batch-description.text

Batch statistics provides descriptive statistical functions for fully-realised
sequences and arrays held in memory.  It implements the standard suite of
summary statistics — mean, variance, standard deviation, quantiles, median,
mode, and modes — as methods on Common Lisp sequence types, extending the
STAT-GENERICS generic function layer.  All moment calculations support
Julia-style weight objects: FREQUENCY-WEIGHTS for repeated observations,
ANALYTIC-WEIGHTS for inverse-variance weighting, and PROBABILITY-WEIGHTS for
survey sampling, each applying the appropriate bias correction denominator.
Quantile computation uses a 0.5-correction empirical algorithm consistent with
the online subsystem, ensuring identical results for the same data regardless
of which subsystem is used.  Higher-level summaries include INTERQUARTILE-RANGE,
FIVENUM for five-number summaries with optional Tukey hinges, and SCALE for
mean-centering and standardisation in the style of R's scale().