Update tidy_distribution_summary_tbl() function

[spsanderson]

Add ci_lo and ci_hi to tidy_distribution_summary_tbl()

[spsanderson]

Function:

#' Tidy Distribution Summary Statistics Tibble
#'
#' @family Summary Statistics
#' @family Table Data
#'
#' @author Steven P. Sanderson II, MPH
#'
#' @details This function takes in a `tidy_` distribution table and
#' will return a tibble of the following information:
#' -  `sim_number`
#' -  `mean_val`
#' -  `median_val`
#' -  `std_val`
#' -  `min_val`
#' -  `max_val`
#' -  `skewness`
#' -  `kurtosis`
#' -  `range`
#' -  `iqr`
#' -  `variance`
#'
#' The kurtosis and skewness come from the package `healthyR.ai`
#'
#' @description This function returns a summary statistics tibble. It will use the
#' y column from the `tidy_` distribution function.
#'
#' @param .data The data that is going to be passed from a a `tidy_` distribution
#' function.
#' @param ... This is the grouping variable that gets passed to [dplyr::group_by()]
#' and [dplyr::select()].
#'
#' @examples
#' library(dplyr)
#'
#' tn <- tidy_normal(.num_sims = 5)
#' tb <- tidy_beta(.num_sims = 5)
#'
#' tidy_distribution_summary_tbl(tn)
#' tidy_distribution_summary_tbl(tn, sim_number)
#'
#' data_tbl <- tidy_combine_distributions(tn, tb)
#'
#' tidy_distribution_summary_tbl(data_tbl)
#' tidy_distribution_summary_tbl(data_tbl, dist_type)
#'
#' @return
#' A summary stats tibble
#'
#' @export
#'

tidy_distribution_summary_tbl <- function(.data, ...) {
  
  # Get the data attributes
  atb <- attributes(.data)
  
  if (!"tibble_type" %in% names(atb) & !"tibble_type" %in% names(atb$all)) {
    rlang::abort("The data passed must come from a `tidy_` distribution function.")
  }
  
  data_tbl <- dplyr::as_tibble(.data)
  
  summary_tbl <- data_tbl %>%
    dplyr::group_by(...) %>%
    dplyr::select(..., y) %>%
    dplyr::summarise(
      mean_val = mean(y, na.rm = TRUE),
      median_val = stats::median(y, na.rm = TRUE),
      std_val = sd(y, na.rm = TRUE),
      min_val = min(y),
      max_val = max(y),
      skewness = tidy_skewness_vec(y),
      kurtosis = tidy_kurtosis_vec(y),
      range = tidy_range_statistic(y),
      iqr = stats::IQR(y),
      variance = stats::var(y),
      ci_low = ci_lo(y),
      ci_high = ci_hi(y)
    ) %>%
    dplyr::ungroup()
  
  return(summary_tbl)
}

Example:

> library(dplyr)
> 
> tn <- tidy_normal(.num_sims = 5)
> tb <- tidy_beta(.num_sims = 5)
> 
> tidy_distribution_summary_tbl(tn)
# A tibble: 1 × 12
  mean_val median_val std_val min_val max_val skewness kurtosis range   iqr
     <dbl>      <dbl>   <dbl>   <dbl>   <dbl>    <dbl>    <dbl> <dbl> <dbl>
1   0.0210     0.0807   0.990   -2.68    2.22  -0.0478     2.78  4.90  1.22
# … with 3 more variables: variance <dbl>, ci_low <dbl>, ci_high <dbl>
> tidy_distribution_summary_tbl(tn, sim_number)
# A tibble: 5 × 13
  sim_number mean_val median_val std_val min_val max_val skewness kurtosis range
  <fct>         <dbl>      <dbl>   <dbl>   <dbl>   <dbl>    <dbl>    <dbl> <dbl>
1 1            0.122      0.0777   0.895   -1.45    2.12   0.480      2.74  3.57
2 2            0.0456     0.153    0.945   -1.79    2.02   0.0614     2.51  3.81
3 3           -0.0754    -0.107    1.19    -2.68    1.99  -0.142      2.34  4.67
4 4           -0.167      0.0313   0.887   -2.17    1.95  -0.195      2.54  4.12
5 5            0.180      0.231    1.00    -2.12    2.22  -0.189      3.07  4.34
# … with 4 more variables: iqr <dbl>, variance <dbl>, ci_low <dbl>, ci_high <dbl>
> 
> data_tbl <- tidy_combine_distributions(tn, tb)
> 
> tidy_distribution_summary_tbl(data_tbl)
# A tibble: 1 × 12
  mean_val median_val std_val min_val max_val skewness kurtosis range   iqr
     <dbl>      <dbl>   <dbl>   <dbl>   <dbl>    <dbl>    <dbl> <dbl> <dbl>
1    0.254      0.338   0.764   -2.68    2.22   -0.755     4.52  4.90 0.671
# … with 3 more variables: variance <dbl>, ci_low <dbl>, ci_high <dbl>
> tidy_distribution_summary_tbl(data_tbl, dist_type)
# A tibble: 2 × 13
  dist_type    mean_val median_val std_val  min_val max_val skewness kurtosis range
  <fct>           <dbl>      <dbl>   <dbl>    <dbl>   <dbl>    <dbl>    <dbl> <dbl>
1 Gaussian c(…   0.0210     0.0807   0.990 -2.68      2.22   -0.0478     2.78 4.90 
2 Beta c(1, 1…   0.486      0.476    0.282  0.00133   0.997   0.113      1.88 0.995
# … with 4 more variables: iqr <dbl>, variance <dbl>, ci_low <dbl>, ci_high <dbl>