freqs_list: Frequencies on Lists and UpSet Plot

In laresbernardo/lares: Analytics & Machine Learning Sidekick

Frequencies on Lists and UpSet Plot

Description

Visualize frequency of elements on a list, list vector, or vector with comma separated values. Detect which combinations and elements are the most frequent and how much they represent of your total observations. This is similar to the UpSet Plots which may be used as an alternative to Venn diagrams.

Usage

freqs_list(
  df,
  var = NULL,
  wt = NULL,
  fx = "mean",
  rm.na = FALSE,
  min_elements = 1,
  limit = 10,
  limit_x = NA,
  limit_y = NA,
  tail = TRUE,
  size = 10,
  unique = TRUE,
  abc = FALSE,
  title = "",
  plot = TRUE
)

Arguments

df	Data.frame
var	Variable. Variables you wish to process.
wt	Variable, numeric. Select a numeric column to use in the colour scale, used as sum, mean... of those values for each of the combinations.
fx	Character. Set operation: mean, sum
rm.na	Boolean. Remove NA value from wt?
min_elements	Integer. Exclude combinations with less than n elements
limit, limit_x, limit_y	Integer. Show top n combinations (x) and/or elements (y). The rest will be grouped into a single element. Set argument to 0 to ignore. limit_x/limit_y answer to limit's argument.
tail	Boolean. Show tail grouped into "..." on the plots?
size	Numeric. Text base size
unique	Boolean. a,b = b,a?
abc	Boolean. Do you wish to sort by alphabetical order?
title	Character. Overwrite plot's title with.
plot	Boolean. Plot viz? Will be generated anyways in the output object

Value

List. data.frame with the data results, elements and combinations.

See Also

Other Frequency: freqs(), freqs_df(), freqs_plot()

Other Exploratory: corr_cross(), corr_var(), crosstab(), df_str(), distr(), freqs(), freqs_df(), freqs_plot(), lasso_vars(), missingness(), plot_cats(), plot_df(), plot_nums(), tree_var()

Other Visualization: distr(), freqs(), freqs_df(), freqs_plot(), noPlot(), plot_chord(), plot_survey(), plot_timeline(), tree_var()

Examples

## Not run: 
df <- dplyr::starwars
head(df[, c(1, 4, 5, 12)], 10)

# Characters per movies combinations in a list column
head(df$films, 2)
freqs_list(df, films)

# Skin colours in a comma-separated column
head(df$skin_color)
x <- freqs_list(df, skin_color, min_elements = 2, limit = 5, plot = FALSE)
# Inside "x" we'll have:
names(x)

# Using the 'wt' argument to add a continuous value metric
# into an already one-hot encoded columns dataset (and hide tail)
csv <- "https://raw.githubusercontent.com/hms-dbmi/UpSetR/master/inst/extdata/movies.csv"
movies <- read.csv(csv, sep = ";")
head(movies)
freqs_list(movies,
  wt = AvgRating, min_elements = 2, tail = FALSE,
  title = "Movies\nMixed Genres\nRanking"
)
# So, please: no more Comedy+SciFi and more Drama+Horror films (based on ~50 movies)!

## End(Not run)

laresbernardo/lares documentation built on Oct. 23, 2024, 12:05 p.m.