gglogistic: Plot logistic regression of probability from 0% to 100%...

In Ergative/logisticat: Plot Inverse Predictions for Logistic Regressions

Plot logistic regression of probability from 0% to 100% across a predictor.

Description

Plot logistic regression of probability from 0% to 100% across a predictor.

Usage

gglogistic(
  data,
  predictor,
  success.counts,
  failure.counts,
  treatment,
  replicate = NULL,
  line.var = "both",
  point.var = "none",
  boxplot.var = "none",
  inverse.var = "line",
  probability.of.interest = 0.5,
  xlabel = NULL,
  ylabel = NULL,
  treatment.label = NULL,
  treatment.colors = NULL,
  integer.x.breaks.by = NULL,
  fill.alpha = 0.25,
  replicate.alpha = 0.5,
  jitter.width = 0,
  dodge.width = 0,
  inverse.predict.digits = 3
)

Arguments

data	Data with continuous predictor and success and failure columns.
predictor	Continuous predictor for x-axis.
success.counts	Column representing number of successes per observation.
failure.counts	Column representing number of failures per observation.
treatment	Treatment variable.
replicate	Within-treatment replicates.
line.var	Variable to plot regression curves for: "both" (default), "treatment", "replicate", or "none".
point.var	Variable to plot points for: "both", "treatment", "replicate", or "none" (default). y-value represents fraction of successes in each treatment or replicate.
boxplot.var	Variable to plot boxplots for: "both", "treatment", "replicate", or "none" (default).
inverse.var	Variable to draw lines at inverse predictions for: "both", "treatment", "replicate", "none" or "line" (same as whatever was specified for line.var; default).
probability.of.interest	Calculate predictor value in each group that would yield this probability.
xlabel	Label for x-axis.
ylabel	Label for y-axis.
treatment.label	Label for treatment.
treatment.colors	Vector mapping treatment levels to colors.
integer.x.breaks.by	If specified, integer breaks in x-axis. Forces ggplot to break at integer values.
fill.alpha	Opacity in range [0,1] for fill of boxplots. Boxplots can be hard to see if there is a lot of data, and having a light fill color helps to distinguish them.
replicate.alpha	Opacity in range [0,1] to draw replicate geoms at when the same geom is also drawn for treatment. Helps distinguish treatment from replicate values.
jitter.width	If specified, amount to jitter points in x-axis. Useful if predictor was sampled at discrete steps that would otherwise cause points to be hard to tell apart.
dodge.width	If specified, amount to dodge replicate boxplots or groups of points by.
inverse.predict.digits	Number of digits to print for inverse predicted values.

Value

A ggplot object.

Examples

metamorphosis.data <- logisticat::metamorphosis()
head(metamorphosis.data) # Has columns for success and failure counts
logisticat::gglogistic(metamorphosis.data,
                       predictor=days,
                       success.counts=metamorphosed,
                       failure.counts=did.not.metamorphose,
                       treatment=ration,
                       replicate=beaker,
                       probability.of.interest=0.5)

Ergative/logisticat documentation built on May 17, 2022, 12:32 a.m.