folivo/PINA
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

edited utils to take list (#115)

Browse Source 
* enhanced difference domain
* refactored utils
* fixed typo
* added tests
---------

Co-authored-by: Dario Coscia <93731561+dario-coscia@users.noreply.github.com>
This commit is contained in:
Kush
2023-06-19 18:47:52 +02:00
committed by Nicola Demo
parent aaf2bed732
commit 62ec69ccac
9 changed files with 73 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
pina/geometry/difference_domain.py
View File

@@ -3,16 +3,17 @@
from .location import Location
from ..label_tensor import LabelTensor
class Difference(Location):
"""
"""
def __init__(self, first, second):
def __init__(self, first, second):
self.first = first
self.second = second
def sample(self, n, mode ='random', variables='all'):
def sample(self, n, mode='random', variables='all'):
"""
"""
assert mode is 'random', 'Only random mode is implemented'
@@ -24,4 +25,4 @@ class Difference(Location):
samples.append(sample.tolist()[0])
import torch
return LabelTensor(torch.tensor(samples), labels=['x', 'y'])
return LabelTensor(torch.tensor(samples), labels=['x', 'y'])

19
pina/geometry/union_domain.py
View File

@@ -25,9 +25,9 @@ class Union(Location):
super().__init__()
# union checks
self._check_union_inheritance(geometries)
self._check_union_consistency(geometries)
check_consistency(geometries, Location)
self._check_union_dimensions(geometries)
# assign geometries
self._geometries = geometries
@@ -36,7 +36,7 @@ class Union(Location):
"""
The geometries."""
return self._geometries
@property
def variables(self):
"""
@@ -116,7 +116,7 @@ class Union(Location):
return LabelTensor(torch.cat(sampled_points), labels=[f'{i}' for i in self.variables])
def _check_union_consistency(self, geometries):
def _check_union_dimensions(self, geometries):
"""Check if the dimensions of the geometries are consistent.
:param geometries: Geometries to be checked.
@@ -126,12 +126,3 @@ class Union(Location):
if geometry.variables != geometries[0].variables:
raise NotImplementedError(
f'The geometries need to be the same dimensions. {geometry.variables} is not equal to {geometries[0].variables}')
def _check_union_inheritance(self, geometries):
"""Check if the geometries are inherited from 'pina.geometry.Location'.
param geometries: Geometries to be checked.
:type geometries: list[Location]
"""
for idx, geometry in enumerate(geometries):
check_consistency(geometry, Location, f'geometry[{idx}]')

4
pina/loss.py
View File

@@ -108,9 +108,9 @@ class LpLoss(LossInterface):
super().__init__(reduction=reduction)
# check consistency
check_consistency(p, (str,int,float), 'degree p')
check_consistency(p, (str,int,float))
self.p = p
check_consistency(relative, bool, 'relative')
check_consistency(relative, bool)
self.relative = relative
def forward(self, input, target):

10
pina/model/network.py
View File

@@ -4,20 +4,20 @@ from ..utils import check_consistency
class Network(torch.nn.Module):
def __init__(self, model, extra_features=None):
super().__init__()
# check model consistency
check_consistency(model, nn.Module, 'torch model')
check_consistency(model, nn.Module)
self._model = model
# check consistency and assign extra fatures
# check consistency and assign extra fatures
if extra_features is None:
self._extra_features = []
else:
for feat in extra_features:
check_consistency(feat, nn.Module, 'extra features')
check_consistency(feat, nn.Module)
self._extra_features = nn.Sequential(*extra_features)
# check model works with inputs
@@ -44,4 +44,4 @@ class Network(torch.nn.Module):
@property
def extra_features(self):
return self._extra_features
return self._extra_features

