R/make_table_popchar.R

In Rapporteket/norspis2: Plots and Tables of Results and Data Quality in NorSpis

#' make_table_popchar
#'
#' Function for table with overview of patient characteristics.
#'
#' The name of this function when maing aarsrapport 2019, in 2020,  was make_popCharTab().
#'
#' @param RegData
#' @param varsInTab
#'
#' @return
#' @export
#'
#' @examples

make_table_popchar <- function(RegData = myFilteredData,
                               varsInTab = myvarString) {

  output_tibble <- tibble() #just making an empty tibble where output of the "for loop" can go

  for(myvar in varsInTab) {

    #This makes a small summary table with counts for each category of the variable
    summarytable_variable <- RegData %>%
      group_by(!!myvar)%>% #we group by this variable (i.e. the variable's categories) to get counts within each category(below, by summarize())
      summarize(counts = sum(!is.na(PasientID))) %>%#summarizes with counts for each category. Here we use Pasient.ID to count/represent each patient
      mutate(counts2 = case_when(!is.na(!!myvar) ~ counts))%>% #sum N, except NAs
      mutate(perc = counts/(sum(counts))*100) %>%
      mutate(perc2 = counts2/(sum(counts2, na.rm = T))*100) %>% #perc2 is the percantages where missing are not brought into calculations
      #mutate(perc2 := counts/(sum(counts)-.[is.na(!!myvar) , "counts"][[1]]) * 100) %>% #"[is.na(!!myvar) , "counts"][[1]]" extracts the cell value
      #mutate(perc2 := case_when(.[is.na(!!myvar),][[1]] ~ .[is.na(!!myvar) , "perc"][[1]],
      #                          TRUE ~ perc2))%>%
      mutate(perc2 = case_when(is.na(perc2) ~ perc,
                               TRUE ~ perc2)) %>%
      mutate(countsum = paste0("(N = ", sum(counts2, na.rm = T), ")")) %>% #sums the valid observations (i.e. does not include NAs)
      #Rounding
      mutate(perc =round(perc,1),
             perc2 =round(perc2,1))%>% #just rounding perc to one decimal
      #Add variable name as new  column or row
      mutate(myvarname := colnames(.[,1])) %>% #variable name added as a column
      #add_row(!!myvar:=colnames(.[,1]), .before = 1) %>% #adds a new row with variable name as value in first cell, and second cell empty, this is to prepare for graphical/table presentation later
      ## If you add it as a row you must bake NA values blank (togehter with converting them to characters)
      #mutate(counts = case_when(is.na(counts) ~ as.character(" "), #make NAs blank in counts column
      #                           TRUE ~ as.character(counts))) %>%
      # mutate(perc2 = case_when(is.na(perc2) ~ as.character(" "), #make NAs blank in perc2 column
      #                           TRUE ~ as.character(perc2))) %>%

      ##Changing names:
      mutate(!!myvar := forcats::fct_explicit_na(!!myvar, "Missing"))%>%
      #make column with variable categories character (from ordered factor) to be able to merge with other variables, else error will appear
      mutate(!!myvar :=as.character(!!myvar))%>%#make column with variable categories character (from ordered factor) to be able to merge with other variables, else error will appear
      #mutate(!!myvar := case_when(is.na(!!myvar) ~ "Missing", #This worked when myvar var character, but failed for factor
      #                            TRUE ~ !!myvar)) %>% #in the first column, named "!!myvar", we give NA the value "Missing"
      #First column...
      rename(" " := !!myvar) %>% #renames the first column to " " empty(before I used the name"categories")
      rename(N = counts) %>%  #rename counts2 column
      rename("%" = perc2) %>%
      select(., -counts2) %>%  #remove the counts2 abnd the perc column, don't need them anymore
      select(., -perc) %>%
      relocate(myvarname, countsum, before=1) %>%#moving myvarname and countsum column first %>%
      rename("  " = myvarname) %>% #rename column to empty (NOTE: double spaces used here because empty with one space is already used)
      mutate("   " = "") %>%  #make new ampty columnt that we can use for, for instance, a  minibar (flextable::minibar())
      rename("    " = countsum)
    #Fix format

    #Here we just merge the summary tables made above (one table/variable is added for each new running of the loop)
    output_tibble <- dplyr::bind_rows(output_tibble,
                                      summarytable_variable)

  }

  return(output_tibble)
}