse.issparse(other.data):
2504
                    self.data = concat((self.data, pd_DataFrame(other.data.toarray())),
Guolin Ke's avatar
Guolin Ke committed
2505
                                       axis=1, ignore_index=True)
2506
                elif isinstance(other.data, pd_DataFrame):
Guolin Ke's avatar
Guolin Ke committed
2507
2508
                    self.data = concat((self.data, other.data),
                                       axis=1, ignore_index=True)
2509
2510
                elif isinstance(other.data, dt_DataTable):
                    self.data = concat((self.data, pd_DataFrame(other.data.to_numpy())),
Guolin Ke's avatar
Guolin Ke committed
2511
2512
2513
                                       axis=1, ignore_index=True)
                else:
                    self.data = None
2514
            elif isinstance(self.data, dt_DataTable):
Guolin Ke's avatar
Guolin Ke committed
2515
                if isinstance(other.data, np.ndarray):
2516
                    self.data = dt_DataTable(np.hstack((self.data.to_numpy(), other.data)))
Guolin Ke's avatar
Guolin Ke committed
2517
                elif scipy.sparse.issparse(other.data):
2518
2519
2520
2521
2522
                    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
2523
2524
2525
2526
2527
                else:
                    self.data = None
            else:
                self.data = None
        if self.data is None:
2528
2529
            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
2530
2531
            err_msg += ("Set free_raw_data=False when construct Dataset to avoid this"
                        if was_none else "Freeing raw data")
2532
            _log_warning(err_msg)
Guolin Ke's avatar
Guolin Ke committed
2533
        self.feature_name = self.get_feature_name()
2534
2535
        _log_warning("Reseting categorical features.\n"
                     "You can set new categorical features via ``set_categorical_feature`` method")
Guolin Ke's avatar
Guolin Ke committed
2536
2537
        self.categorical_feature = "auto"
        self.pandas_categorical = None
2538
2539
        return self

2540
    def _dump_text(self, filename):
2541
2542
2543
2544
2545
2546
        """Save Dataset to a text file.

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

        Parameters
        ----------
2547
        filename : str or pathlib.Path
2548
2549
2550
2551
2552
2553
2554
2555
2556
            Name of the output file.

        Returns
        -------
        self : Dataset
            Returns self.
        """
        _safe_call(_LIB.LGBM_DatasetDumpText(
            self.construct().handle,
2557
            c_str(str(filename))))
2558
2559
        return self

wxchan's avatar
wxchan committed
2560

2561
class Booster:
2562
    """Booster in LightGBM."""
2563

2564
    def __init__(self, params=None, train_set=None, model_file=None, model_str=None, silent='warn'):
2565
        """Initialize the Booster.
wxchan's avatar
wxchan committed
2566
2567
2568

        Parameters
        ----------
Nikita Titov's avatar
Nikita Titov committed
2569
        params : dict or None, optional (default=None)
2570
2571
2572
            Parameters for Booster.
        train_set : Dataset or None, optional (default=None)
            Training dataset.
2573
        model_file : str, pathlib.Path or None, optional (default=None)
wxchan's avatar
wxchan committed
2574
            Path to the model file.
2575
        model_str : str or None, optional (default=None)
2576
            Model will be loaded from this string.
2577
2578
        silent : bool, optional (default=False)
            Whether to print messages during construction.
wxchan's avatar
wxchan committed
2579
        """
2580
        self.handle = None
2581
        self.network = False
wxchan's avatar
wxchan committed
2582
        self.__need_reload_eval_info = True
2583
        self._train_data_name = "training"
wxchan's avatar
wxchan committed
2584
        self.__attr = {}
2585
        self.__set_objective_to_none = False
wxchan's avatar
wxchan committed
2586
        self.best_iteration = -1
wxchan's avatar
wxchan committed
2587
        self.best_score = {}
2588
        params = {} if params is None else deepcopy(params)
2589
        # user can set verbose with params, it has higher priority
2590
2591
2592
2593
2594
        if silent != 'warn':
            _log_warning("'silent' argument is deprecated and will be removed in a future release of LightGBM. "
                         "Pass 'verbose' parameter via 'params' instead.")
        else:
            silent = False
2595
        if not any(verbose_alias in params for verbose_alias in _ConfigAliases.get("verbosity")) and silent:
2596
            params["verbose"] = -1
wxchan's avatar
wxchan committed
2597
        if train_set is not None:
2598
            # Training task
wxchan's avatar
wxchan committed
2599
            if not isinstance(train_set, Dataset):
2600
                raise TypeError(f'Training data should be Dataset instance, met {type(train_set).__name__}')
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
            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"]
                )
2635
            # construct booster object
2636
2637
2638
2639
            train_set.construct()
            # copy the parameters from train_set
            params.update(train_set.get_params())
            params_str = param_dict_to_str(params)
2640
            self.handle = ctypes.c_void_p()
wxchan's avatar
wxchan committed
2641
            _safe_call(_LIB.LGBM_BoosterCreate(
2642
                train_set.handle,
wxchan's avatar
wxchan committed
2643
2644
                c_str(params_str),
                ctypes.byref(self.handle)))
2645
            # save reference to data
wxchan's avatar
wxchan committed
2646
2647
2648
2649
            self.train_set = train_set
            self.valid_sets = []
            self.name_valid_sets = []
            self.__num_dataset = 1
Guolin Ke's avatar
Guolin Ke committed
2650
2651
            self.__init_predictor = train_set._predictor
            if self.__init_predictor is not None:
wxchan's avatar
wxchan committed
2652
2653
                _safe_call(_LIB.LGBM_BoosterMerge(
                    self.handle,
Guolin Ke's avatar
Guolin Ke committed
2654
                    self.__init_predictor.handle))
Guolin Ke's avatar
Guolin Ke committed
2655
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2656
2657
2658
2659
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.__num_class = out_num_class.value
2660
            # buffer for inner predict
wxchan's avatar
wxchan committed
2661
2662
2663
            self.__inner_predict_buffer = [None]
            self.__is_predicted_cur_iter = [False]
            self.__get_eval_info()
2664
            self.pandas_categorical = train_set.pandas_categorical
2665
            self.train_set_version = train_set.version
wxchan's avatar
wxchan committed
2666
        elif model_file is not None:
2667
            # Prediction task
Guolin Ke's avatar
Guolin Ke committed
2668
            out_num_iterations = ctypes.c_int(0)
2669
            self.handle = ctypes.c_void_p()
wxchan's avatar
wxchan committed
2670
            _safe_call(_LIB.LGBM_BoosterCreateFromModelfile(
2671
                c_str(str(model_file)),
wxchan's avatar
wxchan committed
2672
2673
                ctypes.byref(out_num_iterations),
                ctypes.byref(self.handle)))
Guolin Ke's avatar
Guolin Ke committed
2674
            out_num_class = ctypes.c_int(0)
wxchan's avatar
wxchan committed
2675
2676
2677
2678
            _safe_call(_LIB.LGBM_BoosterGetNumClasses(
                self.handle,
                ctypes.byref(out_num_class)))
            self.__num_class = out_num_class.value
2679
            self.pandas_categorical = _load_pandas_categorical(file_name=model_file)
2680
        elif model_str is not None:
2681
            self.model_from_string(model_str, verbose="_silent_false")
wxchan's avatar
wxchan committed
2682
        else:
2683
2684
            raise TypeError('Need at least one training dataset or model file or model string '
                            'to create Booster instance')
2685
        self.params = params
wxchan's avatar
wxchan committed
2686
2687

