scale.sento_measures: Scaling and centering of sentiment measures

In sentometrics: An Integrated Framework for Textual Sentiment Time Series Aggregation and Prediction

Description

Scales and centers the sentiment measures from a sento_measures object, column-per-column. By default, the measures are normalized. NAs are removed first.

Usage

## S3 method for class 'sento_measures'
scale(x, center = TRUE, scale = TRUE)

Arguments

x	a sento_measures object created using sento_measures.
center	a logical or a numeric vector, see documentation for the generic scale. Alternatively, one can provide a matrix of dimensions nobs(sento_measures) times 1 or nmeasures(sento_measures) with values to subtract from each individual observation.
scale	a logical or a numeric vector, see documentation for the generic scale. Alternatively, one can provide a matrix of dimensions nobs(sento_measures) times 1 or nmeasures(sento_measures) with values to divide each individual observation by.

Details

If one of the arguments center or scale is a matrix, this operation will be applied first, and eventual other centering or scaling is computed on that data.

Value

A modified sento_measures object, with the measures replaced by the scaled measures as well as updated statistics.

Author(s)

Samuel Borms

Examples

data("usnews", package = "sentometrics")
data("list_lexicons", package = "sentometrics")
data("list_valence_shifters", package = "sentometrics")

set.seed(505)

# construct a sento_measures object to start with
corpus <- sento_corpus(corpusdf = usnews)
corpusSample <- quanteda::corpus_sample(corpus, size = 500)
l <- sento_lexicons(list_lexicons[c("LM_en", "HENRY_en")])
ctr <- ctr_agg(howTime = c("equal_weight", "linear"), by = "year", lag = 3)
sento_measures <- sento_measures(corpusSample, l, ctr)

# scale sentiment measures to zero mean and unit standard deviation
sc1 <- scale(sento_measures)

n <- nobs(sento_measures)
m <- nmeasures(sento_measures)

# subtract a matrix
sc2 <- scale(sento_measures, center = matrix(runif(n * m), n, m), scale = FALSE)

# divide every row observation based on a one-column matrix, then center
sc3 <- scale(sento_measures, center = TRUE, scale = matrix(runif(n)))

sentometrics documentation built on Aug. 18, 2021, 9:06 a.m.