Skip to content

GitLab

Unverified Commit 0e541cc9 authored by bwdeng20's avatar bwdeng20 Committed by GitHub
Browse files

Merge branch 'master' into master

parents 0090f4ed 056c0bab
Hide whitespace changes
Inline Side-by-side
Showing with 129 additions and 50 deletions
torch_sparse/saint.py 0 → 100644
View file @ 0e541cc9
from typing import Tuple
import torch
from torch_sparse.tensor import SparseTensor
def saint_subgraph(src: SparseTensor, node_idx: torch.Tensor
) -> Tuple[SparseTensor, torch.Tensor]:
row, col, value = src.coo()
rowptr = src.storage.rowptr()
data = torch.ops.torch_sparse.saint_subgraph(node_idx, rowptr, row, col)
row, col, edge_index = data
if value is not None:
value = value[edge_index]
out = SparseTensor(row=row, rowptr=None, col=col, value=value,
sparse_sizes=(node_idx.size(0), node_idx.size(0)),
is_sorted=True)
return out, edge_index
SparseTensor.saint_subgraph = saint_subgraph
torch_sparse/tensor.py
View file @ 0e541cc9
...@@ -12,17 +12,25 @@ from torch_sparse.utils import is_scalar ...@@ -12,17 +12,25 @@ from torch_sparse.utils import is_scalar
class SparseTensor(object): class SparseTensor(object):
storage: SparseStorage storage: SparseStorage
def __init__(self, row: Optional[torch.Tensor] = None, def __init__(self,
row: Optional[torch.Tensor] = None,
rowptr: Optional[torch.Tensor] = None, rowptr: Optional[torch.Tensor] = None,
col: Optional[torch.Tensor] = None, col: Optional[torch.Tensor] = None,
value: Optional[torch.Tensor] = None, value: Optional[torch.Tensor] = None,
sparse_sizes: Optional[Tuple[int, int]] = None, sparse_sizes: Optional[Tuple[int, int]] = None,
is_sorted: bool = False): is_sorted: bool = False):
self.storage = SparseStorage(row=row, rowptr=rowptr, col=col, self.storage = SparseStorage(
value=value, sparse_sizes=sparse_sizes, row=row,
rowcount=None, colptr=None, colcount=None, rowptr=rowptr,
csr2csc=None, csc2csr=None, col=col,
is_sorted=is_sorted) value=value,
sparse_sizes=sparse_sizes,
rowcount=None,
colptr=None,
colcount=None,
csr2csc=None,
csc2csr=None,
is_sorted=is_sorted)
@classmethod @classmethod
def from_storage(self, storage: SparseStorage): def from_storage(self, storage: SparseStorage):
...@@ -45,12 +53,17 @@ class SparseTensor(object): ...@@ -45,12 +53,17 @@ class SparseTensor(object):
if has_value: if has_value:
value = mat[row, col] value = mat[row, col]
return SparseTensor(row=row, rowptr=None, col=col, value=value, return SparseTensor(
sparse_sizes=(mat.size(0), mat.size(1)), row=row,
is_sorted=True) rowptr=None,
col=col,
value=value,
sparse_sizes=(mat.size(0), mat.size(1)),
is_sorted=True)
@classmethod @classmethod
def from_torch_sparse_coo_tensor(self, mat: torch.Tensor, def from_torch_sparse_coo_tensor(self,
mat: torch.Tensor,
has_value: bool = True): has_value: bool = True):
mat = mat.coalesce() mat = mat.coalesce()
index = mat._indices() index = mat._indices()
...@@ -60,13 +73,20 @@ class SparseTensor(object): ...@@ -60,13 +73,20 @@ class SparseTensor(object):
if has_value: if has_value:
value = mat._values() value = mat._values()
return SparseTensor(row=row, rowptr=None, col=col, value=value, return SparseTensor(
sparse_sizes=(mat.size(0), mat.size(1)), row=row,
is_sorted=True) rowptr=None,
col=col,
value=value,
sparse_sizes=(mat.size(0), mat.size(1)),
is_sorted=True)
@classmethod @classmethod
def eye(self, M: int, N: Optional[int] = None, def eye(self,
options: Optional[torch.Tensor] = None, has_value: bool = True, M: int,
N: Optional[int] = None,
options: Optional[torch.Tensor] = None,
has_value: bool = True,
fill_cache: bool = False): fill_cache: bool = False):
N = M if N is None else N N = M if N is None else N
...@@ -84,8 +104,8 @@ class SparseTensor(object): ...@@ -84,8 +104,8 @@ class SparseTensor(object):
value: Optional[torch.Tensor] = None value: Optional[torch.Tensor] = None
if has_value: if has_value:
if options is not None: if options is not None:
value = torch.ones(row.numel(), dtype=options.dtype, value = torch.ones(
device=row.device) row.numel(), dtype=options.dtype, device=row.device)
else: else:
value = torch.ones(row.numel(), device=row.device) value = torch.ones(row.numel(), device=row.device)
...@@ -108,9 +128,17 @@ class SparseTensor(object): ...@@ -108,9 +128,17 @@ class SparseTensor(object):
csr2csc = csc2csr = row csr2csc = csc2csr = row
