Update of LabelTensor class and fix Simplex domain (#362)
*Implement new methods in LabelTensor and fix operators
This commit is contained in:
Filippo Olivo
committed by
Nicola Demo
Nicola Demo
parent
fdb8f65143
commit
7528f6ef74
@@ -1,5 +1,5 @@
|
||||
""" Module for LabelTensor """
|
||||
|
||||
from copy import deepcopy, copy
|
||||
import torch
|
||||
from torch import Tensor
|
||||
|
||||
@@ -35,12 +35,22 @@ class LabelTensor(torch.Tensor):
|
||||
{1: {"name": "space"['a', 'b', 'c'])
|
||||
|
||||
"""
|
||||
self.dim_names = None
|
||||
self.labels = labels
|
||||
|
||||
@property
|
||||
def labels(self):
|
||||
"""Property decorator for labels
|
||||
|
||||
:return: labels of self
|
||||
:rtype: list
|
||||
"""
|
||||
return self._labels[self.tensor.ndim-1]['dof']
|
||||
|
||||
@property
|
||||
def full_labels(self):
|
||||
"""Property decorator for labels
|
||||
|
||||
:return: labels of self
|
||||
:rtype: list
|
||||
"""
|
||||
@@ -65,6 +75,13 @@ class LabelTensor(torch.Tensor):
|
||||
self.update_labels_from_list(labels)
|
||||
else:
|
||||
raise ValueError(f"labels must be list, dict or string.")
|
||||
self.set_names()
|
||||
|
||||
def set_names(self):
|
||||
labels = self.full_labels
|
||||
self.dim_names = {}
|
||||
for dim in range(self.tensor.ndim):
|
||||
self.dim_names[labels[dim]['name']] = dim
|
||||
|
||||
def extract(self, label_to_extract):
|
||||
"""
|
||||
@@ -76,46 +93,63 @@ class LabelTensor(torch.Tensor):
|
||||
:raises TypeError: Labels are not ``str``.
|
||||
:raises ValueError: Label to extract is not in the labels ``list``.
|
||||
"""
|
||||
from copy import deepcopy
|
||||
if isinstance(label_to_extract, (str, int)):
|
||||
label_to_extract = [label_to_extract]
|
||||
if isinstance(label_to_extract, (tuple, list)):
|
||||
last_dim_label = self._labels[self.tensor.ndim - 1]['dof']
|
||||
if set(label_to_extract).issubset(last_dim_label) is False:
|
||||
raise ValueError('Cannot extract a dof which is not in the original LabelTensor')
|
||||
idx_to_extract = [last_dim_label.index(i) for i in label_to_extract]
|
||||
new_tensor = self.tensor
|
||||
new_tensor = new_tensor[..., idx_to_extract]
|
||||
new_labels = deepcopy(self._labels)
|
||||
last_dim_new_label = {self.tensor.ndim - 1: {
|
||||
'dof': label_to_extract,
|
||||
'name': self._labels[self.tensor.ndim - 1]['name']
|
||||
}}
|
||||
new_labels.update(last_dim_new_label)
|
||||
return self._extract_from_list(label_to_extract)
|
||||
elif isinstance(label_to_extract, dict):
|
||||
new_labels = (deepcopy(self._labels))
|
||||
new_tensor = self.tensor
|
||||
for k, v in label_to_extract.items():
|
||||
idx_dim = None
|
||||
for kl, vl in self._labels.items():
|
||||
if vl['name'] == k:
|
||||
idx_dim = kl
|
||||
break
|
||||
dim_labels = self._labels[idx_dim]['dof']
|
||||
if isinstance(label_to_extract[k], (int, str)):
|
||||
label_to_extract[k] = [label_to_extract[k]]
|
||||
if set(label_to_extract[k]).issubset(dim_labels) is False:
|
||||
raise ValueError('Cannot extract a dof which is not in the original LabelTensor')
|
||||
idx_to_extract = [dim_labels.index(i) for i in label_to_extract[k]]
|
||||
indexer = [slice(None)] * idx_dim + [idx_to_extract] + [slice(None)] * (self.tensor.ndim - idx_dim - 1)
|
||||
new_tensor = new_tensor[indexer]
|
||||
dim_new_label = {idx_dim: {
|
||||
'dof': label_to_extract[k],
|
||||
