MyObject.scala

package com.mypackage.mytasks

// for the sake of example let's imagine that you have IO common functions
import com.mypackage.mytasks.IO
import org.apache.spark.sql.{SparkSession, DataFrame}

object MyObject {
  def my_transformation(dataInputPath:String, dataOutputPath:String){
    def runSparkTransformations(): Unit = {
        val spark = SparkSession.builder
        .master("local[*]")
        .getOrCreate()

        // here you are loading your data
        // let's say it's some bucket with parquet files
        // and you using some function from you project 
        // to load your data
        val rawData = IO.loadDataFromBucket(dataInputPath)

        
        // here you performing some transformations
        val transformedData = rawData
            .select(...)
            .groupby(...)
            .agg(...)
            .cache

        // and at the end you wanna store your transformed results
        IO.writeResultsToBucket(transformedData, dataOutputPath)
    }
  }
}