from pyspark.sql import SparkSessionfrom pyspark.sql$.$functions import col, round, max, avg# Initialize SparkSessionspark $=$ SparkSession.builder $\backslash .$appName$("$ChargePointsETLJob$")\backslash .$getOrCreate$()$# Input and Output pathsinput$\_$path $=$ "input$\_$path$/$electric$-$chargepoints$-2017.$csv$"$output$\_$path $=$ "output$\_$path$/$chargepoints$\_$summary.parquet"# Read CSV file into DataFramedf $=$ spark.read $\backslash .$option$("$header$",$ "true"$)\backslash .$option$("$inferSchema$",$ "true"$)\backslash .$csv$($input$\_$path$)$# Clean up column names and select relevant columnscleaned$\_$df $=$ df$.$select$($ col$("$CPID$").$alias$("$chargepoint$\_$id$"),$ col$("$Duratio n$").$cast$("$double$").$alias$("$duration$"))$# Calculate max and average duration grouped by chargepoint$\_$idsummary$\_$df $=$ cleaned$\_$df$.$groupBy$("$chargepoint$\_$id$")\backslash .$agg$($ round$($max$("$duration$"),2).$alias$("$max$\_$duration"$),$ round$($avg$("$duration$"),2).$alias$("$avg$\_$duration"$))$# Write summary$\_$df as Parquet filesummary$\_$df$.$write.mode$("$overwrite$").$parquet$($output$\_$path$)$# Stop SparkSessionspark.stop$()$