Update documentation

Cargo.toml

@@ -4,10 +4,11 @@ name = "average"
 version = "0.5.0"
 license = "MIT/Apache-2.0"
 repository = "https://github.com/vks/average"
-description = "Calculate the average of a sequence and its error iteratively"
+description = "Calculate statistics iteratively"
 readme = "README.md"
 categories = ["science", "no-std"]
-keywords = ["statistics", "stats"]
+keywords = ["statistics", "stats", "mean", "variance", "skewness",
+            "kurtosis", "quantile"]
 
 [dependencies]
 conv = "0.3"

README.md

@@ -1,7 +1,7 @@
 # average
 
-Calculate the average of a sequence and its error iteratively in a single pass,
+Calculate statistics of a sequence iteratively in a single pass, using
-using constant memory and avoiding numerical problems. The calculation can be
+constant memory and avoiding numerical problems. The calculations can be
 easily parallelized by using `merge`.
 
 [Documentation](https://docs.rs/average) |
@@ -9,7 +9,9 @@ easily parallelized by using `merge`.
 
 [![Build Status](https://travis-ci.org/vks/average.svg?branch=master)](https://travis-ci.org/vks/average)
 
-## Advantages over naive calculation of average and variance
+## Implemented statistics
 
-* Avoids loss of precision due to cancellation.
+* Mean and its error.
-* Only needs a single pass over the samples, at the cost of a division inside the loop.
+* Variance, skewness, kurtosis.
+* Minium and maximum.
+* Quantile.

src/lib.rs

@@ -1,14 +1,11 @@
-//! This crate provides estimators for the weighted and unweighted average of a
+//! This crate provides estimators for statistics on a sequence of numbers. The
-//! sequence of numbers, and for their standard errors. The typical workflow
+//! typical workflow looks like this:
-//! looks like this:
 //!
-//! 1. Initialize your estimator of choice ([`Mean`], [`MeanWithError`],
+//! 1. Initialize the estimator of your choice with `new()`.
-//!    [`WeightedMean`] or [`WeightedMeanWithError`]) with `new()`.
+//! 2. Add some subset (called "sample") of the sequence of numbers (called
-//! 2. Add some subset (called "samples") of the sequence of numbers (called
+//!    "population") for which you want to estimate the statistic, using `add()`
-//!    "population") for which you want to estimate the average, using `add()`
 //!    or `collect()`.
-//! 3. Calculate the arithmetic mean with `mean()` and its standard error with
+//! 3. Calculate the statistic with `mean()` or similar.
-//!    `error()`.
 //!
 //! You can run several estimators in parallel and merge them into one with
 //! `merge()`.
@@ -17,10 +14,38 @@
 //! so the sequence of numbers can be an iterator. The used algorithms try to
 //! avoid numerical instabilities.
 //!
-//! [`Mean`]: ./average/struct.Mean.html
+//!
-//! [`MeanWithError`]: ./average/struct.MeanWithError.html
+//! ## Estimators
-//! [`WeightedMean`]: ./weighted_average/struct.WeightedMean.html
+//!
-//! [`WeightedMeanWithError`]: ./weighted_average/struct.WeightedMeanWithError.html
+//! * Mean ([`Mean`]) and its error ([`MeanWithError`]).
+//! * Weighted mean ([`WeightedMean`]) and its error
+//!   ([`WeightedMeanWithError`]).
+//! * Variance ([`Variance`]), skewness ([`Skewness`]) and kurtosis
+//!   ([`Kurtosis`]).
+//! * Quantiles ([`Quantile`]).
+//! * Minimum ([`Min`]) and maximum ([`Max`]).
+//!
+//! [`Mean`]: ./struct.Mean.html
+//! [`MeanWithError`]: ./type.MeanWithError.html
+//! [`WeightedMean`]: ./struct.WeightedMean.html
+//! [`WeightedMeanWithError`]: ./struct.WeightedMeanWithError.html
+//! [`Variance`]: ./struct.Variance.html
+//! [`Skewness`]: ./struct.Skewness.html
+//! [`Kurtosis`]: ./struct.Kurtosis.html
+//! [`Quantile`]: ./struct.Quantile.html
+//! [`Min`]: ./struct.Min.html
+//! [`Max`]: ./struct.Max.html
+//!
+//!
+//! ## Estimating several statistics at once
+//!
+//! The estimators are designed to have minimal state. The recommended way to
+//! calculate several of them at once is to create a struct with all the
+//! estimators you need. You can then implement `add` for your struct by
+//! forwarding to the underlying estimators.
+//!
+//! Note that calculating moments requires calculating the lower moments, so you
+//! only need to include the highest moment in your struct.
 //!
 //!
 //! ## Example

src/moments/mod.rs

@@ -3,4 +3,5 @@ include!("variance.rs");
 include!("skewness.rs");
 include!("kurtosis.rs");
 
+/// Alias for `Variance`.
 pub type MeanWithError = Variance;