'name': self._labels[idx_dim]['name']
|
||||
}}
|
||||
new_labels.update(dim_new_label)
|
||||
return self._extract_from_dict(label_to_extract)
|
||||
else:
|
||||
raise ValueError('labels_to_extract must be str or list or dict')
|
||||
|
||||
def _extract_from_list(self, labels_to_extract):
|
||||
#Store locally all necessary obj/variables
|
||||
ndim = self.tensor.ndim
|
||||
labels = self.full_labels
|
||||
tensor = self.tensor
|
||||
last_dim_label = self.labels
|
||||
|
||||
#Verify if all the labels in labels_to_extract are in last dimension
|
||||
if set(labels_to_extract).issubset(last_dim_label) is False:
|
||||
raise ValueError('Cannot extract a dof which is not in the original LabelTensor')
|
||||
|
||||
#Extract index to extract
|
||||
idx_to_extract = [last_dim_label.index(i) for i in labels_to_extract]
|
||||
|
||||
#Perform extraction
|
||||
new_tensor = tensor[..., idx_to_extract]
|
||||
|
||||
#Manage labels
|
||||
new_labels = copy(labels)
|
||||
|
||||
last_dim_new_label = {ndim - 1: {
|
||||
'dof': list(labels_to_extract),
|
||||
'name': labels[ndim - 1]['name']
|
||||
}}
|
||||
new_labels.update(last_dim_new_label)
|
||||
return LabelTensor(new_tensor, new_labels)
|
||||
|
||||
def _extract_from_dict(self, labels_to_extract):
|
||||
labels = self.full_labels
|
||||
tensor = self.tensor
|
||||
ndim = tensor.ndim
|
||||
new_labels = deepcopy(labels)
|
||||
new_tensor = tensor
|
||||
for k, _ in labels_to_extract.items():
|
||||
idx_dim = self.dim_names[k]
|
||||
dim_labels = labels[idx_dim]['dof']
|
||||
if isinstance(labels_to_extract[k], (int, str)):
|
||||
labels_to_extract[k] = [labels_to_extract[k]]
|
||||
if set(labels_to_extract[k]).issubset(dim_labels) is False:
|
||||
raise ValueError('Cannot extract a dof which is not in the original LabelTensor')
|
||||
idx_to_extract = [dim_labels.index(i) for i in labels_to_extract[k]]
|
||||
indexer = [slice(None)] * idx_dim + [idx_to_extract] + [slice(None)] * (ndim - idx_dim - 1)
|
||||
new_tensor = new_tensor[indexer]
|
||||
dim_new_label = {idx_dim: {
|
||||
'dof': labels_to_extract[k],
|
||||
'name': labels[idx_dim]['name']
|
||||
}}
|
||||
new_labels.update(dim_new_label)
|
||||
return LabelTensor(new_tensor, new_labels)
|
||||
|
||||
def __str__(self):
|
||||
@@ -147,32 +181,42 @@ class LabelTensor(torch.Tensor):
|
||||
return []
|
||||
if len(tensors) == 1:
|
||||
return tensors[0]
|
||||
n_dims = tensors[0].ndim
|
||||
new_labels_cat_dim = []
|
||||
for i in range(n_dims):
|
||||
name = tensors[0].labels[i]['name']
|
||||
if i != dim:
|
||||
dof = tensors[0].labels[i]['dof']
|
||||
for tensor in tensors:
|
||||
dof_to_check = tensor.labels[i]['dof']
|
||||
name_to_check = tensor.labels[i]['name']
|
||||
if dof != dof_to_check or name != name_to_check:
|
||||
raise ValueError('dimensions must have the same dof and name')
|
||||
else:
|
||||
for tensor in tensors:
|
||||
new_labels_cat_dim += tensor.labels[i]['dof']
|
||||
name_to_check = tensor.labels[i]['name']
|
||||
if name != name_to_check:
|
||||
raise ValueError('dimensions must have the same dof and name')
|
||||
new_labels_cat_dim = LabelTensor._check_validity_before_cat(tensors, dim)
|
||||
|
||||
# Perform cat on tensors
|
||||
new_tensor = torch.cat(tensors, dim=dim)
|
||||
labels = tensors[0].labels
|
||||
|
||||
#Update labels
|
||||
labels = tensors[0].full_labels
|
||||
labels.pop(dim)
