Plot the distribution of the target variable — distribution • modelblueprint

Bins the target variable and plots the total exposure per bin as a bar chart — an exposure-weighted distribution of the target. Shares its argument conventions, binning, NA handling, split behaviour, and plotly styling with one_way(); the only difference is that the target itself is the x-axis variable, so there is no obs argument.

Usage

distribution(data, ...)

# Default S3 method
distribution(
  data,
  var = "target",
  exposure = "exposure",
  split = NA_character_,
  bins = 35L,
  time_unit = NA_character_,
  type_agg = c("equal_exposure", "equal_range"),
  ret = c("plot", "data"),
  verbose = FALSE,
  ...
)

Arguments

data: A data frame or data.table. Must contain all columns referenced by other arguments.
...: Arguments passed to methods.
var: Target column whose distribution to plot. Bare name or string. Default "target".
exposure: Column of exposure weights. If the column does not exist in data, every row is given weight 1. Bare name or string. Default "exposure".
split: Optional column to group bars by. Bare name, string, or NA / NULL for no split. Default NA.
bins: [integer(1)] Number of equal-exposure bins for numeric targets with more than bins unique values. Default 35.
time_unit: [character(1)] For Date / POSIXct targets, the width of each bin (passed to base::cut.POSIXt()). Default NA.
type_agg: [character(1)] Binning strategy for numeric targets: "equal_exposure" (default) or "equal_range".
ret: [character(1)] "plot" (default) returns a plotly object; "data" returns the aggregated data.table.
verbose: [logical(1)] Report how many NA values of var were moved to the trailing "NA" category? Default FALSE.

Value

A plotly object, or a data.table when ret = "data", or NULL with a warning if the plot cannot be produced.

Details

Column name arguments (var, exposure, split) accept both bare (unquoted) names and strings, so distribution(df, mpg) and distribution(df, "mpg") are equivalent.

Examples

# \donttest{
distribution(mtcars, mpg, bins = 10)

distribution(mtcars, var = "mpg", split = "am")

# }