R/reshape_wide_tt.R

In marianschmidt/msSPChelpR: Helper Functions for Second Primary Cancer Analyses

#' Reshape dataset to wide format - tidytable version
#'
#' @param df dataframe
#' @param case_id_var String with name of ID variable indicating same patient.
#'                E.g. \code{idvar="PUBCSNUM"} for SEER data.
#' @param time_id_var String with name of variable that indicates diagnosis per patient.
#'                E.g. \code{timevar="SEQ_NUM"} for SEER data.
#' @param timevar_max Numeric; default 6. Maximum number of cases per id. 
#'                    All tumors > timevar_max will be deleted before reshaping.              
#' @param datsize Number of rows to be taken from df. This parameter is mainly for testing. Default is Inf so that df is fully processed.
#' @return wide_df
#' @export
#' @examples 
#' 
#' data(us_second_cancer)
#' 
#' msSPChelpR::reshape_wide_tt(us_second_cancer,
#'                          case_id_var = "fake_id", 
#'                          time_id_var = "SEQ_NUM", 
#'                          timevar_max = 2,
#'                          datsize = 10000)
#'

reshape_wide_tt <- function(df, case_id_var, time_id_var, timevar_max = 6, datsize = Inf){
  
  case_id_var <- rlang::ensym(case_id_var)
  time_id_var <- rlang::ensym(time_id_var)
  
  ### restrict size of data.frame to datsize number of rows
  if(nrow(df) > datsize){
    df <- df[c(1:datsize), ]
  }
  
  ### get names from df to provide to pivot function
  trans_vars <- names(df)[!names(df) %in% c(rlang::as_name(case_id_var), rlang::as_name(time_id_var))]
  
  ### determine maximum number of cases per patient and deleting all cases > timevar_max
  max_time <- max(as.numeric(df[[rlang::as_name(time_id_var)]]), na.rm = TRUE)
  
  if(max_time > timevar_max){
    rlang::inform(paste("Long dataset had too many cases per patient. Wide dataset is limited to ", timevar_max," cases per id as defined in timevar_max option."))
    
    df <- df %>%
      #sort by case_id and time_id_var
      tidytable::arrange(!!case_id_var, !!time_id_var) %>%
      #calculate new renumbered variable group by case_id_var
      tidytable::mutate(counter = as.integer(tidytable::row_number()), .by = !!case_id_var) %>%
      #filter based on new renumbered variable
      tidytable::filter(counter <= timevar_max) %>% 
      tidytable::select(-counter)
    
    max_time <- timevar_max
    
  }
  
  ### prepare name order of cols with all vars per time_id sorted together
  
  col_order <- 
    #create outer product of trans_vars and time_ids
    outer(trans_vars, 1:max_time, paste, sep = ".") %>%
    #make vector binding rows, one col after the other
    c(.)
  
  ### perform pivot_wider
  df %>% tidytable::pivot_wider(
    names_from = {{time_id_var}}, 
    values_from = tidyselect::all_of(trans_vars),
    names_sep = "."
  ) %>%
    #sort by case_id_var
    tidytable::arrange(as.numeric(rlang::eval_tidy(!!case_id_var))) %>%
    #order columns by old col order 
    tidytable::relocate(tidyselect::all_of(c(rlang::as_name(case_id_var), 
                                              col_order)))
  
}