Mentions légales du service

Skip to content
Snippets Groups Projects
Commit f45fd8f7 authored by hhakim's avatar hhakim
Browse files

Update doc of pyfaust.lazylinop.

parent 92e4dec0
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
wrapper/python/pyfaust/lazylinop.py
+ 54
− 69
View file @ f45fd8f7
......@@ -51,7 +51,7 @@ def aslazylinearoperator(obj):
def hstack(tup):
"""
Concatenates lop1 and obj horizontally.
Concatenates tup objects horizontally.
Args:
tup: a tuple whose first argument is a LazyLinearOp and other must be
......@@ -68,7 +68,7 @@ def hstack(tup):
def vstack(tup):
"""
Concatenates lop1 and obj vertically.
Concatenates tup objects vertically.
Args:
tup: a tuple whose first argument is a LazyLinearOp and other must be
......@@ -90,19 +90,21 @@ class LazyLinearOp(LinearOperator):
href="https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.linalg.LinearOperator.html">scipy.linalg.LinearOperator</a>.
The evaluation of any defined operation is delayed until a multiplication
by a dense matrix/vector, a call of LazyLinearOp.toarray or an explicit
evaluation through LazyLinearOp.eval.
by a dense matrix/vector, a call of LazyLinearOp.toarray.
To instantiate a LazyLinearOp look at pyfaust.lazylinop.asLazyLinearOp.
To instantiate a LazyLinearOp look at
pyfaust.lazylinop.aslazylinearoperator or
pyfaust.lazylinop.LazyLinearOperator to instantiate from matmat/matvec
functions.
Warning: This code is in a beta status.
"""
def __init__(self, lambdas, shape, root_obj, dtype=None):
def __init__(self, lambdas, shape, root_obj=None, dtype=None):
"""
Constructor. Not meant to be used directly.
Args:
init_lambda: starting operation.
lambdas: starting operations.
shape: the initial shape of the operator.
root_obj: the initial object the operator is based on.
......@@ -119,8 +121,7 @@ class LazyLinearOp(LinearOperator):
self.lambdas = lambdas
self._check_lambdas()
self.shape = shape
self._root_obj = root_obj #TODO: delete because it can't always be
# defined (create_from_funcs and hybrid operations)
self._root_obj = root_obj
self.dtype = dtype
super(LazyLinearOp, self).__init__(self.dtype, self.shape)
......@@ -136,31 +137,7 @@ class LazyLinearOp(LinearOperator):
@staticmethod
def create_from_op(obj):
"""
Alias of pyfaust.lazylinop.asLazyLinearOp.
Args:
obj: cf. pyfaust.lazylinop.asLazyLinearOp.
Returns:
cf. pyfaust.lazylinop.asLazyLinearOp.
Example:
>>> from pyfaust.lazylinop import LazyLinearOp
>>> import numpy as np
>>> M = np.random.rand(10, 12)
>>> lM = LazyLinearOp.create(M)
>>> twolM = lM + lM
>>> twolM
<pyfaust.lazylinop.LazyLinearOp at 0x7fcd7d7750f0>
>>> import pyfaust as pf
>>> F = pf.rand(10, 12)
>>> lF = LazyLinearOp.create(F)
>>> twolF = lF + lF
>>> twolF
<pyfaust.lazylinop.LazyLinearOp at 0x7fcd7d774730>
<b>See also:</b> pyfaust.rand.
Alias of pyfaust.lazylinop.aslazylinearoperator.
"""
lambdas = {'@': lambda op: obj @ op}
lambdasT = {'@': lambda op: obj.T @ op}
......@@ -194,6 +171,9 @@ class LazyLinearOp(LinearOperator):
@staticmethod
def create_from_scalar(s, shape=None):
"""
Returns a LazyLinearOp L created from the scalar s, such as each L[i, i] == s.
"""
if shape is None:
shape = (1, 1)
a = np.ones(shape) * s
......@@ -201,6 +181,9 @@ class LazyLinearOp(LinearOperator):
@staticmethod
def create_from_funcs(matmat, rmatmat, shape, dtype=None):
"""
Alias of ctor pyfaust.lazylinop.LazyLinearOperator.
"""
from scipy.sparse import eye as seye
#MX = lambda X: matmat(np.eye(shape[1])) @ X
......@@ -256,7 +239,7 @@ class LazyLinearOp(LinearOperator):
@property
def ndim(self):
"""
Returns the number of dimensions of the LazyLinearOp (it always 2).
Returns the number of dimensions of the LazyLinearOp (it is always 2).
"""
return 2
......@@ -459,7 +442,7 @@ class LazyLinearOp(LinearOperator):
def __matmul__(self, op):
"""
Returns self @ op as a sparse matrix / dense array or as a LazyLinearOp.
Computes self @ op as a sparse matrix / dense array or as a LazyLinearOp depending on the type of op.
Args:
op: an object compatible with self for this binary operation.
......@@ -503,15 +486,12 @@ class LazyLinearOp(LinearOperator):
def matvec(self, op):
"""
Returns the LazyLinearOp for the multiplication self.matvec(op) or the np.ndarray
resulting of the evaluation of self.matvec(op) if op is a np.ndarray.
This function is an alias of self @ op, where the multiplication might
be specialized for op a vector (depending on how self has been defined