|
||||
new_labels_cat_dim = new_labels_cat_dim if len(set(new_labels_cat_dim)) == len(new_labels_cat_dim) \
|
||||
else range(new_tensor.shape[dim])
|
||||
labels[dim] = {'dof': new_labels_cat_dim,
|
||||
'name': tensors[1].labels[dim]['name']}
|
||||
'name': tensors[1].full_labels[dim]['name']}
|
||||
return LabelTensor(new_tensor, labels)
|
||||
|
||||
@staticmethod
|
||||
def _check_validity_before_cat(tensors, dim):
|
||||
n_dims = tensors[0].ndim
|
||||
new_labels_cat_dim = []
|
||||
# Check if names and dof of the labels are the same in all dimensions except in dim
|
||||
for i in range(n_dims):
|
||||
name = tensors[0].full_labels[i]['name']
|
||||
if i != dim:
|
||||
dof = tensors[0].full_labels[i]['dof']
|
||||
for tensor in tensors:
|
||||
dof_to_check = tensor.full_labels[i]['dof']
|
||||
name_to_check = tensor.full_labels[i]['name']
|
||||
if dof != dof_to_check or name != name_to_check:
|
||||
raise ValueError('dimensions must have the same dof and name')
|
||||
else:
|
||||
for tensor in tensors:
|
||||
new_labels_cat_dim += tensor.full_labels[i]['dof']
|
||||
name_to_check = tensor.full_labels[i]['name']
|
||||
if name != name_to_check:
|
||||
raise ValueError('Dimensions to concatenate must have the same name')
|
||||
return new_labels_cat_dim
|
||||
|
||||
def requires_grad_(self, mode=True):
|
||||
lt = super().requires_grad_(mode)
|
||||
lt.labels = self._labels
|
||||
@@ -204,7 +248,6 @@ class LabelTensor(torch.Tensor):
|
||||
out = LabelTensor(super().clone(*args, **kwargs), self._labels)
|
||||
return out
|
||||
|
||||
|
||||
def init_labels(self):
|
||||
self._labels = {
|
||||
idx_: {
|
||||
@@ -221,13 +264,14 @@ class LabelTensor(torch.Tensor):
|
||||
:type labels: dict
|
||||
:raises ValueError: dof list contain duplicates or number of dof does not match with tensor shape
|
||||
"""
|
||||
|
||||
tensor_shape = self.tensor.shape
|
||||
#Check dimensionality
|
||||
for k, v in labels.items():
|
||||
if len(v['dof']) != len(set(v['dof'])):
|
||||
raise ValueError("dof must be unique")
|
||||
if len(v['dof']) != tensor_shape[k]:
|
||||
raise ValueError('Number of dof does not match with tensor dimension')
|
||||
#Perform update
|
||||
self._labels.update(labels)
|
||||
|
||||
def update_labels_from_list(self, labels):
|
||||
@@ -237,6 +281,7 @@ class LabelTensor(torch.Tensor):
|
||||
:param labels: The label(s) to update.
|
||||
:type labels: list
|
||||
"""
|
||||
# Create a dict with labels
|
||||
last_dim_labels = {self.tensor.ndim - 1: {'dof': labels, 'name': self.tensor.ndim - 1}}
|
||||
self.update_labels_from_dict(last_dim_labels)
|
||||
|
||||
@@ -246,26 +291,103 @@ class LabelTensor(torch.Tensor):
|
||||
raise ValueError('tensors list must not be empty')
|
||||
if len(tensors) == 1:
|
||||
return tensors[0]
|
||||
labels = tensors[0].labels
|
||||
# Collect all labels
|
||||
labels = tensors[0].full_labels
|
||||
# Check labels of all the tensors in each dimension
|
||||
for j in range(tensors[0].ndim):
|
||||
for i in range(1, len(tensors)):
|
||||
if labels[j] != tensors[i].labels[j]:
|
||||
if labels[j] != tensors[i].full_labels[j]:
|
||||
labels.pop(j)
|
||||
break
|
||||
|
||||
# Sum tensors
|
||||
data = torch.zeros(tensors[0].tensor.shape)
|
||||
for i in range(len(tensors)):
|
||||
data += tensors[i].tensor
|
||||
new_tensor = LabelTensor(data, labels)
|
||||
return new_tensor
