Describe › Standalone plots

Histogram

A histogram shows the distribution of a single numeric column, binned by the Freedman-Diaconis rule (2·IQR / n^{1/3}) unless you set the bin count yourself.

What is Histogram?

Histograms are the workhorse visualisation for univariate continuous data — they show centre, spread, skew, and modality at a glance. Bin choice is the only real lever: too few bins and you smooth over real structure (bimodality, gaps); too many and you over-fit to noise.

Freedman-Diaconis (default) sets bin width = 2 · IQR / n^{1/3}. It's robust to outliers and adapts to the data's spread. Scott's rule (3.5 · SD / n^{1/3}) assumes normality and is wider; Sturges' (log₂ n + 1) is older and tends to under-bin for n > 100. We expose a manual bin-count override for when none of these gets the right shape.

When should I use Histogram?

First look at a continuous variable to see its shape.
Before normality testing — visual gestalt is faster than W and tells you why.
With "Cumulative" mode for an empirical CDF (and an inline comparison against a theoretical CDF).

What data does it need?

One numeric column + optional bin count + optional cumulative-mode toggle.

What does it report?

Bar chart of counts per bin (or running proportion in cumulative mode).

How do I interpret the result?

Multiple peaks ⇒ mixture of subpopulations — split by the relevant categorical variable before testing.

Long right tail ⇒ try log-transform; long left tail ⇒ try reflecting and log-transforming, or use a non-parametric test.

Single bin with most of the data + sparse tail ⇒ likely a coding issue (e.g. "99 = missing" left in as a numeric value).