languageR (version 1.5.0)

spanishFunctionWords: Relative frequencies of function words in selected Spanish texts

Description

Relative frequencies of the 120 most frequent function words in 15 texts contributed by 3 authors.

Usage

data(spanishFunctionWords)

Format

A data frame with 120 observations on 15 variables documented in spanishMeta.

References

Spassova, M. S. (2006) Las marcas sintacticas de atribucion forense de autoria de textos escritos en espanol, Masters thesis, Institut Universitari de Linguistica Aplicada, Universitat Pompeu Fabra, Barcelona.

Examples

# NOT RUN {
data(spanishFunctionWords)
data(spanishMeta)

# principal components analysis

spanishFunctionWords.t = t(spanishFunctionWords)
spanishFunctionWords.t = 
  spanishFunctionWords.t[order(rownames(spanishFunctionWords.t)), ]
spanishFunctionWords.pca = 
  prcomp(spanishFunctionWords.t, center = TRUE, scale = TRUE)

sdevs = spanishFunctionWords.pca$sdev^2
n = sum(sdevs/sum(sdevs)> 0.05) 

# linear discriminant analysis with cross-validation

library(MASS)

predictedClasses = rep("", 15)
for (i in 1:15) {
  training = spanishFunctionWords.t[-i,]                   
  trainingAuthor = spanishMeta[-i,]$Author
  training.pca = prcomp(training, center = TRUE, scale = TRUE)
  training.x = data.frame(training.pca$x)
  training.x = training.x[order(rownames(training.x)), ]
  training.pca.lda = lda(training[ , 1:n], trainingAuthor)
  cl = predict(training.pca.lda, spanishFunctionWords.t[,1:n])$class[i]
  predictedClasses[i] = as.character(cl)
}

ncorrect = sum(predictedClasses==spanishMeta$Author)
sum(dbinom(ncorrect:15, 15, 1/3))
# }