plot.word_coverage: Plot method for word_coverage objects

In sbo: Text Prediction via Stupid Back-Off N-Gram Models

Description

Plot cumulative corpus coverage fraction of a dictionary.

Usage

## S3 method for class 'word_coverage'
plot(
  x,
  include_EOS = FALSE,
  show_limit = TRUE,
  type = "l",
  xlim = c(0, length(x)),
  ylim = c(0, 1),
  xticks = seq(from = 0, to = length(x), by = length(x)/5),
  yticks = seq(from = 0, to = 1, by = 0.25),
  xlab = "Rank",
  ylab = "Covered fraction",
  title = "Cumulative corpus coverage fraction of dictionary",
  subtitle = "_default_",
  ...
)

Arguments

x	a word_coverage object.
include_EOS	length one logical. Should End-Of-Sentence tokens be considered in the computation of coverage fraction?
show_limit	length one logical. If TRUE, plots an horizontal line corresponding to the total coverage fraction.
type	what type of plot should be drawn, as detailed in ?plot.
xlim	length two numeric. Extremes of the x-range.
ylim	length two numeric. Extremes of the y-range.
xticks	numeric vector. position of the x-axis ticks.
yticks	numeric vector. position of the y-axis ticks.
xlab	length one character. The x-axis label.
ylab	length one character. The y-axis label.
title	length one character. Plot title.
subtitle	length one character. Plot subtitle; if "default", prints dictionary length and total covered fraction.
...	further arguments passed to or from other methods.

Details

This function generates nice plots of cumulative corpus coverage fractions. The x coordinate in the resulting plot is the word rank in the underlying dictionary; the y coordinate at x is the cumulative coverage fraction for rank <= x.

Author(s)

Valerio Gherardi

Examples

c <- word_coverage(twitter_dict, twitter_test)
plot(c)

sbo documentation built on Dec. 6, 2020, 1:06 a.m.