; upon on a operator object or through a matvec/matmat function).
Args:
op: must be a vector (row or column).
<b>See also</b>: LazyLinearOp.__matmul__,<a
href="https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.linalg.LinearOperator.matvec.html#scipy.sparse.linalg.LinearOperator.matvec">
scipy.linalg.LinearOperator.matvec</a>
<b>See also:</b> pyfaust.lazylinop.LazyLinearOperator.
"""
if not hasattr(op, 'shape') or not hasattr(op, 'ndim'):
raise TypeError('op must have shape and ndim attributes')
......@@ -524,7 +504,11 @@ class LazyLinearOp(LinearOperator):
def _rmatvec(self, op):
"""
Returns the LazyLinearOp for self^H @ v, where self^H is the conjugate transpose of A.
Returns self^H @ op, where self^H is the conjugate transpose of A.
Returns:
It might be a LazyLinearOp or an array depending on the op type
(cf. pyfaust.lazylinop.LazyLinearOp.__matmul__).
"""
# LinearOperator need.
return self.T.conj() @ op
......@@ -542,7 +526,11 @@ class LazyLinearOp(LinearOperator):
def _rmatmat(self, op):
"""
Returns the LazyLinearOp for self^H @ v, where self^H is the conjugate transpose of A.
Returns self^H @ op, where self^H is the conjugate transpose of A.
Returns:
It might be a LazyLinearOp or an array depending on the op type
(cf. pyfaust.lazylinop.LazyLinearOp.__matmul__).
"""
# LinearOperator need.
return self.T.conj() @ op
......@@ -555,11 +543,12 @@ class LazyLinearOp(LinearOperator):
def __rmatmul__(self, op):
"""
Returns the LazyLinearOp for op @ self.
Returns op @ self which can be a LazyLinearOp or an array depending on op type.
Args:
op: an object compatible with self for this binary operation.
<b>See also:</b> pyfaust.lazylinop.LazyLinearOp.__matmul__)
"""
self._checkattr('__rmatmul__')
from scipy.sparse import issparse
......@@ -581,11 +570,13 @@ class LazyLinearOp(LinearOperator):
def __mul__(self, other):
"""
Returns the LazyLinearOp for self * other.
Returns the LazyLinearOp for self * other if other is a scalar
otherwise returns self @ other.
Args:
other: a scalar or a vector.
other: a scalar or a vector/array.
<b>See also:</b> pyfaust.lazylinop.LazyLinearOp.__matmul__)
"""
self._checkattr('__mul__')
from scipy.sparse import eye, diags
......@@ -597,20 +588,18 @@ class LazyLinearOp(LinearOperator):
new_op = self @ other
return new_op
def __rmul__(self, s):
def __rmul__(self, other):
"""
Returns the LazyLinearOp for op * self.
Returns other * self.
Args:
s: a scalar.
other: a scalar or a vector/array.
"""
# because s is a scalar, it is commutative
# for vector broadcasting too
if np.isscalar(s):
return self * s
if np.isscalar(other):
return self * other
else:
return s @ self
return other @ self
def __imul__(self, op):
......@@ -621,7 +610,7 @@ class LazyLinearOp(LinearOperator):
def toarray(self):
"""
Returns the numpy array resulting from self evaluation.
Returns self as a numpy array.
"""
#from scipy.sparse import eye
#return self @ eye(self.shape[1], self.shape[1], format='csr')
......@@ -635,7 +624,8 @@ class LazyLinearOp(LinearOperator):
Args:
indices: array of length 1 or 2 which elements must be slice, integer or
Ellipsis (...). Note that using Ellipsis for more than two indices is forbidden.
Ellipsis (...). Note that using Ellipsis for more than two indices
is normally forbidden.
"""
self._checkattr('__getitem__')
......@@ -777,6 +767,9 @@ class LazyLinearOp(LinearOperator):
def vstack(self, op):
"""
See pyfaust.lazylinop.vstack.
"""
if self.shape[1] != op.shape[1]:
raise ValueError('The number of columns of self and op are not the'
' same')
......@@ -825,6 +818,9 @@ class LazyLinearOp(LinearOperator):
+ op @ o._slice((slice(self.shape[1], o.shape[0]), slice(0, o.shape[1])))
def hstack(self, op):
"""
See pyfaust.lazylinop.hstack.
"""
if self.shape[0] != op.shape[0]:
raise ValueError('The number of columns of self and op are not the'
' same')
......@@ -1004,17 +1000,6 @@ def kron(A, B):
<b>See also:</b> numpy.kron.
"""
def _kron(A, B, shape, op):
"""
Returns the LazyLinearOpKron for the multiplication self @ op or if op is a np.ndarray it returns the np.ndarray (self @ op).
Note: this specialization is particularly optimized for multiplying the
operator by a vector.
Args:
op: an object compatible with self for this binary operation.
<b>See also:</b> pyfaust.lazylinop.kron.
"""
from threading import Thread
from multiprocessing import cpu_count
from os import environ
......
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment