chunker: Break Text Into Ordered Word Chunks

In trinker/qdap: Bridging the Gap Between Qualitative Data and Quantitative Analysis

Description

Some visualizations and algorithms require text to be broken into chunks of ordered words. chunker breaks text, optionally by grouping variables, into equal chunks. The chunk size can be specified by giving number of words to be in each chunk or the number of chunks.

Usage

chunker(
  text.var,
  grouping.var = NULL,
  n.words,
  n.chunks,
  as.string = TRUE,
  rm.unequal = FALSE
)

Arguments

text.var	The text variable
grouping.var	The grouping variables. Default NULL generates one word list for all text. Also takes a single grouping variable or a list of 1 or more grouping variables.
n.words	An integer specifying the number of words in each chunk (must specify n.chunks or n.words).
n.chunks	An integer specifying the number of chunks (must specify n.chunks or n.words).
as.string	logical. If TRUE the chunks are returned as a single string. If FALSE the chunks are returned as a vector of single words.
rm.unequal	logical. If TRUE final chunks that are unequal in length to the other chunks are removed.

Value

Returns a list of text chunks.

Examples

with(DATA, chunker(state, n.chunks = 10))
with(DATA, chunker(state, n.words = 10))
with(DATA, chunker(state, n.chunks = 10, as.string=FALSE))
with(DATA, chunker(state, n.chunks = 10, rm.unequal=TRUE))
with(DATA, chunker(state, person, n.chunks = 10))
with(DATA, chunker(state, list(sex, adult), n.words = 10))
with(DATA, chunker(state, person, n.words = 10, rm.unequal=TRUE))

## Bigger data
with(hamlet, chunker(dialogue, person, n.chunks = 10))
with(hamlet, chunker(dialogue, person, n.words = 300))

## Not run: 
## with polarity hedonmetrics
dat <- with(pres_debates2012[pres_debates2012$person %in% qcv(OBAMA, ROMNEY), ], 
    chunker(dialogue, list(person, time), n.words = 300))

dat2 <- colsplit2df(list2df(dat, "dialogue", "person&time")[, 2:1])

dat3 <- split(dat2[, -2], dat2$time)
ltruncdf(dat3, 10, 50)

poldat <- lapply(dat3, function(x) with(x, polarity(dialogue, person, constrain = TRUE)))


m <- lapply(poldat, function(x) plot(cumulative(x)))
m <- Map(function(w, x, y, z) {
        w + ggtitle(x) + xlab(y) + ylab(z)
    }, 
        m, 
        paste("Debate", 1:3), 
        list(NULL, NULL, "Duration (300 Word Segment)"), 
        list(NULL, "Cumulative Average Polarity", NULL)
)

library(gridExtra)
do.call(grid.arrange, m)

## By person
## By person
poldat2 <- Map(function(x, x2){

    scores <- with(counts(x), split(polarity, person))
    setNames(lapply(scores, function(y) {
        y <- list(cumulative_average_polarity = y)
        attributes(y)[["constrained"]] <- TRUE
        qdap:::plot.cumulative_polarity(y) + xlab(NULL) + ylab(x2)
    }), names(scores))

}, poldat, paste("Debate", 1:3))

poldat2 <- lapply(poldat2, function(x) {
    x[[2]] <- x[[2]] + ylab(NULL)
    x
})

poldat2[[1]] <- Map(function(x, y) {
        x + ggtitle(y)
    },
        poldat2[[1]], qcv(Obama, Romney)
)

library(gridExtra)
do.call(grid.arrange, unlist(poldat2, recursive=FALSE))

## End(Not run)

trinker/qdap documentation built on Sept. 30, 2020, 6:28 p.m.