DSP Project first push, date: 29/01/2026

r_scripts/descriptive_stats.R

@@ -0,0 +1,83 @@
+#!/usr/bin/env Rscript
+
+`%||%` <- function(lhs, rhs) {
+  if (is.null(lhs) || is.na(lhs)) rhs else lhs
+}
+
+suppressPackageStartupMessages({
+  library(jsonlite)
+  library(readr)
+  library(dplyr)
+})
+
+args <- commandArgs(trailingOnly = TRUE)
+
+if (length(args) < 2) {
+  stop("Expected arguments: <data_file_path> <params_json>")
+}
+
+data_file_path <- args[1]
+params_json <- args[2]
+
+if (!file.exists(data_file_path)) {
+  stop(sprintf("Data file not found at %s", data_file_path))
+}
+
+params <- tryCatch(
+  fromJSON(params_json),
+  error = function(e) list()
+)
+
+read_opts <- list(
+  locale = locale(encoding = params$encoding %||% "UTF-8")
+)
+
+data <- tryCatch(
+  read_csv(
+    file = data_file_path,
+    show_col_types = FALSE,
+    progress = FALSE,
+    guess_max = params$guess_max %||% 1000,
+    locale = read_opts$locale
+  ),
+  error = function(e) stop(sprintf("Failed to read CSV: %s", e$message))
+)
+
+numeric_columns <- select(data, where(is.numeric))
+
+column_stats <- lapply(numeric_columns, function(column) {
+  list(
+    count = sum(!is.na(column)),
+    mean = mean(column, na.rm = TRUE),
+    median = median(column, na.rm = TRUE),
+    sd = sd(column, na.rm = TRUE),
+    min = min(column, na.rm = TRUE),
+    max = max(column, na.rm = TRUE),
+    missing = sum(is.na(column))
+  )
+})
+
+message_text <- if (length(column_stats) > 0) {
+  "Descriptive statistics generated successfully."
+} else {
+  "No numeric columns detected; returning dataset preview only."
+}
+
+preview_rows <- head(data, n = min(5, nrow(data)))
+
+output <- list(
+  message = message_text,
+  numeric_columns = column_stats,
+  sample_rows = preview_rows,
+  params = params
+)
+
+cat(
+  toJSON(
+    output,
+    pretty = TRUE,
+    auto_unbox = TRUE,
+    na = "null",
+    dataframe = "rows"
+  )
+)