Dispersion Function¶

The convex location-indexed functional \(D_X(u)=\mathbb{E}|X-u|\), whose slopes recover an integrable real distribution.

Version: v2 · 2026-09-19 · History
Domain-specific #: 1692
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomains: Probability Theory, Stochastic Orders → Mathematics

Core Idea¶

For an integrable real random variable \(X\), the dispersion function is

\[ D_X(u)=\mathbb{E}|X-u|,\qquad u\in\mathbb{R}. \]

It turns one distribution into an entire location-indexed curve: at every proposed reference point \(u\), the curve reports expected absolute distance from that point. Muñoz-Pérez and Sánchez-Gómez introduced this named functional as a characterization of a distribution and as a basis for comparing dispersion. The crucial identity is stronger than “another variability measure.” Convexity and the one-sided slopes preserve the whole cumulative distribution:

How would you explain it like I'm…

How-Far-From-Here Curve

Imagine your friends standing at different spots along a line. Pick any spot and ask: on average, how far is everyone from here? Do that for every spot on the line and you get a curve. That curve is the dispersion function, and how steep it is at each spot even tells you how many friends are standing on each side of it.

The Average-Distance Curve

In probability, we often think about a random number, like the result of a spinner. The dispersion function picks a point u and asks: on average, how far is the random number from u? It does this for every possible point, making a curve. The curve is lowest near the middle of where the numbers usually land. Amazingly, the shape of this curve holds all the information about the random number: from how steep the curve is at each point, you can work out the chance of landing at or below that point.

Expected Absolute Loss Profile

For a random variable X with a finite mean, the dispersion function is D_X(u) = E|X - u|: the expected absolute distance from X to a reference point u, computed for every real u. So instead of a single spread measure, you get a curve over all locations. The curve is convex, and its slope from the right at u equals 2F_X(u) - 1, where F_X is the cumulative distribution function. That means you can recover the entire distribution from the curve, including jumps where X has positive probability of equaling a single value; at those jumps the curve has a corner, so it isn't differentiable there. The mean absolute deviation around the mean or median is just one point on this curve, not the whole thing. It was introduced by Muñoz-Pérez and Sánchez-Gómez as a way to characterize distributions and compare their spread.

For an integrable real random variable X, the dispersion function is D_X(u) = E|X - u| for u in R, mapping a distribution to a location-indexed curve of expected absolute loss. Muñoz-Pérez and Sánchez-Gómez introduced it as a characterization of the distribution and a basis for dispersion comparisons. D_X is convex, with one-sided derivatives D'_{X,+}(u) = 2F_X(u) - 1 and D'_{X,-}(u) = 2F_X(u-) - 1. Hence F_X(u) = (D'_{X,+}(u) + 1)/2 recovers the full CDF, including atoms; at an atom the left and right slopes differ, so ordinary differentiability should not be claimed there. Its minimizers are the medians, since the slope changes sign where F_X crosses 1/2. The recognizing invariant is the complete absolute-loss profile over all locations, not a single mean absolute deviation evaluated at the mean or median. The functional ties together robust loss, distribution recovery, convex geometry, medians, moments, and stochastic ordering in one object.

Scope of Application¶

The primary scope is univariate probability theory and mathematical statistics for laws with finite first moment. It provides a representation of distributions, a graphical absolute-distance profile, a route to medians, and an object for studying spread comparisons. The original paper explicitly establishes distribution characterization and an induced dispersive ordering.

In stochastic-order theory, the curve belongs among integrated-distribution transforms used to compare laws beyond their means. Shaked and Shanthikumar document stochastic orders as tools for uncertainty comparison and for structural analysis of stochastic systems.

Clarity¶

The name separates three information levels. A mean gives one location number. Mean absolute deviation gives one spread number relative to a selected center. The dispersion function records expected absolute deviation from every center, thereby retaining the entire CDF. This explains why two laws can share mean and mean absolute deviation yet have different dispersion functions.

Manages Complexity¶

A probability law may require a density, a mass function, or a mixture of both. The dispersion function represents all three through one finite convex curve. Its geometry compresses multiple tasks: slopes give probabilities, kinks give atoms, minima give medians, and tail intercepts give the mean. This is useful when convex analysis is more tractable than manipulating a distribution measure directly.

Abstract Reasoning¶

Characterization: equal dispersion functions imply equal one-sided derivatives and hence equal CDFs. Location: any minimizer of a convex \(D_X\) satisfies a zero-in-subgradient condition, which is exactly the median condition \(F_X(u-)\le1/2\le F_X(u)\). Atom detection: \(D'_{+}(u)-D'_{-}(u)=2P(X=u)\). Translation: for \(Y=X+a\), \(D_Y(u)=D_X(u-a)\). Scaling: for real \(b\), \(D_{bX}(u)=|b|D_X(u/b)\) when \(b\ne0\). These relations can be checked without assuming a density.

Knowledge Transfer¶

Within probability and statistics, the full object transfers from distribution characterization to robust location, empirical absolute-loss curves, and stochastic comparison. The equations survive unchanged. A sample version \(n^{-1}\sum_i|X_i-u|\) approximates the population curve and makes the geometry visible, although inferential guarantees need their own assumptions.

Outside those fields, absolute-loss minimization and convexity transfer through existing abstractions such as Expected Value, but “dispersion function” does not. The same phrase also denotes unrelated objects in waves and plasma physics, demonstrating that lexical travel is not structural recurrence.

Relationships to Other Abstractions¶

Current abstraction Dispersion Function Domain-specific

Parents (1) — outgoing typed relations

Dispersion Function presupposes Expected Value Prime

The dispersion function uses prime:expected_value: for each \(u\), it averages the random quantity \(|X-u|\) against the probability measure.

Hierarchy paths (2) — routes to 2 parentless roots

Dispersion Function → presupposes Expected Value → is a kind of Aggregation → is a kind of Transformation → is a kind of, typical Function (Mapping)

Show alternative path (1)

Neighborhood in Abstraction Space¶

Dispersion Function sits in a sparse region of the domain-specific corpus (65^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (2769 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25