A plot function to plot multiple rm_main_results

Source: R/rm_main_results.R

This function can create a line plot of the results or temporal cross-decoding results for the the zero-one loss, normalized rank and/or decision values after the decoding analysis has been run (and all results have been aggregated).

plot_main_results(
  results_dir_name,
  results_to_plot,
  results_to_show = "zero_one_loss",
  type = "line",
  errorbar = NULL,
  display_names = NULL
)

Arguments

results_dir_name

A string specifying the directory name that contains files with DECODING_RESULTS that have rm_main_results as one of the result metrics.

results_to_plot

This can be set to a vector of strings specifying result_names for the results to plot, or a vector of numbers that contain the rows in the results_manifest file of the results that should be compared. The results_manifest file should be created from saving results using the log_save_results() function. Finally, if this is set to a single string that is a regular expression, all results in the results_manifest file result_name variable that match the regular expression will be plotted.

results_to_show

A string specifying the types of results to plot. Options are: 'zero_one_loss', 'normalized_rank', 'decision_values', or 'all'.

type

A string specifying the type of results to plot. Options are 'TCD' to plot a temporal cross decoding matrix or 'line' to create a line plot of the decoding results as a function of time.

errorbar

A string specifying if error bars should be plotted. Options are: 'sd', 'se', or '2se'. If this is set to NULL, then no error bars will be plotted. If this is set to 'sd', then the standard deviation of the results will be plotted. If this is set to 'se', then the standard error of the results will be plotted. If this is set to '2se', then two times the standard error of the results will be plotted (which is often used to represent a 95% confidence interval). Note, these error bars are slight underestimates of the sd and sderr because when using cross-validation the test data is not independent of the training data. Also, note that error bars can only be plotted for line plots and not for TCD plots.

display_names

A vector of strings specifying what the labels on the plots should say for each result. If this is NULL, the result names will be the names from the manifest file's result_name column, or if these are set to "No result name set" then the analysisID will be the label.

Value

Returns a ggplot object that a comparison of main decoding results.