    def __del__(self):
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
        try:
            if self.network:
                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
2698

wxchan's avatar
wxchan committed
2699
2700
2701
2702
    def __copy__(self):
        return self.__deepcopy__(None)

    def __deepcopy__(self, _):
2703
        model_str = self.model_to_string(num_iteration=-1)
2704
        booster = Booster(model_str=model_str)
2705
        return booster
wxchan's avatar
wxchan committed
2706
2707
2708
2709
2710
2711
2712

    def __getstate__(self):
        this = self.__dict__.copy()
        handle = this['handle']
        this.pop('train_set', None)
        this.pop('valid_sets', None)
        if handle is not None:
2713
            this["handle"] = self.model_to_string(num_iteration=-1)
wxchan's avatar
wxchan committed
2714
2715
2716
        return this

    def __setstate__(self, state):
2717
2718
        model_str = state.get('handle', None)
        if model_str is not None:
wxchan's avatar
wxchan committed
2719
            handle = ctypes.c_void_p()
Guolin Ke's avatar
Guolin Ke committed
2720
            out_num_iterations = ctypes.c_int(0)
2721
2722
2723
2724
            _safe_call(_LIB.LGBM_BoosterLoadModelFromString(
                c_str(model_str),
                ctypes.byref(out_num_iterations),
                ctypes.byref(handle)))
wxchan's avatar
wxchan committed
2725
2726
2727
            state['handle'] = handle
        self.__dict__.update(state)

wxchan's avatar
wxchan committed
2728
    def free_dataset(self):
Nikita Titov's avatar
Nikita Titov committed
2729
2730
2731
2732
2733
2734
2735
        """Free Booster's Datasets.

        Returns
        -------
        self : Booster
            Booster without Datasets.
        """
wxchan's avatar
wxchan committed
2736
2737
        self.__dict__.pop('train_set', None)
        self.__dict__.pop('valid_sets', None)
2738
        self.__num_dataset = 0
Nikita Titov's avatar
Nikita Titov committed
2739
        return self
wxchan's avatar
wxchan committed
2740

2741
2742
2743
    def _free_buffer(self):
        self.__inner_predict_buffer = []
        self.__is_predicted_cur_iter = []
Nikita Titov's avatar
Nikita Titov committed
2744
        return self
2745

2746
2747
2748
2749
2750
2751
2752
    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":
2753
2754
2755
2756
        """Set the network configuration.

        Parameters
        ----------
2757
        machines : list, set or str
2758
            Names of machines.
Nikita Titov's avatar
Nikita Titov committed
2759
        local_listen_port : int, optional (default=12400)
2760
            TCP listen port for local machines.
Nikita Titov's avatar
Nikita Titov committed
2761
        listen_time_out : int, optional (default=120)
2762
            Socket time-out in minutes.
Nikita Titov's avatar
Nikita Titov committed
2763
        num_machines : int, optional (default=1)
2764
            The number of machines for distributed learning application.
Nikita Titov's avatar
Nikita Titov committed
2765
2766
2767
2768
2769

        Returns
        -------
        self : Booster
            Booster with set network.
2770
        """
2771
2772
        if isinstance(machines, (list, set)):
            machines = ','.join(machines)
2773
2774
2775
2776
2777
        _safe_call(_LIB.LGBM_NetworkInit(c_str(machines),
                                         ctypes.c_int(local_listen_port),
                                         ctypes.c_int(listen_time_out),
                                         ctypes.c_int(num_machines)))
        self.network = True
Nikita Titov's avatar
Nikita Titov committed
2778
        return self
2779
2780

    def free_network(self):
Nikita Titov's avatar
Nikita Titov committed
2781
2782
2783
2784
2785
2786
2787
        """Free Booster's network.

        Returns
        -------
        self : Booster
            Booster with freed network.
        """
2788
2789
        _safe_call(_LIB.LGBM_NetworkFree())
        self.network = False
Nikita Titov's avatar
Nikita Titov committed
2790
        return self
2791

2792
2793
2794
    def trees_to_dataframe(self):
        """Parse the fitted model and return in an easy-to-read pandas DataFrame.

2795
2796
2797
2798
        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.
2799
2800
2801
2802
2803
            - ``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.
2804
2805
            - ``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.
2806
            - ``decision_type`` : str, logical operator describing how to compare a value to ``threshold``.
2807
2808
              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.
2809
2810
            - ``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.
2811
2812
2813
2814
            - ``value`` : float64, predicted value for this leaf node, multiplied by the learning rate.
            - ``weight`` : float64 or int64, sum of hessian (second-order derivative of objective), summed over observations that fall in this node.
            - ``count`` : int64, number of records in the training data that fall into this node.

2815
2816
2817
2818
2819
2820
        Returns
        -------
        result : pandas DataFrame
            Returns a pandas DataFrame of the parsed model.
        """
        if not PANDAS_INSTALLED:
2821
2822
            raise LightGBMError('This method cannot be run without pandas installed. '
                                'You must install pandas and restart your session to use this method.')
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833

        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):
2834
                tree_num = f'{tree_index}-' if tree_index is not None else ''
2835
2836
2837
                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
2838
2839
                node_num = tree.get('split_index' if is_split else 'leaf_index', 0)
                return f"{tree_num}{node_type}{node_num}"
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851

            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):
2852
                return set(tree.keys()) == {'leaf_value'}
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925

            # 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))

2926
        return pd_DataFrame(model_list, columns=model_list[0].keys())
2927

wxchan's avatar
wxchan committed
2928
    def set_train_data_name(self, name):
2929
2930
2931
2932
        """Set the name to the training Dataset.

