Package: fastmit 0.1.1

Shiyun Lin

fastmit: Fast Mutual Information Based Independence Test

A mutual information estimator based on k-nearest neighbor method proposed by A. Kraskov, et al. (2004) <doi:10.1103/PhysRevE.69.066138> to measure general dependence and the time complexity for our estimator is only squared to the sample size, which is faster than other statistics. Besides, an implementation of mutual information based independence test is provided for analyzing multivariate data in Euclidean space (T B. Berrett, et al. (2019) <doi:10.1093/biomet/asz024>); furthermore, we extend it to tackle datasets in metric spaces.

Authors:Shiyun Lin [aut, cre], Jin Zhu [aut], Wenliang Pan [aut], Xueqin Wang [aut], SC2S2 [cph]

fastmit_0.1.1.tar.gz
fastmit_0.1.1.tar.gz(r-4.5-noble)fastmit_0.1.1.tar.gz(r-4.4-noble)
fastmit_0.1.1.tgz(r-4.4-emscripten)fastmit_0.1.1.tgz(r-4.3-emscripten)
fastmit.pdf |fastmit.html✨
fastmit/json (API)
NEWS

# Install 'fastmit' in R:
install.packages('fastmit', repos = 'https://cloud.r-project.org')
Uses libs:
  • c++– GNU Standard C++ Library v3

On CRAN:

Conda:

This package does not link to any Github/Gitlab/R-forge repository. No issue tracker or development information is available.

cpp

1.70 score 149 downloads 2 exports 2 dependencies

Last updated 5 years agofrom:4adb14365d. Checks:3 OK. Indexed: yes.

TargetResultLatest binary
Doc / VignettesOKMar 26 2025
R-4.5-linux-x86_64OKMar 26 2025
R-4.4-linux-x86_64OKMar 26 2025

Exports:mimi.test

Dependencies:RcppRcppArmadillo

Citation

To cite package ‘fastmit’ in publications use:

Lin S, Zhu J, Pan W, Wang X (2019). fastmit: Fast Mutual Information Based Independence Test. R package version 0.1.1, https://CRAN.R-project.org/package=fastmit.

Corresponding BibTeX entry: 

  @Manual{,
    title = {fastmit: Fast Mutual Information Based Independence Test},
    author = {Shiyun Lin and Jin Zhu and Wenliang Pan and Xueqin Wang},
    year = {2019},
    note = {R package version 0.1.1},
    url = {https://CRAN.R-project.org/package=fastmit},
  }

Readme and manuals

Fast Mutual Information based independence Test (fastmit)

Introduction

The fundamental problem for data mining and statistical analysis is:

  • Whether two random variables are dependent?

fastmit package provides solutions for this issue. It implements the kNN method described by Kraskov, et. al (2004) to estimate the empirical mutual information and furthermore uses permutation test to detect whether two random variables are independent. The core functions in fastmit package are mi and mi.test.

These functions based on mutual information have two main advantages:

  • It's applicable to complex data in metric spaces.

  • It is faster than other statistics (e.g., distance covariance and ball covariance), which makes it advantageous in large sample situations.

Installation

CRAN version

You can install the released version of fastmit from CRAN with:

install.packages("fastmit")
Github version

You can install the released version of fastmit from GitHub with:

library(devtools)
install_github("Mamba413/fastmit/R-package")

Windows user will need to install Rtools first.

Example

  • mi function
## simulate data
set.seed(1)
x <- rnorm(100)
y <- x + rnorm(100)
## estimate the empirical mutual information
mi(x, y)

In this example, the result is:

# [1] 0.3320034
  • mi.test function
## simulate data
set.seed(1)
error <- runif(50, min = -0.3, max = 0.3)
x <- runif(50, 0, 4*pi)
y <- cos(x) + error

## perform independence test via mutual information
mi.test(x, y)

In this example, the result is:

	Mutual Information test of independence

data:  x and y
number of observations = 50
replicates = 99
p-value = 0.01
alternative hypothesis: random variables are dependent
sample estimates:
       MI 
0.6953105

If you find any bugs, or if you experience any crashes, please report to us. Also, if you have any questions, feel free to ask.

License

GPL-3

Reference