calc_stats_pred_obs: Calculate Statistics for Model Prediction and/or Observed...

View source: R/core_stats.R

calc_stats_pred_obsR Documentation

Calculate Statistics for Model Prediction and/or Observed Data

Description

This function derives statistics that can be calculated for model predictions and/or observed data. However, it does not calculate it, but rather calls the respective backend functions. Supported statistics currently include:

  • Basic Summary Statistics (i.e., means and response percentages calc_basic_stats())

  • Conditional Accuracy Functions (CAFs; calc_cafs())

  • Quantiles (calc_quantiles())

  • Delta Functions (calc_delta_funs()).

  • Density Estimates (calc_dens()).

Usage

calc_stats_pred_obs(type, b_coding, conds, ..., scale_mass = FALSE)

Arguments

type

character string, specifying the type of statistic to calculate. Available options are "basic_stats", "cafs", "quantiles", "delta_funs", and "densities".

b_coding

list for the boundary coding (see b_coding).

conds

character vector, specifying the conditions to include in calculations (used for labeling and subsetting the model PDFs and the observed data).

...

Additional parameters passed on to the specific statistic calculation function (see Details).

scale_mass

a single logical, only relevant for density estimation. If TRUE, PDF masses are scaled proportional to the number of trials per condition.

Details

When calling this function the arguments all_rts_u/all_rts_l and/or all_pdfs must always be specified (see re_evaluate_model, obs_data). Otherwise, the backend functions won't work properly. Further arguments are:

  • for CAFS: n_bins controls the number of bins, with a default of 5.

  • for Quantiles and Delta Functions: probs controls the quantiles to calculate. Default is seq(0.1, 0.9, 0.1) (see drift_dm_default_probs()).

  • for basic summary satistics, Quantiles, and Delta Function: skip_if_contr_low controls if quantiles and means are calculated for PDFs with very small contribution (see also drift_dm_skip_if_contr_low()).

  • for densities: discr controls the bin width for the observed data. Default is 0.015 seconds

This function gets called by calc_stats()

Value

A data frame with the calculated statistic across conds (ordered according to Source).


dRiftDM documentation built on Dec. 1, 2025, 5:08 p.m.