        Parameters
        ----------
2933
        name : str
Nikita Titov's avatar
Nikita Titov committed
2934
2935
2936
2937
2938
2939
            Name for the training Dataset.

        Returns
        -------
        self : Booster
            Booster with set training Dataset name.
2940
        """
2941
        self._train_data_name = name
Nikita Titov's avatar
Nikita Titov committed
2942
        return self
wxchan's avatar
wxchan committed
2943
2944

    def add_valid(self, data, name):
2945
        """Add validation data.
wxchan's avatar
wxchan committed
2946
2947
2948
2949

        Parameters
        ----------
        data : Dataset
2950
            Validation data.
2951
        name : str
2952
            Name of validation data.
Nikita Titov's avatar
Nikita Titov committed
2953
2954
2955
2956
2957

        Returns
        -------
        self : Booster
            Booster with set validation data.
wxchan's avatar
wxchan committed
2958
        """
Guolin Ke's avatar
Guolin Ke committed
2959
        if not isinstance(data, Dataset):
2960
            raise TypeError(f'Validation data should be Dataset instance, met {type(data).__name__}')
Guolin Ke's avatar
Guolin Ke committed
2961
        if data._predictor is not self.__init_predictor:
2962
2963
            raise LightGBMError("Add validation data failed, "
                                "you should use same predictor for these data")
wxchan's avatar
wxchan committed
2964
2965
        _safe_call(_LIB.LGBM_BoosterAddValidData(
            self.handle,
wxchan's avatar
wxchan committed
2966
            data.construct().handle))
wxchan's avatar
wxchan committed
2967
2968
2969
2970
2971
        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
2972
        return self
wxchan's avatar
wxchan committed
2973
2974

    def reset_parameter(self, params):
2975
        """Reset parameters of Booster.
wxchan's avatar
wxchan committed
2976
2977
2978
2979

        Parameters
        ----------
        params : dict
2980
            New parameters for Booster.
Nikita Titov's avatar
Nikita Titov committed
2981
2982
2983
2984
2985

        Returns
        -------
        self : Booster
            Booster with new parameters.
wxchan's avatar
wxchan committed
2986
2987
2988
2989
2990
2991
        """
        params_str = param_dict_to_str(params)
        if params_str:
            _safe_call(_LIB.LGBM_BoosterResetParameter(
                self.handle,
                c_str(params_str)))
Guolin Ke's avatar
Guolin Ke committed
2992
        self.params.update(params)
Nikita Titov's avatar
Nikita Titov committed
2993
        return self
wxchan's avatar
wxchan committed
2994
2995

    def update(self, train_set=None, fobj=None):
Nikita Titov's avatar
Nikita Titov committed
2996
        """Update Booster for one iteration.
2997

wxchan's avatar
wxchan committed
2998
2999
        Parameters
        ----------
3000
3001
3002
3003
        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
3004
            Customized objective function.
3005
3006
3007
3008
3009
            Should accept two parameters: preds, train_data,
            and return (grad, hess).

                preds : list or numpy 1-D array
                    The predicted values.
3010
3011
                    Predicted values are returned before any transformation,
                    e.g. they are raw margin instead of probability of positive class for binary task.
3012
3013
3014
                train_data : Dataset
                    The training dataset.
                grad : list or numpy 1-D array
3015
3016
                    The value of the first order derivative (gradient) of the loss
                    with respect to the elements of preds for each sample point.
3017
                hess : list or numpy 1-D array
3018
3019
                    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
3020

3021
3022
            For multi-class task, the preds is group by class_id first, then group by row_id.
            If you want to get i-th row preds in j-th class, the access way is score[j * num_data + i]
3023
3024
            and you should group grad and hess in this way as well.

wxchan's avatar
wxchan committed
3025
3026
        Returns
        -------
3027
3028
        is_finished : bool
            Whether the update was successfully finished.
wxchan's avatar
wxchan committed
3029
        """
3030
        # need reset training data
3031
3032
3033
3034
3035
3036
        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
3037
            if not isinstance(train_set, Dataset):
3038
                raise TypeError(f'Training data should be Dataset instance, met {type(train_set).__name__}')
Guolin Ke's avatar
Guolin Ke committed
3039
            if train_set._predictor is not self.__init_predictor:
3040
3041
                raise LightGBMError("Replace training data failed, "
                                    "you should use same predictor for these data")
wxchan's avatar
wxchan committed
3042
3043
3044
            self.train_set = train_set
            _safe_call(_LIB.LGBM_BoosterResetTrainingData(
                self.handle,
wxchan's avatar
wxchan committed
3045
                self.train_set.construct().handle))
wxchan's avatar
wxchan committed
3046
            self.__inner_predict_buffer[0] = None
3047
            self.train_set_version = self.train_set.version
wxchan's avatar
wxchan committed
3048
3049
        is_finished = ctypes.c_int(0)
        if fobj is None:
3050
            if self.__set_objective_to_none:
3051
                raise LightGBMError('Cannot update due to null objective function.')
wxchan's avatar
wxchan committed
3052
3053
3054
            _safe_call(_LIB.LGBM_BoosterUpdateOneIter(
                self.handle,
                ctypes.byref(is_finished)))
3055
            self.__is_predicted_cur_iter = [False for _ in range(self.__num_dataset)]
wxchan's avatar
wxchan committed
3056
3057
            return is_finished.value == 1
        else:
3058
            if not self.__set_objective_to_none:
Nikita Titov's avatar
Nikita Titov committed
3059
                self.reset_parameter({"objective": "none"}).__set_objective_to_none = True
wxchan's avatar
wxchan committed
3060
3061
3062
3063
            grad, hess = fobj(self.__inner_predict(0), self.train_set)
            return self.__boost(grad, hess)

    def __boost(self, grad, hess):
3064
        """Boost Booster for one iteration with customized gradient statistics.
Nikita Titov's avatar
Nikita Titov committed
3065

Nikita Titov's avatar
Nikita Titov committed
3066
3067
        .. note::

3068
3069
            Score is returned before any transformation,
            e.g. it is raw margin instead of probability of positive class for binary task.
Nikita Titov's avatar
Nikita Titov committed
3070
3071
3072
            For multi-class task, the score is group by class_id first, then group by row_id.
            If you want to get i-th row score in j-th class, the access way is score[j * num_data + i]
            and you should group grad and hess in this way as well.
3073

wxchan's avatar
wxchan committed
3074
3075
        Parameters
        ----------
3076
        grad : list or numpy 1-D array
3077
3078
            The value of the first order derivative (gradient) of the loss
            with respect to the elements of score for each sample point.
3079
        hess : list or numpy 1-D array
3080
3081
            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
3082
3083
3084