|
||||
|
||||
def last_dim_dof(self):
|
||||
return self._labels[self.tensor.ndim - 1]['dof']
|
||||
|
||||
def append(self, tensor, mode='std'):
|
||||
print(self.labels)
|
||||
print(tensor.labels)
|
||||
if mode == 'std':
|
||||
# Call cat on last dimension
|
||||
new_label_tensor = LabelTensor.cat([self, tensor], dim=self.tensor.ndim - 1)
|
||||
|
||||
elif mode=='cross':
|
||||
# Crete tensor and call cat on last dimension
|
||||
tensor1 = self
|
||||
tensor2 = tensor
|
||||
n1 = tensor1.shape[0]
|
||||
n2 = tensor2.shape[0]
|
||||
tensor1 = LabelTensor(tensor1.repeat(n2, 1), labels=tensor1.labels)
|
||||
tensor2 = LabelTensor(tensor2.repeat_interleave(n1, dim=0), labels=tensor2.labels)
|
||||
new_label_tensor = LabelTensor.cat([tensor1, tensor2], dim=self.tensor.ndim-1)
|
||||
else:
|
||||
raise ValueError('mode must be either "std" or "cross"')
|
||||
return new_label_tensor
|
||||
|
||||
@staticmethod
|
||||
def vstack(label_tensors):
|
||||
"""
|
||||
Stack tensors vertically. For more details, see
|
||||
:meth:`torch.vstack`.
|
||||
|
||||
:param list(LabelTensor) label_tensors: the tensors to stack. They need
|
||||
to have equal labels.
|
||||
:return: the stacked tensor
|
||||
:rtype: LabelTensor
|
||||
"""
|
||||
return LabelTensor.cat(label_tensors, dim=0)
|
||||
|
||||
def __getitem__(self, index):
|
||||
"""
|
||||
Return a copy of the selected tensor.
|
||||
"""
|
||||
|
||||
if isinstance(index, str) or (isinstance(index, (tuple, list)) and all(isinstance(a, str) for a in index)):
|
||||
return self.extract(index)
|
||||
|
||||
selected_lt = super().__getitem__(index)
|
||||
|
||||
try:
|
||||
len_index = len(index)
|
||||
except TypeError:
|
||||
len_index = 1
|
||||
|
||||
if isinstance(index, int) or len_index == 1:
|
||||
if selected_lt.ndim == 1:
|
||||
selected_lt = selected_lt.reshape(1, -1)
|
||||
if hasattr(self, "labels"):
|
||||
new_labels = deepcopy(self.full_labels)
|
||||
new_labels.pop(0)
|
||||
selected_lt.labels = new_labels
|
||||
elif len(index) == self.tensor.ndim:
|
||||
new_labels = deepcopy(self.full_labels)
|
||||
if selected_lt.ndim == 1:
|
||||
selected_lt = selected_lt.reshape(-1, 1)
|
||||
for j in range(selected_lt.ndim):
|
||||
if hasattr(self, "labels"):
|
||||
if isinstance(index[j], list):
|
||||
new_labels.update({j: {'dof': [new_labels[j]['dof'][i] for i in index[1]],
|
||||
'name':new_labels[j]['name']}})
|
||||
else:
|
||||
new_labels.update({j: {'dof': new_labels[j]['dof'][index[j]],
|
||||
'name':new_labels[j]['name']}})
|
||||
|
||||
selected_lt.labels = new_labels
|
||||
else:
|
||||
new_labels = deepcopy(self.full_labels)
|
||||
new_labels.update({0: {'dof': list[index], 'name': new_labels[0]['name']}})
|
||||
selected_lt.labels = self.labels
|
||||
|
||||
return selected_lt
|
||||
|
||||
def sort_labels(self, dim=None):
|
||||
def argsort(lst):
|
||||
return sorted(range(len(lst)), key=lambda x: lst[x])
|
||||
if dim is None:
|
||||
dim = self.tensor.ndim-1
|
||||
labels = self.full_labels[dim]['dof']
|
||||
sorted_index = argsort(labels)
|
||||
indexer = [slice(None)] * self.tensor.ndim
|
||||
indexer[dim] = sorted_index
|
||||
new_labels = deepcopy(self.full_labels)
|
||||
new_labels[dim] = {'dof': sorted(labels), 'name': new_labels[dim]['name']}
|
||||
return LabelTensor(self.tensor[indexer], new_labels)
|
||||
Reference in New Issue
Block a user