R/summary_ggoutlier.R

In GGoutlieR: Identify Individuals with Unusual Geo-Genetic Patterns

#' Summarize GGoutlieR results
#' @description
#' Get a summary table from the `ggoutlier` output
#' @param  ggoutlier_res output from the function `ggoutlier`
#' @returns a table contains IDs of outliers and p-values.
#' @examples
#' library(GGoutlieR)
#' data(ggoutlier_example) # get an example output of ggoutlier
#' head(summary_ggoutlier(ggoutlier_example))
#' @export

summary_ggoutlier <- function(ggoutlier_res){
  method <- attributes(ggoutlier_res)$model
  if(method == "composite"){
    ## get significant outliers from genetic space KNN result
    sig_index1 <- ggoutlier_res$geneticKNN_result$statistics$significant
    sig_geneticKNN <- data.frame(
      "ID" = rownames(ggoutlier_res$geneticKNN_result$statistics)[sig_index1],
      "method" = rep("geneticKNN", sum(sig_index1)),
      "p.value" = ggoutlier_res$geneticKNN_result$statistics[sig_index1,"p.value"])
    sig_geneticKNN <- sig_geneticKNN[order(sig_geneticKNN$p.value, decreasing = F),]

    ## get significant outliers from geographical space KNN result
    sig_index2 <- ggoutlier_res$geoKNN_result$statistics$significant
    sig_geoKNN <- data.frame(
      "ID" = rownames(ggoutlier_res$geoKNN_result$statistics)[sig_index2],
      "method" = rep("geoKNN", sum(sig_index2)),
      "p.value" = ggoutlier_res$geoKNN_result$statistics[sig_index2,"p.value"])
    sig_geoKNN <- sig_geoKNN[order(sig_geoKNN$p.value, decreasing = F),]

    prop_sig <- c(mean(ggoutlier_res$geneticKNN_result$statistics$significant),
                  mean(ggoutlier_res$geoKNN_result$statistics$significant))
    names(prop_sig) <- c("geneticKNN", "geoKNN")
    summary_out <- rbind.data.frame(sig_geneticKNN, sig_geoKNN)
    summary_out <- summary_out[order(summary_out$p.value, decreasing = F),]
  }
  if(method == "ggoutlier_geoKNN"){
    ## get significant outliers from genetic space KNN result
    sig_index <- ggoutlier_res$statistics$significant
    sig_geneticKNN <- data.frame(
      "ID" = rownames(ggoutlier_res$statistics)[sig_index],
      "method" = rep("geoKNN", sum(sig_index)),
      "p.value" = ggoutlier_res$statistics[sig_index,"p.value"])
    sig_geneticKNN <- sig_geneticKNN[order(sig_geneticKNN$p.value, decreasing = F),]
    prop_sig <- c(mean(ggoutlier_res$geneticKNN_result$statistics$significant),
                  mean(ggoutlier_res$geoKNN_result$statistics$significant))
    names(prop_sig) <- c("geneticKNN")
    summary_out <- sig_geneticKNN
  }
  if(method == "ggoutlier_geoKNN"){
    ## get significant outliers from geographical space KNN result
    sig_index <- ggoutlier_res$geoKNN_result$statistics$significant
    sig_geoKNN <- data.frame(
      "ID" = rownames(ggoutlier_res$geoKNN_result$statistics)[sig_index],
      "method" = rep("geneticKNN", nrow(ggoutlier_res$statistics)),
      "p.value" = ggoutlier_res$geoKNN_result$statistics[sig_index,"p.value"])
    sig_geoKNN <- sig_geoKNN[order(sig_geoKNN$p.value, decreasing = F),]

    prop_sig <- c(mean(ggoutlier_res$geoKNN_result$statistics$significant))
    names(prop_sig) <- c("geoKNN")
    summary_out <- sig_geoKNN
  }
  rownames(summary_out) <- NULL
  #sig_ID_list <- unique(summary_out$ID)
  attr(summary_out, "proportion_of_outliers") = prop_sig
  #attr(summary_out, "significant_samples" = sig_ID_list)
  class(summary_out) <- c("data.frame", "ggoutlier_summary")
  return(summary_out)
} # summary_ggoutlier end