dist_kde

dist_kde(data, kernel='Gaussian', bandwidth=None, bounded_by_data=True)

Kernel density distribution.

A nonparametric density estimate built from a sample by summing a kernel centered at each observation. Mirrors the C# KernelDensity composite.

Parameters

Name Type Description Default
data array-like of float Observations the density is built from. required
kernel ('Gaussian', 'Epanechnikov', 'Triangular', 'Uniform') The kernel shape. "Gaussian"
bandwidth float The kernel bandwidth; None (the default) uses Silverman’s rule of thumb. None
bounded_by_data bool Whether the reported minimum and maximum are the smallest and largest observation (True, the default) or extend three bandwidths past each. Those bounds gate Distribution.cdf() and Distribution.quantile(). The density is summed wherever you ask it, either way. True

Returns

Name Type Description
Distribution Family "KernelDensity", accepted by every method every other :class:Distribution accepts.

Examples

>>> d = dist_kde([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
>>> d.cdf(5.5)
0.5