__all__ = ["MemoizeOperator"]
import numpy as np
from pylops import LinearOperator
from pylops.utils.typing import NDArray
[docs]
class MemoizeOperator(LinearOperator):
r"""Memoize Operator.
This operator can be used to wrap any PyLops operator and add a memoize
functionality
and stores the last ``max_neval`` model/data
vector pairs
Parameters
----------
Op : :obj:`pylops.LinearOperator`
PyLops linear operator
max_neval : :obj:`int`, optional
Maximum number of previous evaluations stored,
use ``np.inf`` for infinite memory
Attributes
----------
store : :obj:`list`
List of stored (x, y) pairs.
neval : :obj:`int`
Number of evaluations of the operator.
shape : :obj:`tuple`
Operator shape. :math:`[n \times m]`.
"""
def __init__(
self,
Op: LinearOperator,
max_neval: int = 10,
) -> None:
super().__init__(Op=Op)
self.max_neval = max_neval
self.store: list[tuple[NDArray, NDArray]] = [] # Store a list of (x, y)
self.neval = 0 # Number of evaluations of the operator
def _matvec(self, x: NDArray) -> NDArray:
for xstored, ystored in self.store:
if np.allclose(xstored, x):
return ystored
if len(self.store) + 1 > self.max_neval:
del self.store[0] # Delete oldest
y = self.Op._matvec(x)
self.neval += 1
self.store.append((x.copy(), y.copy()))
return y
def _rmatvec(self, y: NDArray) -> NDArray:
for xstored, ystored in self.store:
if np.allclose(ystored, y):
return xstored
if len(self.store) + 1 > self.max_neval:
del self.store[0] # Delete oldest
x = self.Op._rmatvec(y)
self.neval += 1
self.store.append((x.copy(), y.copy()))
return x
PyLops