storage: SparseStorage = SparseStorage( storage: SparseStorage = SparseStorage(
row=row, rowptr=rowptr, col=col, value=value, sparse_sizes=(M, N), row=row,
rowcount=rowcount, colptr=colptr, colcount=colcount, rowptr=rowptr,
csr2csc=csr2csc, csc2csr=csc2csr, is_sorted=True) col=col,
value=value,
sparse_sizes=(M, N),
rowcount=rowcount,
colptr=colptr,
colcount=colcount,
csr2csc=csr2csc,
csc2csr=csc2csr,
is_sorted=True)
self = SparseTensor.__new__(SparseTensor) self = SparseTensor.__new__(SparseTensor)
self.storage = storage self.storage = storage
...@@ -153,12 +181,14 @@ class SparseTensor(object): ...@@ -153,12 +181,14 @@ class SparseTensor(object):
def has_value(self) -> bool: def has_value(self) -> bool:
return self.storage.has_value() return self.storage.has_value()
def set_value_(self, value: Optional[torch.Tensor], def set_value_(self,
value: Optional[torch.Tensor],
layout: Optional[str] = None): layout: Optional[str] = None):
self.storage.set_value_(value, layout) self.storage.set_value_(value, layout)
return self return self
def set_value(self, value: Optional[torch.Tensor], def set_value(self,
value: Optional[torch.Tensor],
layout: Optional[str] = None): layout: Optional[str] = None):
return self.from_storage(self.storage.set_value(value, layout)) return self.from_storage(self.storage.set_value(value, layout))
...@@ -187,23 +217,31 @@ class SparseTensor(object): ...@@ -187,23 +217,31 @@ class SparseTensor(object):
# Utility functions ####################################################### # Utility functions #######################################################
def fill_value_(self, fill_value: float, def fill_value_(self,
fill_value: float,
options: Optional[torch.Tensor] = None): options: Optional[torch.Tensor] = None):
if options is not None: if options is not None:
value = torch.full((self.nnz(), ), fill_value, dtype=options.dtype, value = torch.full((self.nnz(), ),
fill_value,
dtype=options.dtype,
device=self.device()) device=self.device())
else: else:
value = torch.full((self.nnz(), ), fill_value, value = torch.full((self.nnz(), ),
fill_value,
device=self.device()) device=self.device())
return self.set_value_(value, layout='coo') return self.set_value_(value, layout='coo')
def fill_value(self, fill_value: float, def fill_value(self,
fill_value: float,
options: Optional[torch.Tensor] = None): options: Optional[torch.Tensor] = None):
if options is not None: if options is not None:
value = torch.full((self.nnz(), ), fill_value, dtype=options.dtype, value = torch.full((self.nnz(), ),
fill_value,
dtype=options.dtype,
device=self.device()) device=self.device())
else: else:
value = torch.full((self.nnz(), ), fill_value, value = torch.full((self.nnz(), ),
fill_value,
device=self.device()) device=self.device())
return self.set_value(value, layout='coo') return self.set_value(value, layout='coo')
...@@ -270,8 +308,13 @@ class SparseTensor(object): ...@@ -270,8 +308,13 @@ class SparseTensor(object):
N = max(self.size(0), self.size(1)) N = max(self.size(0), self.size(1))
out = SparseTensor(row=row, rowptr=None, col=col, value=value, out = SparseTensor(
sparse_sizes=(N, N), is_sorted=False) row=row,
rowptr=None,
col=col,
value=value,
sparse_sizes=(N, N),
is_sorted=False)
out = out.coalesce(reduce) out = out.coalesce(reduce)
return out return out
...@@ -294,7 +337,8 @@ class SparseTensor(object): ...@@ -294,7 +337,8 @@ class SparseTensor(object):
else: else:
return False return False
def requires_grad_(self, requires_grad: bool = True, def requires_grad_(self,
requires_grad: bool = True,
options: Optional[torch.Tensor] = None): options: Optional[torch.Tensor] = None):
if requires_grad and not self.has_value(): if requires_grad and not self.has_value():
self.fill_value_(1., options=options) self.fill_value_(1., options=options)
...@@ -315,8 +359,8 @@ class SparseTensor(object): ...@@ -315,8 +359,8 @@ class SparseTensor(object):
if value is not None: if value is not None:
return value return value
else: else:
return torch.tensor(0., dtype=torch.float, return torch.tensor(
device=self.storage.col().device) 0., dtype=torch.float, device=self.storage.col().device)
def device(self): def device(self):
return self.storage.col().device return self.storage.col().device
...@@ -324,7 +368,8 @@ class SparseTensor(object): ...@@ -324,7 +368,8 @@ class SparseTensor(object):
def cpu(self): def cpu(self):
return self.device_as(torch.tensor(0.), non_blocking=False) return self.device_as(torch.tensor(0.), non_blocking=False)
def cuda(self, options: Optional[torch.Tensor] = None, def cuda(self,
options: Optional[torch.Tensor] = None,
non_blocking: bool = False): non_blocking: bool = False):
if options is not None: if options is not None:
return self.device_as(options, non_blocking) return self.device_as(options, non_blocking)
...@@ -387,19 +432,19 @@ class SparseTensor(object): ...@@ -387,19 +432,19 @@ class SparseTensor(object):
row, col, value = self.coo() row, col, value = self.coo()
if value is not None: if value is not None:
mat = torch.zeros(self.sizes(), dtype=value.dtype, mat = torch.zeros(
device=self.device()) self.sizes(), dtype=value.dtype, device=self.device())
elif options is not None: elif options is not None:
mat = torch.zeros(self.sizes(), dtype=options.dtype, mat = torch.zeros(
device=self.device()) self.sizes(), dtype=options.dtype, device=self.device())
else: else:
mat = torch.zeros(self.sizes(), device=self.device()) mat = torch.zeros(self.sizes(), device=self.device())
if value is not None: if value is not None:
mat[row, col] = value mat[row, col] = value
else: else:
mat[row, col] = torch.ones(self.nnz(), dtype=mat.dtype, mat[row, col] = torch.ones(
device=mat.device) self.nnz(), dtype=mat.dtype, device=mat.device)
return mat return mat
...@@ -409,8 +454,8 @@ class SparseTensor(object): ...@@ -409,8 +454,8 @@ class SparseTensor(object):
index = torch.stack([row, col], dim=0) index = torch.stack([row, col], dim=0)
if value is None: if value is None:
if options is not None: if options is not None:
value = torch.ones(self.nnz(), dtype=options.dtype, value = torch.ones(
device=self.device()) self.nnz(), dtype=options.dtype, device=self.device())
else: else:
value = torch.ones(self.nnz(), device=self.device()) value = torch.ones(self.nnz(), device=self.device())
...@@ -434,7 +479,7 @@ def is_shared(self: SparseTensor) -> bool: ...@@ -434,7 +479,7 @@ def is_shared(self: SparseTensor) -> bool:
def to(self, *args: Optional[List[Any]], def to(self, *args: Optional[List[Any]],
**kwargs: Optional[Dict[str, Any]]) -> SparseTensor: **kwargs: Optional[Dict[str, Any]]) -> SparseTensor:
device, dtype, non_blocking, _ = torch._C._nn._parse_to(*args, **kwargs) device, dtype, non_blocking = torch._C._nn._parse_to(*args, **kwargs)
if dtype is not None: if dtype is not None:
self = self.type_as(torch.tensor(0., dtype=dtype)) self = self.type_as(torch.tensor(0., dtype=dtype))
...@@ -515,8 +560,8 @@ SparseTensor.__repr__ = __repr__ ...@@ -515,8 +560,8 @@ SparseTensor.__repr__ = __repr__
# Scipy Conversions ########################################################### # Scipy Conversions ###########################################################
ScipySparseMatrix = Union[scipy.sparse.coo_matrix, scipy.sparse.csr_matrix, ScipySparseMatrix = Union[scipy.sparse.coo_matrix, scipy.sparse.
scipy.sparse.csc_matrix] csr_matrix, scipy.sparse.csc_matrix]
@torch.jit.ignore @torch.jit.ignore
...@@ -535,16 +580,25 @@ def from_scipy(mat: ScipySparseMatrix, has_value: bool = True) -> SparseTensor: ...@@ -535,16 +580,25 @@ def from_scipy(mat: ScipySparseMatrix, has_value: bool = True) -> SparseTensor:
value = torch.from_numpy(mat.data) value = torch.from_numpy(mat.data)
sparse_sizes = mat.shape[:2] sparse_sizes = mat.shape[:2]
storage = SparseStorage(row=row, rowptr=rowptr, col=col, value=value, storage = SparseStorage(
sparse_sizes=sparse_sizes, rowcount=None, row=row,
colptr=colptr, colcount=None, csr2csc=None, rowptr=rowptr,
csc2csr=None, is_sorted=True) col=col,
value=value,
sparse_sizes=sparse_sizes,
rowcount=None,
colptr=colptr,
colcount=None,
csr2csc=None,
csc2csr=None,
is_sorted=True)
return SparseTensor.from_storage(storage) return SparseTensor.from_storage(storage)
@torch.jit.ignore @torch.jit.ignore
def to_scipy(self: SparseTensor, layout: Optional[str] = None, def to_scipy(self: SparseTensor,
layout: Optional[str] = None,
dtype: Optional[torch.dtype] = None) -> ScipySparseMatrix: dtype: Optional[torch.dtype] = None) -> ScipySparseMatrix:
assert self.dim() == 2 assert self.dim() == 2
layout = get_layout(layout) layout = get_layout(layout)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment