Update kpi_averaging.R

src/main/R/drtAnalysis/kpi_averaging.R

@@ -1,17 +1,19 @@
 #####libraries####
 library(stringr)
+library(tidyverse)
 
 #####global variables####
-path_to_data <- "runs"
-
+path_to_data <- "path/to/data (folder of a specific case, with different seeds)"
+stats = c("mean","median","sd" ,"max", "min")
 
 ##### Collect all folder names####
-
 folders_list<-list.files(path_to_data,full.names = TRUE)
-
 folders_seeded = list()
 
 for(i in 1:length(folders_list)){
+  if (endsWith(folders_list[i],".tsv")){
+    next
+  }
 
   case_name = tail(str_split(folders_list[i],"-")[[1]],n = 1)
 
@@ -22,83 +24,88 @@ for(i in 1:length(folders_list)){
   }
 }
 
-#####Reading and averaging av tables#####
-
-
-folders_av_averaged_table = list()
+#########################################
+#####Reading and averaging drt tables####
+folders_drt_averaged_table = list()
 
 for(case_name in names(folders_seeded)){
-
-
-
   for(folder in folders_seeded[[case_name]]){
     files_list<- list.files(paste0(folder,"/analysis-drt-service-quality"),full.names = TRUE)
-    av_KPI_file <- files_list[grepl(pattern = "av_KPI.tsv",files_list)]
-    cat("processing ",av_KPI_file," \r\n")
-    av_KPI_table <- read.delim(av_KPI_file)
+    drt_KPI_file <- files_list[grepl(pattern = "drt_KPI.tsv",files_list)]
+    cat("processing ",drt_KPI_file," \r\n")
+    drt_KPI_table <- read.delim(drt_KPI_file)
 
-    if(!case_name %in% names(folders_av_averaged_table)){
-      folders_av_averaged_table[[case_name]] = av_KPI_table
+    if(!case_name %in% names(folders_drt_averaged_table)){
+      folders_drt_averaged_table[[case_name]] = drt_KPI_table
     }else{
-      folders_av_averaged_table[[case_name]] = folders_av_averaged_table[[case_name]]+av_KPI_table
+      folders_drt_averaged_table[[case_name]] = rbind(folders_drt_averaged_table[[case_name]],drt_KPI_table)
     }
 
   }
+
+  tbl_colnames = c("stat",colnames(folders_drt_averaged_table[[case_name]]))
+  result_tibble = tbl_colnames %>% purrr::map_dfc(setNames, object = list(numeric()))
+  for(stat in stats){
+    func = get(stat)
+    new_row = c(stat)
+    for(column in colnames(folders_drt_averaged_table[[case_name]])){
+           new_row = append(new_row,func(folders_drt_averaged_table[[case_name]][[column]]))
+    }
+
+    result_tibble = rbind(result_tibble,new_row)
+  }
+  colnames(result_tibble) = tbl_colnames
 
-  folders_av_averaged_table[[case_name]] = folders_av_averaged_table[[case_name]]/length(folders_seeded[[case_name]])
-
+  folders_drt_averaged_table[[case_name]] = result_tibble
 
 }
 
-print(folders_av_averaged_table)
-
-#####Write averaged av tables####
-
-dir_output_name <- "results_kpi_av"
-dir.create(dir_output_name)
+print(folders_drt_averaged_table)
 
-for(case_name in names(folders_av_averaged_table)){
-  write.table(folders_av_averaged_table[[case_name]],paste0("results_kpi_av/result_av_",case_name,".tsv"),quote = FALSE,row.names = FALSE)
+#Write averaged drt tables####
+for(case_name in names(folders_drt_averaged_table)){
+  write.table(folders_drt_averaged_table[[case_name]],paste0(path_to_data, "/kpi_summary_drt_", case_name, ".tsv"),quote = FALSE,row.names = FALSE)
 }
 
 
 
-#####Reading and averaging drt tables####
-
-
-
-folders_drt_averaged_table = list()
-
+#########################################
+#####Reading and averaging av tables#####
+folders_av_averaged_table = list()
 for(case_name in names(folders_seeded)){
-
-
-
   for(folder in folders_seeded[[case_name]]){
     files_list<- list.files(paste0(folder,"/analysis-drt-service-quality"),full.names = TRUE)
-    drt_KPI_file <- files_list[grepl(pattern = "drt_KPI.tsv",files_list)]
-    cat("processing ",drt_KPI_file," \r\n")
-    drt_KPI_table <- read.delim(drt_KPI_file)
+    av_KPI_file <- files_list[grepl(pattern = "av_KPI.tsv",files_list)]
+    cat("processing ",av_KPI_file," \r\n")
+    av_KPI_table <- read.delim(av_KPI_file)
 
-    if(!case_name %in% names(folders_drt_averaged_table)){
-      folders_drt_averaged_table[[case_name]] = drt_KPI_table
+    if(!case_name %in% names(folders_av_averaged_table)){
+      folders_av_averaged_table[[case_name]] = av_KPI_table
     }else{
-      folders_drt_averaged_table[[case_name]] = folders_drt_averaged_table[[case_name]]+drt_KPI_table
+      folders_av_averaged_table[[case_name]] = rbind(folders_av_averaged_table[[case_name]],av_KPI_table)
     }
-
   }
 
-  folders_drt_averaged_table[[case_name]] = folders_drt_averaged_table[[case_name]]/length(folders_seeded[[case_name]])
-
-
+  tbl_colnames = c("stat",colnames(folders_av_averaged_table[[case_name]]))
+  result_tibble = tbl_colnames %>% purrr::map_dfc(setNames, object = list(numeric()))
+  for(stat in stats){
+    func = get(stat)
+    new_row = c(stat)
+    for(column in colnames(folders_av_averaged_table[[case_name]])){
+      new_row = append(new_row,func(folders_av_averaged_table[[case_name]][[column]]))
+    }
+    result_tibble = rbind(result_tibble,new_row)
+  }
+  colnames(result_tibble) = tbl_colnames
+  folders_av_averaged_table[[case_name]] = result_tibble
 }
 
-print(folders_drt_averaged_table)
+print(folders_av_averaged_table)
+
+#Write averaged av tables####
+for(case_name in names(folders_av_averaged_table)){
+  write.table(folders_av_averaged_table[[case_name]],paste0(path_to_data, "/kpi_summary_av_",case_name,".tsv"),quote = FALSE,row.names = FALSE)
+}
 
-#####Write averaged drt tables####
 
-output_dir_name <- "results_kpi_drt"
-dir.create(output_dir_name)
 
-for(case_name in names(folders_drt_averaged_table)){
-  write.table(folders_drt_averaged_table[[case_name]],paste0("results_kpi_drt/result_drt_",case_name,".tsv"),quote = FALSE,row.names = FALSE)
-}