plot_polymr: Plot results of PolyMR
In JonSulc/PolyMR: Model Non-Linear Causal Effects Through Polynomial Regression

View source: R/plot_polymr.R

plot_polymr

R Documentation

Plot results of PolyMR

Description

This function provides an easy interface to quickly visualize the results from PolyMR using ggplot2.

Usage

plot_polymr(
  polymr_results,
  show_polymr = TRUE,
  show_observational = TRUE,
  show_binned_observations = TRUE,
  show_confidence_ribbons = TRUE,
  scale_values = TRUE,
  output_filename = NULL,
  main_title = "Effects of exposure on outcome",
  subtitle = NULL,
  auto_subtitle = TRUE,
  xlab = ifelse(scale_values, "Exposure (scaled)", "Exposure"),
  ylab = ifelse(scale_values, "Outcome (scaled)", "Outcome"),
  ylims = NULL,
  show_legend = TRUE,
  ggsave_options = list(width = 18, height = 18, units = "cm"),
  n_points = 1000
)

Arguments

`polymr_results`	A list containing the results from the `polymr` function.
`show_polymr`	Boolean indicating whether or not to plot the inferred causal function (default is TRUE).
`show_observational`	Boolean indicating whether or not to plot the observed association. Requires the observational association to have been estimated by the `polymr` function using `return_observational_function = TRUE`. Default is TRUE.
`show_binned_observations`	Boolean indicating whether or not to plot the binned observations, specifically the median exposure and outcome for each bin. Requires the binned values to have been returned by the `polymr` function using `return_binned_observations = TRUE`. Default is TRUE.
`show_confidence_ribbons`	Boolean indicating whether or not to display the 95% confidence hulls (default is TRUE).
`scale_values`	Boolean indicating whether or not to leave phenotypes (exposure and outcome) standardized with mean 0 and variance 1 or convert them back to their original range using the values in `phenotypes_summary`. Default is TRUE, indicating that the values will be plotted on a standard scale.
`output_filename`	The file name of the plot to be saved, if any. This will override any `filename` provided in `ggsave_options`. Default is `NULL`, in which case the plot will not be saved.
`main_title`	Main title of the plot (default is "Effect of exposure on outcome").
`subtitle`	Subtitle of the plot. If `auto_subtitle` is enabled, this will be prepended to the automatic subtitle. Default is `NULL`.
`auto_subtitle`	Whether to create and include a summary of results as subtitle. This summary includes the p-values and variance explained for the plotted functions (total variance for the observational association, causal variance for PolyMR). Default is `TRUE`.
`xlab`	Label for the X axis. Default is "Exposure (scaled)" if `scale_values` is `TRUE` and "Exposure" otherwise.
`ylab`	Label for the Y axis. Default is "Outcome (scaled)" if `scale_values` is `TRUE` and "Outcome" otherwise.
`ylims`	An optional vector of length 2 defining the boundaries of the plotted region on the Y axis. Default is NULL, relying on `ggplot2` standard behavior.
`show_legend`	Whether or not to display the legend next to the plot (default is `TRUE`).
`ggsave_options`	A list of options to provide to `ggplot2::ggsave`. Ignored if no filename is specified here or using `output_filename`. Default is a list specifying a `width` and `height` both equal to 18 `cm`.
`n_points`	Number of points to plot in rendering the curve. Default is 1000, resulting in a smooth enough curve for most purposes.