Learn R Programming

sortinghat (version 0.1)

sortinghat

Description

sortinghat is a classification framework to streamline the evaluation of classifiers (classification models and algorithms) and seeks to determine the best classifiers on a variety of simulated and benchmark data sets. Several error-rate estimators are included to evaluate the performance of a classifier. This package is intended to complement the well-known 'caret' package.

Copy Link

Version

Install

install.packages('sortinghat')

Monthly Downloads

1

Version

0.1

License

MIT + file LICENSE

Issues

Pull Requests

Stars

Forks

Repository

http://github.com/ramhiser/sortinghat

Maintainer

John Ramey

Last Published

December 7th, 2013

Functions in sortinghat (0.1)

Constructs a p-dimensional intraclass covariance matrix.

Calculates the Cross-Validation Error Rate for a specified classifier given a data set.

cov_block_autocorrelation

Generates a p-dimensional block-diagonal covariance matrix with autocorrelated blocks.

Wrapper function to estimate the error rate of a classifier

Partitions data for cross-validation.

Helper function that partitions a data set into training and test data sets.

cov_autocorrelation

Constructs a p-dimensional covariance matrix with an autocorrelation (autoregressive) structure.

check_arguments

Checks the arguments passed to the error rate estimator functions.

Calculates the Bootstrap Error Rate for a specified classifier given a data set.

simdata_contaminated

Generates random variates from K multivariate contaminated normal populations.

Calculates the Bootstrap Cross-Validation (BCV) Error Rate Estimator for a specified classifier given a data set.

errorest_loo_boot

Calculates the Leave-One-Out (LOO) Bootstrap Error Rate for a specified classifier given a data set.

Generates random variates from K multivariate normal populations.

Wrapper function to generate data from a variety of data-generating families for classification studies.

Generates data from K multivariate normal data populations having the covariance structure from Guo et al. (2007).

errorest_632plus

Calculates the .632+ Error Rate for a specified classifier given a data set.

Helper function that determines which element in a vector is the minimum. Ties can be broken randomly or via first/last ordering.

simdata_friedman

Generates data from 3 multivariate normal data populations having the covariance structure from Friedman (1989).

Function to check whether all elements in a numeric vector are equal within some tolerance

errorest_apparent

Calculates the Apparent Error Rate for a specified classifier given a data set.

simdata_uniform

Generates random variates from multivariate uniform populations.

Calculates the .632 Error Rate for a specified classifier given a data set.

Generates random variates from K multivariate Student's t populations.