Histogram

Shipping

array_histogram

Equal-width histogram of an array: bin counts and bin edges over [min, max]

Signature

Inputs

dataArrayrequired— Any array, flattened to 1-D and binned over its own [min, max].

Outputs

countsVector— Count in each of the `bins` equal-width buckets (length = bins).
edgesVector— The bin edges (length = bins + 1).

Parameters

Key	Type	Default	Notes
`bins`	int	10	Number of equal-width buckets spanning $[\min, \max]$ of the data.

Description

Histogram buckets the flattened input into bins equal-width intervals spanning the data's own $[min, max]$ range and returns two vectors: counts (one value per bin) and edges (the $bins + 1$ bin boundaries). Because the range is derived from the data, the histogram always covers the full extent of the samples.

The edges vector has one more element than counts — bin $k$ spans $[edges_{k}, edges_{k + 1})$ . Downstream you can plot counts against edge midpoints, feed counts into Array Entropy or KL Divergence as an empirical distribution, or find the mode. The node is stateless.

Mathematics

w = \frac{max - min}{bins}, edges_{k} = min + k w (k = 0, \dots, bins)

counts_{k} = {x_{i} : edges_{k} \leq x_{i} < edges_{k + 1}}

Examples

Five-bin histogram

For $[0, 1, 2, 3, 4]$ with bins = 5: the range $[0, 4]$ splits into width- $0.8$ buckets, so edges has 6 entries $[0, 0.8, 1.6, 2.4, 3.2, 4.0]$ and each of the five samples lands in its own bin (counts sum to 5).

Applications

Visualizing a signal's amplitude distribution as a bar chart.
Building an empirical probability mass to feed Entropy / KL Divergence.
Detecting multi-modality or clipping in a data batch.

Neat

counts and edges follow the NumPy contract (edges = bins + 1), so the output drops straight into any histogram plotter or `numpy.histogram`-style consumer.

The auto-derived [min, max] range guarantees no sample falls outside the bins — the histogram is always complete for the input.