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3085
3086
        is_finished : bool
            Whether the boost was successfully finished.
wxchan's avatar
wxchan committed
3087
        """
3088
3089
        grad = list_to_1d_numpy(grad, name='gradient')
        hess = list_to_1d_numpy(hess, name='hessian')
3090
3091
        assert grad.flags.c_contiguous
        assert hess.flags.c_contiguous
wxchan's avatar
wxchan committed
3092
        if len(grad) != len(hess):
3093
            raise ValueError(f"Lengths of gradient({len(grad)}) and hessian({len(hess)}) don't match")
wxchan's avatar
wxchan committed
3094
3095
3096
3097
3098
3099
        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)))
3100
        self.__is_predicted_cur_iter = [False for _ in range(self.__num_dataset)]
wxchan's avatar
wxchan committed
3101
3102
3103
        return is_finished.value == 1

    def rollback_one_iter(self):
Nikita Titov's avatar
Nikita Titov committed
3104
3105
3106
3107
3108
3109
3110
        """Rollback one iteration.

        Returns
        -------
        self : Booster
            Booster with rolled back one iteration.
        """
wxchan's avatar
wxchan committed
3111
3112
        _safe_call(_LIB.LGBM_BoosterRollbackOneIter(
            self.handle))
3113
        self.__is_predicted_cur_iter = [False for _ in range(self.__num_dataset)]
Nikita Titov's avatar
Nikita Titov committed
3114
        return self
wxchan's avatar
wxchan committed
3115
3116

    def current_iteration(self):
3117
3118
3119
3120
3121
3122
3123
        """Get the index of the current iteration.

        Returns
        -------
        cur_iter : int
            The index of the current iteration.
        """
Guolin Ke's avatar
Guolin Ke committed
3124
        out_cur_iter = ctypes.c_int(0)
wxchan's avatar
wxchan committed
3125
3126
3127
3128
3129
        _safe_call(_LIB.LGBM_BoosterGetCurrentIteration(
            self.handle,
            ctypes.byref(out_cur_iter)))
        return out_cur_iter.value

3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
    def num_model_per_iteration(self):
        """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

    def num_trees(self):
        """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

3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
    def upper_bound(self):
        """Get upper bound value of a model.

        Returns
        -------
        upper_bound : double
            Upper bound value of the model.
        """
        ret = ctypes.c_double(0)
        _safe_call(_LIB.LGBM_BoosterGetUpperBoundValue(
            self.handle,
            ctypes.byref(ret)))
        return ret.value

    def lower_bound(self):
        """Get lower bound value of a model.

        Returns
        -------
        lower_bound : double
            Lower bound value of the model.
        """
        ret = ctypes.c_double(0)
        _safe_call(_LIB.LGBM_BoosterGetLowerBoundValue(
            self.handle,
            ctypes.byref(ret)))
        return ret.value

wxchan's avatar
wxchan committed
3186
    def eval(self, data, name, feval=None):
3187
        """Evaluate for data.
wxchan's avatar
wxchan committed
3188
3189
3190

        Parameters
        ----------
3191
3192
        data : Dataset
            Data for the evaluating.
3193
        name : str
3194
3195
            Name of the data.
        feval : callable or None, optional (default=None)
3196
            Customized evaluation function.
3197
            Should accept two parameters: preds, eval_data,
3198
            and return (eval_name, eval_result, is_higher_better) or list of such tuples.
3199
3200
3201

                preds : list or numpy 1-D array
                    The predicted values.
3202
3203
                    If ``fobj`` is specified, predicted values are returned before any transformation,
                    e.g. they are raw margin instead of probability of positive class for binary task in this case.
3204
3205
                eval_data : Dataset
                    The evaluation dataset.
3206
                eval_name : str
Andrew Ziem's avatar
Andrew Ziem committed
3207
                    The name of evaluation function (without whitespace).
3208
3209
3210
3211
3212
                eval_result : float
                    The eval result.
                is_higher_better : bool
                    Is eval result higher better, e.g. AUC is ``is_higher_better``.

3213
3214
            For multi-class task, the preds is group by class_id first, then group by row_id.
            If you want to get i-th row preds in j-th class, the access way is preds[j * num_data + i].
3215

wxchan's avatar
wxchan committed
3216
3217
        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3218
        result : list
3219
            List with evaluation results.
wxchan's avatar
wxchan committed
3220
        """
Guolin Ke's avatar
Guolin Ke committed
3221
3222
        if not isinstance(data, Dataset):
            raise TypeError("Can only eval for Dataset instance")
wxchan's avatar
wxchan committed
3223
3224
3225
3226
        data_idx = -1
        if data is self.train_set:
            data_idx = 0
        else:
3227
            for i in range(len(self.valid_sets)):
wxchan's avatar
wxchan committed
3228
3229
3230
                if data is self.valid_sets[i]:
                    data_idx = i + 1
                    break
3231
        # need to push new valid data
wxchan's avatar
wxchan committed
3232
3233
3234
3235
3236
3237
3238
        if data_idx == -1:
            self.add_valid(data, name)
            data_idx = self.__num_dataset - 1

        return self.__inner_eval(name, data_idx, feval)

    def eval_train(self, feval=None):
3239
        """Evaluate for training data.
wxchan's avatar
wxchan committed
3240
3241
3242

        Parameters
        ----------
3243
        feval : callable or None, optional (default=None)
3244
            Customized evaluation function.
3245
3246
            Should accept two parameters: preds, train_data,
            and return (eval_name, eval_result, is_higher_better) or list of such tuples.
3247
3248
3249

                preds : list or numpy 1-D array
                    The predicted values.
3250
3251
                    If ``fobj`` is specified, predicted values are returned before any transformation,
                    e.g. they are raw margin instead of probability of positive class for binary task in this case.
3252
3253
                train_data : Dataset
                    The training dataset.
3254
                eval_name : str
Andrew Ziem's avatar
Andrew Ziem committed
3255
                    The name of evaluation function (without whitespace).
3256
3257
3258
3259
3260
                eval_result : float
                    The eval result.
                is_higher_better : bool
                    Is eval result higher better, e.g. AUC is ``is_higher_better``.

