create_training_matrix: Create a matrix from a vector of text. This function does not...

In jlricon/MaxClassify: An efficient maximum entropy classifier, with some convenience functions

Description

Create a matrix from a vector of text. This function does not do any text cleaning.

Usage

create_training_matrix(text, labels, minDocFreq = 0, maxDocFreq = 1,
  minWordLength = 0, maxWordLength = Inf, weighting = "count",
  ngrams = 1, features = NULL, verbose = FALSE)

Arguments

text	A vector of strings to be made into a matrix
labels	Labels for classification
minDocFreq	Minimum frequency of a word to be included (count or fraction)
maxDocFreq	Maxmimum frequency of a word to be included (count or fraction)
minWordLength	Minimum wordlength for a word to be included
maxWordLength	Maximum wordlength for a word to be included
weighting	Weighting to use. Choose between frequency, relFreq, relMaxFreq, logFreq, tfidf (See dfm_weight)
ngrams	Vector that specifies type of ngrams to generate

Value

A document feature matrix

jlricon/MaxClassify documentation built on May 29, 2019, 2:56 p.m.