Update dev version

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: ale
 Title: Interpretable Machine Learning and Statistical Inference with Accumulated Local Effects (ALE)
-Version: 0.3.0.20241110
+Version: 0.3.0.20241111
 Authors@R: c(
     person("Chitu", "Okoli", , "[email protected]", role = c("aut", "cre"),
            comment = c(ORCID = "0000-0001-5574-7572"))

R/stats.R

@@ -515,8 +515,6 @@ summarize_conf_regions_1D <- function(
           mid_bar = case_when(
             .data$.y_hi < y_zeroed_summary['med_lo'] ~ 'below',
             .data$.y_lo > y_zeroed_summary['med_hi'] ~ 'above',
-            # .data$.y_hi < y_summary[['med_lo', 1]] ~ 'below',
-            # .data$.y_lo > y_summary[['med_hi', 1]] ~ 'above',
             .default = 'overlap'
           ) |>
             factor(ordered = TRUE, levels = c('below', 'overlap', 'above')),
@@ -574,14 +572,13 @@ summarize_conf_regions_1D <- function(
 
         cr <- cr |>
           rename(
-            # x = '.x',
             n = '.n',
             y = '.y',
           ) |>
           mutate(
             pct = (.data$n / sum(it.ale_data$.n)) * 100,
             # Convert x column from ordinal to character for consistency across terms
-            x = as.character(x),
+            x = as.character(.data$x),
           ) |>
           select('x', 'y', 'mid_bar', 'n', 'pct')
       }
@@ -597,33 +594,13 @@ summarize_conf_regions_1D <- function(
     # https://bard.google.com/chat/ea68c7b9e8437179
     select(
       'term',
-      # any_of is used because categorical variables do not have 'start_x', 'end_x', 'x_span_pct'
-      # while numeric values do not have 'x'
+      # any_of is used because categorical variables do not have 'start_x', 'end_x', 'x_span_pct' while numeric values do not have 'x'
       any_of(c('x', 'start_x', 'end_x', 'x_span_pct')),
       'n', 'pct',
       any_of(c('y', 'start_y', 'end_y', 'trend')),
       'mid_bar'
     )
 
-  # browser()
-
-
-  # Highlight which confidence regions are statistically significant
-  # sig_conf_regions <- map2(
-  #   cr_by_term, names(cr_by_term),
-  #   \(it.conf_tbl, it.term) {
-  #     it.conf_tbl$term <- it.term
-  #
-  #     if ('x' %in% names(it.conf_tbl)) {
-  #       # Convert x column from ordinal to character for consistency across terms
-  #       it.conf_tbl$x <- as.character(it.conf_tbl$x)
-  #     }
-  #
-  #     it.conf_tbl
-  #   }
-  # ) |>
-  #   bind_rows() |>
-
   sig_conf_regions <- cr_by_term |>
     filter(.data$mid_bar != 'overlap')
 
@@ -665,16 +642,14 @@ summarize_conf_regions_2D <- function(
           mid_bar = case_when(
             .data$.y_hi < y_zeroed_summary['med_lo'] ~ 'below',
             .data$.y_lo > y_zeroed_summary['med_hi'] ~ 'above',
-            # .data$.y_hi < y_summary[['med_lo', 1]] ~ 'below',
-            # .data$.y_lo > y_summary[['med_hi', 1]] ~ 'above',
             .default = 'overlap'
           ) |>
             factor(ordered = TRUE, levels = c('below', 'overlap', 'above')),
         ) |>
         select(-c('.y_lo':'.y_hi')) |>
         rename(
-          n = .n,
-          y = .y
+          n = '.n',
+          y = '.y'
         )
 
       # Initialize cr_groups, used only if one or both x variables is non-numeric
@@ -716,20 +691,11 @@ summarize_conf_regions_2D <- function(
           n   = sum(.data$n),
           pct = (n / total_n) * 100,
           y   = mean(.data$y),
-          # n_below = sum(mid_bar == 'below'),
-          # n_overlap = sum(mid_bar == 'overlap'),
-          # n_above = sum(mid_bar == 'above'),
-          # pct_below = (n_below / total_n) * 100,
-          # pct_overlap = (n_overlap / total_n) * 100,
-          # pct_above = (n_above / total_n) * 100,
         )
 
       # Rename the x variables with their original variable names
       x1_x2_names <- x1_x2_names |>
         stringr::str_remove("\\.bin$|\\.ceil$")
-      # names(cr)[1] <- x1_x2_names[1]
-      # names(cr)[2] <- x1_x2_names[2]
-
 
       # Convert x data columns uniformly to character format
       cr[[1]] <- as.character(cr[[1]])
@@ -738,39 +704,19 @@ summarize_conf_regions_2D <- function(
       # Rename the x data columns consistently
       names(cr)[1:2] <- c('x1', 'x2')
 
+      # Return value for map function
       cr |>
         mutate(
           term1 = x1_x2_names[1],
           term2 = x1_x2_names[2],
         ) |>
         select('term1', 'x1', 'term2', 'x2', everything())
-
-
     }) |>
     set_names(names(ale_data_list)) |>
     bind_rows()
 
-  # browser()
-
   # Highlight which confidence regions are statistically significant
   sig_conf_regions <- cr_by_term |>
-    # map(\(it.conf_tbl) {
-    #   # browser()
-    #   x1_x2_names <- names(it.conf_tbl)[1:2]
-    #   # Convert x data columns uniformly to character format
-    #   it.conf_tbl[[1]] <- as.character(it.conf_tbl[[1]])
-    #   it.conf_tbl[[2]] <- as.character(it.conf_tbl[[2]])
-    #
-    #   # Rename the x data columns consistently
-    #   names(it.conf_tbl)[1:2] <- c('x1', 'x2')
-    #
-    #   it.conf_tbl |>
-    #     mutate(
-    #       term1 = x1_x2_names[1],
-    #       term2 = x1_x2_names[2],
-    #     ) |>
-    #     select('term1', 'x1', 'term2', 'x2', everything())
-    # }) |>
     filter(.data$mid_bar != 'overlap')
 
 
@@ -791,7 +737,6 @@ summarize_conf_regions_1D_in_words <- function(
     band_type = 'median'
 ) {
   map_chr(1:nrow(conf_region_summary), \(.row_num) {
-    # browser()
     with(
       conf_region_summary[.row_num, ],
       if (exists('start_x')) { # conf_region_summary is numeric