10
pina/pinn.py
View File

@@ -48,11 +48,11 @@ class PINN(SolverInterface):
super().__init__(model=model, problem=problem, extra_features=extra_features)
# check consistency
check_consistency(optimizer, torch.optim.Optimizer, 'optimizer', subclass=True)
check_consistency(optimizer_kwargs, dict, 'optimizer_kwargs')
check_consistency(scheduler, LRScheduler, 'scheduler', subclass=True)
check_consistency(scheduler_kwargs, dict, 'scheduler_kwargs')
check_consistency(loss, (LossInterface, _Loss), 'loss', subclass=False)
check_consistency(optimizer, torch.optim.Optimizer, subclass=True)
check_consistency(optimizer_kwargs, dict)
check_consistency(scheduler, LRScheduler, subclass=True)
check_consistency(scheduler_kwargs, dict)
check_consistency(loss, (LossInterface, _Loss), subclass=False)
# assign variables
self._optimizer = optimizer(self.model.parameters(), **optimizer_kwargs)

2
pina/solver.py
View File

@@ -20,7 +20,7 @@ class SolverInterface(pl.LightningModule, metaclass=ABCMeta):
super().__init__()
# check inheritance for pina problem
check_consistency(problem, AbstractProblem, 'pina problem')
check_consistency(problem, AbstractProblem)
# assigning class variables (check consistency inside Network class)
self._pina_model = Network(model=model, extra_features=extra_features)

2
pina/trainer.py
View File

@@ -11,7 +11,7 @@ class Trainer(pl.Trainer):
super().__init__(**kwargs)
# check inheritance consistency for solver
check_consistency(solver, SolverInterface, 'Solver model')
check_consistency(solver, SolverInterface)
self._model = solver
# create dataloader

36
pina/utils.py
View File

@@ -1,4 +1,5 @@
"""Utils module"""
from torch.utils.data import Dataset, DataLoader
from functools import reduce
import types
@@ -10,14 +11,14 @@ from .label_tensor import LabelTensor
import torch
def check_consistency(object, object_instance, object_name, subclass=False):
def check_consistency(object, object_instance, subclass=False):
"""Helper function to check object inheritance consistency.
Given a specific ``'object'`` we check if the object is
instance of a specific ``'object_instance'``, or in case
``'subclass=True'`` we check if the object is subclass
if the ``'object_instance'``.
:param Object object: The object to check the inheritance
:param (iterable or class object) object: The object to check the inheritance
:param Object object_instance: The parent class from where the object
is expected to inherit
:param str object_name: The name of the object
@@ -25,12 +26,17 @@ def check_consistency(object, object_instance, object_name, subclass=False):
:raises ValueError: If the object does not inherit from the
specified class
"""
if not subclass:
if not isinstance(object, object_instance):
raise ValueError(f"{object_name} must be {object_instance}")
else:
if not issubclass(object, object_instance):
raise ValueError(f"{object_name} must be {object_instance}")
if not isinstance(object, (list, set, tuple)):
object = [object]
for obj in object:
try:
if not subclass:
assert isinstance(obj, object_instance)
else:
assert issubclass(obj, object_instance)
except AssertionError:
raise ValueError(f"{type(obj).__name__} must be {object_instance}.")
def number_parameters(model, aggregate=True, only_trainable=True): # TODO: check
@@ -180,13 +186,13 @@ def chebyshev_roots(n):
# def __len__(self):
# return self._len
from torch.utils.data import Dataset, DataLoader
class LabelTensorDataset(Dataset):
def __init__(self, d):
for k, v in d.items():
setattr(self, k, v)
self.labels = list(d.keys())
def __getitem__(self, index):
print(index)
result = {}
@@ -201,7 +207,7 @@ class LabelTensorDataset(Dataset):
result[label] = sample_tensor[index]
except IndexError:
result[label] = torch.tensor([])
print(result)
return result
@@ -229,13 +235,13 @@ class LabelTensorDataLoader(DataLoader):
# def __len__(self):
# return self._len
from torch.utils.data import Dataset, DataLoader
class LabelTensorDataset(Dataset):
def __init__(self, d):
for k, v in d.items():
setattr(self, k, v)
self.labels = list(d.keys())
def __getitem__(self, index):
print(index)
result = {}
@@ -250,7 +256,7 @@ class LabelTensorDataset(Dataset):
result[label] = sample_tensor[index]
except IndexError:
result[label] = torch.tensor([])
print(result)
return result
@@ -261,4 +267,4 @@ class LabelTensorDataset(Dataset):
class LabelTensorDataLoader(DataLoader):
def collate_fn(self, data):
pass
pass