3261
3262
            For multi-class task, the preds is group by class_id first, then group by row_id.
            If you want to get i-th row preds in j-th class, the access way is preds[j * num_data + i].
wxchan's avatar
wxchan committed
3263
3264
3265

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3266
        result : list
3267
            List with evaluation results.
wxchan's avatar
wxchan committed
3268
        """
3269
        return self.__inner_eval(self._train_data_name, 0, feval)
wxchan's avatar
wxchan committed
3270
3271

    def eval_valid(self, feval=None):
3272
        """Evaluate for validation data.
wxchan's avatar
wxchan committed
3273
3274
3275

        Parameters
        ----------
3276
        feval : callable or None, optional (default=None)
3277
            Customized evaluation function.
3278
            Should accept two parameters: preds, valid_data,
3279
            and return (eval_name, eval_result, is_higher_better) or list of such tuples.
3280
3281
3282

                preds : list or numpy 1-D array
                    The predicted values.
3283
3284
                    If ``fobj`` is specified, predicted values are returned before any transformation,
                    e.g. they are raw margin instead of probability of positive class for binary task in this case.
3285
3286
                valid_data : Dataset
                    The validation dataset.
3287
                eval_name : str
Andrew Ziem's avatar
Andrew Ziem committed
3288
                    The name of evaluation function (without whitespace).
3289
3290
3291
3292
3293
                eval_result : float
                    The eval result.
                is_higher_better : bool
                    Is eval result higher better, e.g. AUC is ``is_higher_better``.

3294
3295
            For multi-class task, the preds is group by class_id first, then group by row_id.
            If you want to get i-th row preds in j-th class, the access way is preds[j * num_data + i].
wxchan's avatar
wxchan committed
3296
3297
3298

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3299
        result : list
3300
            List with evaluation results.
wxchan's avatar
wxchan committed
3301
        """
3302
        return [item for i in range(1, self.__num_dataset)
wxchan's avatar
wxchan committed
3303
                for item in self.__inner_eval(self.name_valid_sets[i - 1], i, feval)]
wxchan's avatar
wxchan committed
3304

3305
    def save_model(self, filename, num_iteration=None, start_iteration=0, importance_type='split'):
3306
        """Save Booster to file.
wxchan's avatar
wxchan committed
3307
3308
3309

        Parameters
        ----------
3310
        filename : str or pathlib.Path
3311
            Filename to save Booster.
3312
3313
3314
3315
        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
3316
        start_iteration : int, optional (default=0)
3317
            Start index of the iteration that should be saved.
3318
        importance_type : str, optional (default="split")
3319
3320
3321
            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
3322
3323
3324
3325
3326

        Returns
        -------
        self : Booster
            Returns self.
wxchan's avatar
wxchan committed
3327
        """
3328
        if num_iteration is None:
3329
            num_iteration = self.best_iteration
3330
        importance_type_int = FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
wxchan's avatar
wxchan committed
3331
3332
        _safe_call(_LIB.LGBM_BoosterSaveModel(
            self.handle,
3333
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
3334
            ctypes.c_int(num_iteration),
3335
            ctypes.c_int(importance_type_int),
3336
            c_str(str(filename))))
3337
        _dump_pandas_categorical(self.pandas_categorical, filename)
Nikita Titov's avatar
Nikita Titov committed
3338
        return self
wxchan's avatar
wxchan committed
3339

3340
    def shuffle_models(self, start_iteration=0, end_iteration=-1):
3341
        """Shuffle models.
Nikita Titov's avatar
Nikita Titov committed
3342

3343
3344
3345
        Parameters
        ----------
        start_iteration : int, optional (default=0)
3346
            The first iteration that will be shuffled.
3347
3348
        end_iteration : int, optional (default=-1)
            The last iteration that will be shuffled.
3349
            If <= 0, means the last available iteration.
3350

Nikita Titov's avatar
Nikita Titov committed
3351
3352
3353
3354
        Returns
        -------
        self : Booster
            Booster with shuffled models.
3355
        """
3356
3357
        _safe_call(_LIB.LGBM_BoosterShuffleModels(
            self.handle,
Guolin Ke's avatar
Guolin Ke committed
3358
3359
            ctypes.c_int(start_iteration),
            ctypes.c_int(end_iteration)))
Nikita Titov's avatar
Nikita Titov committed
3360
        return self
3361

3362
    def model_from_string(self, model_str, verbose='warn'):
3363
3364
3365
3366
        """Load Booster from a string.

        Parameters
        ----------
3367
        model_str : str
3368
            Model will be loaded from this string.
Nikita Titov's avatar
Nikita Titov committed
3369
3370
        verbose : bool, optional (default=True)
            Whether to print messages while loading model.
3371
3372
3373

        Returns
        -------
Nikita Titov's avatar
Nikita Titov committed
3374
        self : Booster
3375
3376
            Loaded Booster object.
        """
3377
3378
3379
3380
        if self.handle is not None:
            _safe_call(_LIB.LGBM_BoosterFree(self.handle))
        self._free_buffer()
        self.handle = ctypes.c_void_p()
3381
3382
3383
3384
3385
3386
3387
3388
3389
        out_num_iterations = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterLoadModelFromString(
            c_str(model_str),
            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)))
3390
3391
3392
3393
        if verbose in {'warn', '_silent_false'}:
            verbose = verbose == 'warn'
        else:
            _log_warning("'verbose' argument is deprecated and will be removed in a future release of LightGBM.")
3394
        if verbose:
3395
            _log_info(f'Finished loading model, total used {int(out_num_iterations.value)} iterations')
3396
        self.__num_class = out_num_class.value
3397
        self.pandas_categorical = _load_pandas_categorical(model_str=model_str)
3398
3399
        return self

3400
    def model_to_string(self, num_iteration=None, start_iteration=0, importance_type='split'):
3401
        """Save Booster to string.
3402

3403
3404
3405
3406
3407
3408
        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
3409
        start_iteration : int, optional (default=0)
3410
            Start index of the iteration that should be saved.
3411
        importance_type : str, optional (default="split")
3412
3413
3414
            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.
3415
3416
3417

        Returns
        -------
3418
        str_repr : str
3419
3420
            String representation of Booster.
        """
3421
        if num_iteration is None:
3422
            num_iteration = self.best_iteration
3423
        importance_type_int = FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
3424
        buffer_len = 1 << 20
3425
        tmp_out_len = ctypes.c_int64(0)
3426
3427
3428
3429
        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,
3430
            ctypes.c_int(start_iteration),
3431
            ctypes.c_int(num_iteration),
3432
            ctypes.c_int(importance_type_int),
3433
            ctypes.c_int64(buffer_len),
3434
3435
3436
            ctypes.byref(tmp_out_len),
            ptr_string_buffer))
        actual_len = tmp_out_len.value
3437
        # if buffer length is not long enough, re-allocate a buffer
3438
3439
3440
3441
3442
        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,
3443
                ctypes.c_int(start_iteration),
3444
                ctypes.c_int(num_iteration),
3445
                ctypes.c_int(importance_type_int),
3446
                ctypes.c_int64(actual_len),
3447
3448
                ctypes.byref(tmp_out_len),
                ptr_string_buffer))
3449
        ret = string_buffer.value.decode('utf-8')
3450
3451
        ret += _dump_pandas_categorical(self.pandas_categorical)
        return ret
3452

