A small package with fast numpy routines written in cython
pip install numpyx
All functions here are specialized for double arrays only
These functions are similar to numpy functions but are faster by exiting out of a loop when one element satisfies the given condition
- any_less_than
- any_less_or_equal_than
- any_greater_than
- any_greater_or_equal_than
- any_equal_to
- array_is_sorted
- allequal
Calculate min. and max. value in one go
like search sorted, but for 1d double arrays. It is faster than the more generic numpy version
like search sorted but allows to search across any column of a 2d array
Return the index of the item in an array which is nearest to a given value. The array does not need to be sorted (this is a simple linear search)
For any value of an array, search the nearest item in another array and put its value in the output result
Weighted averageof a time-series
trapz integration specialized for contiguous / double arrays. Quite faster than generic numpy/scipy