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

Added 'sample_surface' parameter to CartesianDomain (#190)

Browse Source 
* Added 'sample_surface' parameter to CartesianDomain
* Changed comments from 'Span' to 'CartesianDomain'
This commit is contained in:
Ben Volokh
2023-10-24 03:43:33 -04:00
committed by Nicola Demo
parent dab5595ee6
commit da38b6456a
1 changed files with 18 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
pina/geometry/cartesian.py
View File

@@ -2,22 +2,31 @@ import torch
from .location import Location
from ..label_tensor import LabelTensor
from ..utils import torch_lhs, chebyshev_roots
from ..utils import torch_lhs, chebyshev_roots, check_consistency
class CartesianDomain(Location):
"""PINA implementation of Hypercube domain."""
def __init__(self, span_dict):
def __init__(self, span_dict, sample_surface=False):
"""
:param span_dict: A dictionary with dict-key a string representing
the input variables for the pinn, and dict-value a list with
the domain extrema.
:type span_dict: dict
:param sample_surface: A variable for choosing sample strategies. If
`sample_surface=True` only samples on the Cartesian surface
frontier are taken. If `sample_surface=False`, no such criteria
is followed.
:type sample_surface: bool
:Example:
>>> spatial_domain = CartesianDomain({'x': [0, 1], 'y': [0, 1]})
"""
# check consistency of sample_surface as bool
check_consistency(sample_surface, bool)
self._sample_surface = sample_surface
self.fixed_ = {}
self.range_ = {}
@@ -42,13 +51,13 @@ class CartesianDomain(Location):
"""Adding new dimensions on the span
:param new_span: A new span object to merge
:type new_span: Span
:type new_span: CartesianDomain
:Example:
>>> spatial_domain = Span({'x': [0, 1], 'y': [0, 1]})
>>> spatial_domain = CartesianDomain({'x': [0, 1], 'y': [0, 1]})
>>> spatial_domain.variables
['x', 'y']
>>> spatial_domain_2 = Span({'z': [3, 4], 'w': [0, 1]})
>>> spatial_domain_2 = CartesianDomain({'z': [3, 4], 'w': [0, 1]})
>>> spatial_domain.update(spatial_domain_2)
>>> spatial_domain.variables
['x', 'y', 'z', 'w']
@@ -74,7 +83,7 @@ class CartesianDomain(Location):
"""
dim = bounds.shape[0]
if mode in ['chebyshev', 'grid'] and dim != 1:
raise RuntimeError('Something wrong in Span...')
raise RuntimeError('Something wrong in CartesianDomain...')
if mode == 'random':
pts = torch.rand(size=(n, dim))
@@ -115,10 +124,10 @@ class CartesianDomain(Location):
are sampled all together, and the final number of points
.. warning::
The extrema values of Span are always sampled only for 'grid' mode.
The extrema values of CartesianDomain are always sampled only for 'grid' mode.
:Example:
>>> spatial_domain = Span({'x': [0, 1], 'y': [0, 1]})
>>> spatial_domain = CartesianDomain({'x': [0, 1], 'y': [0, 1]})
>>> spatial_domain.sample(n=4, mode='random')
tensor([[0.0108, 0.7643],
[0.4477, 0.8015],
@@ -143,7 +152,7 @@ class CartesianDomain(Location):
[1.0000, 1.0000]])
"""
def _1d_sampler(n, mode, variables):
""" Sample independentely the variables and cross the results"""
""" Sample independently the variables and cross the results"""
tmp = []
for variable in variables:
if variable in self.range_.keys():