createWordcloud: Create Word Cloud Visualization.

In toaster: Big Data in-Database Analytics that Scales with Teradata Aster Distributed Platform

Description

Wrapper around wordcloud function that optionally saves graphics to the file of one of supported formats.

Usage

createWordcloud(words, freq, title = "Wordcloud", scale = c(8, 0.2),
  minFreq = 10, maxWords = 40, filename, format = c("png", "bmp", "jpeg",
  "tiff", "pdf"), width = 480, height = 480, units = "px",
  palette = brewer.pal(8, "Dark2"), titleFactor = 1)

Arguments

words	the words
freq	their frequencies
title	plot title
scale	a vector indicating the range of the size of the words (default c(4,.5))
minFreq	words with frequency below minFreq will not be displayed
maxWords	Maximum number of words to be plotted (least frequent terms dropped).
filename	file name to use where to save graphics
format	format of graphics device to save wordcloud image
width	the width of the output graphics device
height	the height of the output graphics device
units	the units in which height and width are given. Cab be px (pixels, the default), in (inches), cm or mm.
palette	color words from least to most frequent
titleFactor	numeric title character expansion factor; multiplied by par("cex") yields the final title character size. NULL and NA are equivalent to a factor of 1.

Details

Uses base graphics and worldcloud package to create a word cloud (tag cloud) visual reprsentation of for text data. Function uses 2 vectors of equal lengths: one contains list of words and the other has their frequencies.

Resulting graphics is saved in file in one of available graphical formats (png, bmp, jpeg, tiff, or pdf).

Word Cloud visuals apply to any concept that satisfies following conditions: * each data point (artifact) can be expressed with distinct word or compact text in distinct and self-explanatory fashion and * it assigns each artifact scalar non-negative metric. Given these two conditions we can use Word Clouds to visualize top, bottom or all artifacts in single word cloud visual.

Value

nothing

See Also

wordcloud

Examples

if(interactive()){
# initialize connection to Dallas database in Aster 
conn = odbcDriverConnect(connection="driver={Aster ODBC Driver};
                         server=<dbhost>;port=2406;database=<dbname>;uid=<user>;pwd=<pw>")

stopwords = c("a", "an", "the", "with")

# 2-gram tf-idf on offense table
daypart_tfidf_2gram = computeTfIdf(conn, "public.dallaspoliceall", 
                                   docId="extract('hour' from offensestarttime)::int/6",  
                                   textColumns=c('offensedescription','offensenarrative'),
                                   parser=nGram(2, delimiter='[  \\t\\b\\f\\r:\"]+'),
                                   stopwords=stopwords)

toRace <- function(ch) {
  switch(as.character(ch),
         "M" = "Male",
         "F" = "Female",
         "0" = "Night",
         "1" = "Morning",
         "2" = "Day",
         "3" = "Evening",
         "C" = "C",
         "Unknown")
}
                                  
createDallasWordcloud <- function(tf_df, metric, slice, n, maxWords=25, size=750) {
  words=with(tf_df$rs, tf_df$rs[docid==slice,])
  
  ## palette 
  pal = rev(brewer.pal(8, "Set1"))[c(-3,-1)]
  
  createWordcloud(words$term, words[, metric], maxWords=maxWords, scale=c(4, 0.5), palette=pal, 
                  title=paste("Top ", metric, "Offense", n, "- grams for", toRace(race)),
                  file=paste0('wordclouds/',metric,'_offense_',n,'gram_',toRace(slice),'.png'), 
                  width=size, height=size)
}

createDallasWordcloud(daypart_tfidf_2gram, 'tf_idf', 0, n=2, maxWords=200, size=1300)

}

Example output

Loading required package: RODBC

toaster documentation built on May 30, 2017, 3:51 a.m.