Remove unneeded file

code/calculate-summated-perf.R

-args <- R.utils::commandArgs(trailingOnly = TRUE, asValues = TRUE)
-
-set.seed(20220502)
-
-suppressPackageStartupMessages({
-  library(dplyr)
-  library(tidyr)
-  library(purrr)
-  library(SingleCellExperiment)
-  library(pROC)
-})
-
-source(here::here("code", "utils.R"))
-source(here::here("code", "analysis-utils.R"))
-
-summated_prediction <- function(result,
-                                method_id, data_id,
-                                n_genes) {
-
-  sce <- readRDS(here::here("data", "real_data", paste0(data_id, ".rds")))
-
-  unique_clusters <- unique(colLabels(sce))
-  n_clusters <- length(unique_clusters)
-
-  res <- result$result
-
-  auc <- numeric(n_clusters)
-  for (i in 1:n_clusters) {
-
-    cluster <- unique_clusters[[i]]
-
-    top_genes <- res %>%
-      filter(cluster == !!cluster) %>%
-      get_top_sel_mgs(n = n_genes) %>%
-      pull(gene)
-
-    if (is.na(top_genes) || is.null(top_genes)) {
-      auc[[i]] <- NA
-    } else {
-      score <- sce[top_genes, ] %>%
-        logcounts() %>%
-        colSums() %>%
-        unname()
-
-      cluster_status <- as.numeric(cluster == colLabels(sce))
-      auc[[i]] <- auc(cluster_status, score)
-    }
-  }
-
-  out <- tibble(auc, cluster = unique_clusters, method_id = method_id,
-                data_id = data_id)
-
-  out
-}
-
-# Debugging.
-print(args)
-print(args$wcs)
-
-wcs <- parse_wildcards(args$wcs)
-
-print(wcs)
-
-result <- readRDS(here::here(args$result))
-
-summated_perf <- summated_prediction(
-  result,
-  method_id = wcs$method_id,
-  data_id = wcs$data_id,
-  n_genes = 5
-)
-
-saveRDS(summated_perf, here::here(args$out))