Calculate Simpson's diversity, Gini-Simpson index, ENS, and evenness

Computes four related measures from Simpson's D (sum of squared relative abundances). Note that GINI_SIMPSON (1 - D) is a diversity index, not an evenness measure — it is the probability that two randomly drawn individuals belong to different species. True evenness is derived from the number of species (SIMPSON_ENS).

Usage

get_alpha_simpson(df, feature_id = "feature_id", id_col = "SAMPLE")

Arguments

df

A dataframe with features as rows and samples as columns

feature_id

The name of the column with feature IDs (default is "feature_id")

id_col

The name of the column with sample IDs (default is "SAMPLE")

Value

A tibble with one row per sample and columns:

SIMPSON_D

Simpson's dominance index (sum of squared proportions); lower = more diverse

GINI_SIMPSON

Gini-Simpson index (1 - D); a diversity index, not evenness

SIMPSON_ENS

Effective number of species (1 / D); the q=2 Hill number; interpretable as the number of equally abundant dominant species that would yield the same D

SIMPSON_ENS_EVENNESS

ENS-based evenness (SIMPSON_ENS / S.obs); ranges from 1/S.obs (single dominant) to 1 (all species equal)

Breaking changes

SIMPSON_EVENNESS

Previously 1 - SIMPSON_D (Gini-Simpson index). Renamed to GINI_SIMPSON. The name SIMPSON_EVENNESS is no longer returned — old workflows using it will error rather than silently return wrong values.

SIMPSON_INVERSE

Renamed to SIMPSON_ENS. Old workflows using SIMPSON_INVERSE will error.