counts: Utilities for count matrices

In maptpx: MAP Estimation of Topic Models

Description

Tools for manipulating (sparse) count matrices.

Usage

normalize(x,byrow=TRUE)
stm_tfidf(x)

Arguments

x	A simple_triplet_matrix or matrix of counts.
byrow	Whether to normalize by row or column totals.

Value

normalize divides the counts by row or column totals, and stm_tfidf returns a matrix with entries x_{ij} \log[ n/(d_j+1) ], where x_{ij} is term-j frequency in document-i, and d_j is the number of documents containing term-j.

Author(s)

Matt Taddy [e-mail]

Examples

normalize( matrix(1:9, ncol=3) )
normalize( matrix(1:9, ncol=3), byrow=FALSE )

(x <- matrix(rbinom(15,size=2,prob=.25),ncol=3))
stm_tfidf(x)

maptpx documentation built on July 1, 2020, 10:35 p.m.