docs/08-lorenz.md

In ppernot/ErrView: Analysis of error sets

Lorenz module

Plot of Lorenz curves of absolute errors and associated statistics [1].

Controls

• Gini vs. LAC: 2D plot of Lorenz Asymmetry coefficients vs. Gini coefficients (experimental)

• Remove glob. outl.: remove global outliers (if any) defined in module Outliers

• Correct trend: remove trend from errors

• Trend degree: degree of the polynomial to fit trend

• Identity grid: plot parallel y=a+x lines to evaluate asymmetry of Lorenz curves

• Normal curve: plot Lorenz curve for half-normal distrubution

• Var-based: variance-based Lorenz curves (experimental)

• Fix colors: use a color palette adapted to the selected methods, or a colr palette based on the full set of methods (if checked)

Methods selector

• Choose metods: a set of checkboxes to select the methods to be plotted.

• Check all: select all methods

• Reset: uncheck all methods except the first one

Plot

Plot of the Lorenz curves or from the 'Gini va. LAC' data.

References

1. P. Pernot and A. Savin (2021) Using the Gini coefficient to characterize the shape of computational chemistry error distributions. arXiv https://arxiv.org/abs/2012.09589

ppernot/ErrView documentation built on Jan. 30, 2022, 6:59 a.m.