gelman

Welcome to Gelman! 📊📈📉

Gelman is a statistical library written entirely in 🌟 Gleam 🌟. It is named after the noted Bayesian statistician Andrew Gelman, whose name happens to be a near-anagram of Gleam.

Wherever possible, Gelman is:

🥚 Simple to use, with a consistent interface. A dataset is a list of floats, so any summary or frequency statistics are returned as floats. If you have any integers, please convert them first.
⭐ Pure Gleam, and purely functional. Wherever possible, the fold combinator is used. map followed by fold is avoided as to reduce memory overhead.
🦦 Efficient. Gelman endeavors to perform one-pass over the data, even for higher order moments like skewness and kurtosis. If sorting is required, Gelman sorts only once, unless absolutely necessary.
🧪 Extensively tested. Tests are borrowed from scipy/stats, and so the results are guaranteed to be at least as accurate.

Gelman functions are grouped according to their purpose.

Module	Contains
`summary`	Summary statistics of a sample dataset, such as `mean`, `variance`. `interquartile_range`. These typically take in a list of values and return one single value.
`transform`	Applies a transformation to the entire dataset. Attention: these can either preserve the size of the dataset, or they can drop some elements.
`discretize`	Produce a range of values that describe its frequences.

gleam add gelman@1

import gelman/summarize
import gelman/transform
import gelman/discretize 

pub fn main() -> Nil {
  let dataset = [0.0, 50.0, 100.0]
  let average = summarize.mean(dataset)
  let quantiles = 
    dataset
    |> discretize.quantiles([0.25, 0.5, 0.75])
  let windsorized_values = 
    dataset
    |> transform.winsorize(0.05, 0.95)
  let 
}

Further documentation can be found at https://hexdocs.pm/gelman.

Development

Currently, the library has most of the functions available in scipy/stats, for summary, descriptive and frequency statistics. I will implement a few more:

discretize: histogram, cumulative_frequency, counts
transform: trim, rank
summary: standard_error_of_mean

I am also planning on implementing another module, test, which will contain statistical tests. Currently, there is no unified mathematics library which offers all the functions required to perform parametric tests. Until I can can figure out how to either implement these functions or augment these libraries, only nonparametric tests can be performed.

gleam run   # Run the project
gleam test  # Run the tests