xyzt 0.0.1

Compute and plot timing scalings for functions with numpy array inputs

xyzt: compute and plot time cost scaling for functions with numpy array input

I often implement functions on atom positions in terms of Nx3 numpy arrays, and need to compare speed of different implementations. This small library is a thin wrapper over timeit for that purpose.

xyzt.timethem(functions_to_check, nmin=10, nmax=1e5, number=1e5) -> (n, times)
xyzt.plot(n, times, labels=None)

for fi in functions:
    for n in valid_range:
        generate random n-dimensional input vector
        timeit fi(v_rand)

example: scipy.spatial.distance.cdist+min versus scipy.spatial.cKDTree