cartesian-explorer 0.1.13

Utility to efficiently explore functions on their domains

Cartesian Explorer

A handy tool to map functions over their domains.

Works:

• Map over cartesian product of arguments: pass arrays of values for function argument

• Built-in caching

• Handy plotting utilities

• Resolving dependencies between functions that require and provide variables

Usage

Map

from cartesian_explorer import Explorer

explorer = Explorer()

def my_function(x, y):
    return x+y
data = explorer.map(my_function, x=range(5), y=range(3))
print(data)
assert data.shape == (5, 3)
assert data[1, 2] == my_function(1, 2)

Cache

from cartesian_explorer import Explorer
explorer = Explorer()
mock = MagicMock()
my_function = mock.my_function
wrapped = explorer.cache_function(my_function)
wrapped(a=1, b=2)
wrapped(a=1, b=2)
my_function.assert_called_once_with(a=1, b=2)

TODO

• [ ] Interpolation, extrapolation

• [ ] Vectorized providers

• [x] Plot distribution props along additional distribution var

History

0.1.12 (2020-09-29)

• Row-level variables: now can visualize 4 dimensions

0.1.3 (2020-09-29)

• Joblib disc caching and parallelism

• Dependency graph visualization with networkx

• Plot-level variables

0.0.1 (2020-09-23)

• First release on PyPI.