3453
    def dump_model(self, num_iteration=None, start_iteration=0, importance_type='split', object_hook=None):
Nikita Titov's avatar
Nikita Titov committed
3454
        """Dump Booster to JSON format.
wxchan's avatar
wxchan committed
3455

3456
3457
        Parameters
        ----------
3458
3459
3460
3461
        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
3462
        start_iteration : int, optional (default=0)
3463
            Start index of the iteration that should be dumped.
3464
        importance_type : str, optional (default="split")
3465
3466
3467
            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.
3468
3469
3470
3471
3472
3473
3474
3475
3476
        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.
3477

wxchan's avatar
wxchan committed
3478
3479
        Returns
        -------
3480
        json_repr : dict
Nikita Titov's avatar
Nikita Titov committed
3481
            JSON format of Booster.
wxchan's avatar
wxchan committed
3482
        """
3483
        if num_iteration is None:
3484
            num_iteration = self.best_iteration
3485
        importance_type_int = FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
wxchan's avatar
wxchan committed
3486
        buffer_len = 1 << 20
3487
        tmp_out_len = ctypes.c_int64(0)
wxchan's avatar
wxchan committed
3488
3489
3490
3491
        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,
3492
            ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
3493
            ctypes.c_int(num_iteration),
3494
            ctypes.c_int(importance_type_int),
3495
            ctypes.c_int64(buffer_len),
wxchan's avatar
wxchan committed
3496
            ctypes.byref(tmp_out_len),
Guolin Ke's avatar
Guolin Ke committed
3497
            ptr_string_buffer))
wxchan's avatar
wxchan committed
3498
        actual_len = tmp_out_len.value
3499
        # if buffer length is not long enough, reallocate a buffer
wxchan's avatar
wxchan committed
3500
3501
3502
3503
3504
        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,
3505
                ctypes.c_int(start_iteration),
Guolin Ke's avatar
Guolin Ke committed
3506
                ctypes.c_int(num_iteration),
3507
                ctypes.c_int(importance_type_int),
3508
                ctypes.c_int64(actual_len),
wxchan's avatar
wxchan committed
3509
                ctypes.byref(tmp_out_len),
Guolin Ke's avatar
Guolin Ke committed
3510
                ptr_string_buffer))
3511
        ret = json.loads(string_buffer.value.decode('utf-8'), object_hook=object_hook)
3512
3513
3514
        ret['pandas_categorical'] = json.loads(json.dumps(self.pandas_categorical,
                                                          default=json_default_with_numpy))
        return ret
wxchan's avatar
wxchan committed
3515

3516
    def predict(self, data, start_iteration=0, num_iteration=None,
3517
                raw_score=False, pred_leaf=False, pred_contrib=False,
3518
                data_has_header=False, is_reshape=True, **kwargs):
3519
        """Make a prediction.
wxchan's avatar
wxchan committed
3520
3521
3522

        Parameters
        ----------
3523
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame or scipy.sparse
3524
            Data source for prediction.
3525
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM).
3526
        start_iteration : int, optional (default=0)
3527
            Start index of the iteration to predict.
3528
            If <= 0, starts from the first iteration.
3529
        num_iteration : int or None, optional (default=None)
3530
3531
3532
3533
            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).
3534
3535
3536
3537
        raw_score : bool, optional (default=False)
            Whether to predict raw scores.
        pred_leaf : bool, optional (default=False)
            Whether to predict leaf index.
3538
3539
        pred_contrib : bool, optional (default=False)
            Whether to predict feature contributions.
3540

Nikita Titov's avatar
Nikita Titov committed
3541
3542
3543
3544
3545
3546
3547
            .. 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.
3548

3549
3550
        data_has_header : bool, optional (default=False)
            Whether the data has header.
3551
            Used only if data is str.
3552
3553
        is_reshape : bool, optional (default=True)
            If True, result is reshaped to [nrow, ncol].
3554
3555
        **kwargs
            Other parameters for the prediction.
wxchan's avatar
wxchan committed
3556
3557
3558

        Returns
        -------
3559
        result : numpy array, scipy.sparse or list of scipy.sparse
3560
            Prediction result.
3561
            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
3562
        """
3563
        predictor = self._to_predictor(deepcopy(kwargs))
3564
        if num_iteration is None:
3565
            if start_iteration <= 0:
3566
3567
3568
3569
                num_iteration = self.best_iteration
            else:
                num_iteration = -1
        return predictor.predict(data, start_iteration, num_iteration,
3570
3571
                                 raw_score, pred_leaf, pred_contrib,
                                 data_has_header, is_reshape)
wxchan's avatar
wxchan committed
3572

3573
    def refit(self, data, label, decay_rate=0.9, **kwargs):
Guolin Ke's avatar
Guolin Ke committed
3574
3575
3576
3577
        """Refit the existing Booster by new data.

        Parameters
        ----------
3578
        data : str, pathlib.Path, numpy array, pandas DataFrame, H2O DataTable's Frame or scipy.sparse
Guolin Ke's avatar
Guolin Ke committed
3579
            Data source for refit.
3580
            If str or pathlib.Path, it represents the path to a text file (CSV, TSV, or LibSVM).
3581
        label : list, numpy 1-D array or pandas Series / one-column DataFrame
Guolin Ke's avatar
Guolin Ke committed
3582
3583
            Label for refit.
        decay_rate : float, optional (default=0.9)
3584
3585
            Decay rate of refit,
            will use ``leaf_output = decay_rate * old_leaf_output + (1.0 - decay_rate) * new_leaf_output`` to refit trees.
3586
3587
        **kwargs
            Other parameters for refit.
3588
            These parameters will be passed to ``predict`` method.
Guolin Ke's avatar
Guolin Ke committed
3589
3590
3591
3592
3593
3594

        Returns
        -------
        result : Booster
            Refitted Booster.
        """
3595
3596
        if self.__set_objective_to_none:
            raise LightGBMError('Cannot refit due to null objective function.')
3597
        predictor = self._to_predictor(deepcopy(kwargs))
3598
        leaf_preds = predictor.predict(data, -1, pred_leaf=True)
3599
        nrow, ncol = leaf_preds.shape
3600
3601
3602
3603
        out_is_linear = ctypes.c_bool(False)
        _safe_call(_LIB.LGBM_BoosterGetLinear(
            self.handle,
            ctypes.byref(out_is_linear)))
Nikita Titov's avatar
Nikita Titov committed
3604
3605
3606
3607
3608
        new_params = _choose_param_value(
            main_param_name="linear_tree",
            params=self.params,
            default_value=None
        )
3609
3610
        new_params["linear_tree"] = out_is_linear.value
        train_set = Dataset(data, label, silent=True, params=new_params)
3611
        new_params['refit_decay_rate'] = decay_rate
3612
        new_booster = Booster(new_params, train_set)
Guolin Ke's avatar
Guolin Ke committed
3613
3614
3615
3616
3617
        # Copy models
        _safe_call(_LIB.LGBM_BoosterMerge(
            new_booster.handle,
            predictor.handle))
        leaf_preds = leaf_preds.reshape(-1)
3618
        ptr_data, _, _ = c_int_array(leaf_preds)
Guolin Ke's avatar
Guolin Ke committed
3619
3620
3621
        _safe_call(_LIB.LGBM_BoosterRefit(
            new_booster.handle,
            ptr_data,
3622
3623
            ctypes.c_int32(nrow),
            ctypes.c_int32(ncol)))
3624
3625
        new_booster.network = self.network
        new_booster.__attr = self.__attr.copy()
Guolin Ke's avatar
Guolin Ke committed
3626
3627
        return new_booster

3628
    def get_leaf_output(self, tree_id, leaf_id):
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
        """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.
        """
3643
3644
3645
3646
3647
3648
3649
3650
        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

3651
    def _to_predictor(self, pred_parameter=None):
3652
        """Convert to predictor."""
3653
        predictor = _InnerPredictor(booster_handle=self.handle, pred_parameter=pred_parameter)
3654
        predictor.pandas_categorical = self.pandas_categorical
wxchan's avatar
wxchan committed
3655
3656
        return predictor

3657
    def num_feature(self):
3658
3659
3660
3661
3662
3663
3664
        """Get number of features.

        Returns
        -------
        num_feature : int
            The number of features.
        """
3665
3666
3667
3668
3669
3670
        out_num_feature = ctypes.c_int(0)
        _safe_call(_LIB.LGBM_BoosterGetNumFeature(
            self.handle,
            ctypes.byref(out_num_feature)))
        return out_num_feature.value

wxchan's avatar
wxchan committed
3671
    def feature_name(self):
3672
        """Get names of features.
wxchan's avatar
wxchan committed
3673
3674
3675

        Returns
        -------
3676
3677
        result : list
            List with names of features.
wxchan's avatar
wxchan committed
3678
        """
3679
        num_feature = self.num_feature()
3680
        # Get name of features
wxchan's avatar
wxchan committed
3681
        tmp_out_len = ctypes.c_int(0)
3682
3683
        reserved_string_buffer_size = 255
        required_string_buffer_size = ctypes.c_size_t(0)
3684
        string_buffers = [ctypes.create_string_buffer(reserved_string_buffer_size) for _ in range(num_feature)]
wxchan's avatar
wxchan committed
3685
3686
3687
        ptr_string_buffers = (ctypes.c_char_p * num_feature)(*map(ctypes.addressof, string_buffers))
        _safe_call(_LIB.LGBM_BoosterGetFeatureNames(
            self.handle,
3688
            ctypes.c_int(num_feature),
wxchan's avatar
wxchan committed
3689
            ctypes.byref(tmp_out_len),
3690
            ctypes.c_size_t(reserved_string_buffer_size),
3691
            ctypes.byref(required_string_buffer_size),
wxchan's avatar
wxchan committed
3692
3693
3694
            ptr_string_buffers))
        if num_feature != tmp_out_len.value:
            raise ValueError("Length of feature names doesn't equal with num_feature")
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
        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))
3707
        return [string_buffers[i].value.decode('utf-8') for i in range(num_feature)]
wxchan's avatar
wxchan committed
3708

3709
    def feature_importance(self, importance_type='split', iteration=None):
3710
        """Get feature importances.
3711

3712
3713
        Parameters
        ----------
3714
        importance_type : str, optional (default="split")
3715
3716
3717
            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.
3718
3719
3720
3721
        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).
3722

3723
3724
        Returns
        -------
3725
3726
        result : numpy array
            Array with feature importances.
3727
        """
3728
3729
        if iteration is None:
            iteration = self.best_iteration
3730
        importance_type_int = FEATURE_IMPORTANCE_TYPE_MAPPER[importance_type]
3731
        result = np.empty(self.num_feature(), dtype=np.float64)
3732
3733
3734
3735
3736
3737
        _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))))
        if importance_type_int == 0:
3738
            return result.astype(np.int32)
3739
3740
        else:
            return result
3741

3742
3743
3744
3745
3746
    def get_split_value_histogram(self, feature, bins=None, xgboost_style=False):
        """Get split value histogram for the specified feature.

        Parameters
        ----------
3747
        feature : int or str
3748
3749
            The feature name or index the histogram is calculated for.
            If int, interpreted as index.
3750
            If str, interpreted as name.
3751

Nikita Titov's avatar
Nikita Titov committed
3752
3753
3754
            .. warning::

                Categorical features are not supported.
3755

3756
        bins : int, str or None, optional (default=None)
3757
3758
3759
            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.
3760
            If str, it should be one from the list of the supported values by ``numpy.histogram()`` function.
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
        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.
        """
        def add(root):
            """Recursively add thresholds."""
            if 'split_index' in root:  # non-leaf
3778
                if feature_names is not None and isinstance(feature, str):
3779
3780
3781
3782
                    split_feature = feature_names[root['split_feature']]
                else:
                    split_feature = root['split_feature']
                if split_feature == feature:
3783
                    if isinstance(root['threshold'], str):
3784
3785
3786
                        raise LightGBMError('Cannot compute split value histogram for the categorical feature')
                    else:
                        values.append(root['threshold'])
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
                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'])

3797
        if bins is None or isinstance(bins, int) and xgboost_style:
3798
3799
3800
3801
3802
3803
3804
            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:
3805
                return pd_DataFrame(ret, columns=['SplitValue', 'Count'])
3806
3807
3808
3809
3810
            else:
                return ret
        else:
            return hist, bin_edges

wxchan's avatar
wxchan committed
3811
    def __inner_eval(self, data_name, data_idx, feval=None):
3812
        """Evaluate training or validation data."""
wxchan's avatar
wxchan committed
3813
        if data_idx >= self.__num_dataset:
3814
            raise ValueError("Data_idx should be smaller than number of dataset")
wxchan's avatar
wxchan committed
3815
3816
3817
        self.__get_eval_info()
        ret = []
        if self.__num_inner_eval > 0:
3818
            result = np.empty(self.__num_inner_eval, dtype=np.float64)
Guolin Ke's avatar
Guolin Ke committed
3819
            tmp_out_len = ctypes.c_int(0)
wxchan's avatar
wxchan committed
3820
3821
            _safe_call(_LIB.LGBM_BoosterGetEval(
                self.handle,
Guolin Ke's avatar
Guolin Ke committed
3822
                ctypes.c_int(data_idx),
wxchan's avatar
wxchan committed
3823
                ctypes.byref(tmp_out_len),
Guolin Ke's avatar
Guolin Ke committed
3824
                result.ctypes.data_as(ctypes.POINTER(ctypes.c_double))))
wxchan's avatar
wxchan committed
3825
            if tmp_out_len.value != self.__num_inner_eval:
3826
                raise ValueError("Wrong length of eval results")
3827
            for i in range(self.__num_inner_eval):
3828
3829
                ret.append((data_name, self.__name_inner_eval[i],
                            result[i], self.__higher_better_inner_eval[i]))
3830
3831
        if callable(feval):
            feval = [feval]
wxchan's avatar
wxchan committed
3832
3833
3834
3835
3836
        if feval is not None:
            if data_idx == 0:
                cur_data = self.train_set
            else:
                cur_data = self.valid_sets[data_idx - 1]
3837
3838
3839
3840
3841
3842
3843
3844
3845
            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
3846
3847
3848
3849
                    ret.append((data_name, eval_name, val, is_higher_better))
        return ret

    def __inner_predict(self, data_idx):
3850
        """Predict for training and validation dataset."""
wxchan's avatar
wxchan committed
3851
        if data_idx >= self.__num_dataset:
3852
            raise ValueError("Data_idx should be smaller than number of dataset")
wxchan's avatar
wxchan committed
3853
3854
3855
3856
3857
        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
3858
            self.__inner_predict_buffer[data_idx] = np.empty(n_preds, dtype=np.float64)
3859
        # avoid to predict many time in one iteration
wxchan's avatar
wxchan committed
3860
3861
        if not self.__is_predicted_cur_iter[data_idx]:
            tmp_out_len = ctypes.c_int64(0)
Guolin Ke's avatar
Guolin Ke committed
3862
            data_ptr = self.__inner_predict_buffer[data_idx].ctypes.data_as(ctypes.POINTER(ctypes.c_double))
wxchan's avatar
wxchan committed
3863
3864
            _safe_call(_LIB.LGBM_BoosterGetPredict(
                self.handle,
Guolin Ke's avatar
Guolin Ke committed
3865
                ctypes.c_int(data_idx),
wxchan's avatar
wxchan committed
3866
3867
3868
                ctypes.byref(tmp_out_len),
                data_ptr))
            if tmp_out_len.value != len(self.__inner_predict_buffer[data_idx]):
3869
                raise ValueError(f"Wrong length of predict results for data {data_idx}")
wxchan's avatar
wxchan committed
3870
3871
3872
3873
            self.__is_predicted_cur_iter[data_idx] = True
        return self.__inner_predict_buffer[data_idx]

    def __get_eval_info(self):
3874
        """Get inner evaluation count and names."""
wxchan's avatar
wxchan committed
3875
3876
        if self.__need_reload_eval_info:
            self.__need_reload_eval_info = False
Guolin Ke's avatar
Guolin Ke committed
3877
            out_num_eval = ctypes.c_int(0)
3878
            # Get num of inner evals
wxchan's avatar
wxchan committed
3879
3880
3881
3882
3883
            _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:
3884
                # Get name of eval metrics
Guolin Ke's avatar
Guolin Ke committed
3885
                tmp_out_len = ctypes.c_int(0)
3886
3887
3888
                reserved_string_buffer_size = 255
                required_string_buffer_size = ctypes.c_size_t(0)
                string_buffers = [
3889
                    ctypes.create_string_buffer(reserved_string_buffer_size) for _ in range(self.__num_inner_eval)
3890
                ]
wxchan's avatar
wxchan committed
3891
                ptr_string_buffers = (ctypes.c_char_p * self.__num_inner_eval)(*map(ctypes.addressof, string_buffers))
wxchan's avatar
wxchan committed
3892
3893
                _safe_call(_LIB.LGBM_BoosterGetEvalNames(
                    self.handle,
3894
                    ctypes.c_int(self.__num_inner_eval),
wxchan's avatar
wxchan committed
3895
                    ctypes.byref(tmp_out_len),
3896
                    ctypes.c_size_t(reserved_string_buffer_size),
3897
                    ctypes.byref(required_string_buffer_size),
wxchan's avatar
wxchan committed
3898
3899
                    ptr_string_buffers))
                if self.__num_inner_eval != tmp_out_len.value:
3900
                    raise ValueError("Length of eval names doesn't equal with num_evals")
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
                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
                ]
3921

wxchan's avatar
wxchan committed
3922
    def attr(self, key):
3923
        """Get attribute string from the Booster.
wxchan's avatar
wxchan committed
3924
3925
3926

        Parameters
        ----------
3927
        key : str
3928
            The name of the attribute.
wxchan's avatar
wxchan committed
3929
3930
3931

        Returns
        -------
3932
        value : str or None
3933
            The attribute value.
Nikita Titov's avatar
Nikita Titov committed
3934
            Returns None if attribute does not exist.
wxchan's avatar
wxchan committed
3935
        """
3936
        return self.__attr.get(key, None)
wxchan's avatar
wxchan committed
3937
3938

    def set_attr(self, **kwargs):
3939
        """Set attributes to the Booster.
wxchan's avatar
wxchan committed
3940
3941
3942
3943

        Parameters
        ----------
        **kwargs
3944
3945
            The attributes to set.
            Setting a value to None deletes an attribute.
Nikita Titov's avatar
Nikita Titov committed
3946
3947
3948
3949

        Returns
        -------
        self : Booster
3950
            Booster with set attributes.
wxchan's avatar
wxchan committed
3951
3952
3953
        """
        for key, value in kwargs.items():
            if value is not None:
3954
                if not isinstance(value, str):
Nikita Titov's avatar
Nikita Titov committed
3955
                    raise ValueError("Only string values are accepted")
wxchan's avatar
wxchan committed
3956
3957
3958
                self.__attr[key] = value
            else:
                self.__attr.pop(key, None)
Nikita Titov's avatar
Nikita Titov committed
3959
        return self