Wine Quality Prediction with Spark Scala and UDF

WineQualityScalaSparkPrediction.md

import org.apache.spark.sql.functions.col
import org.apache.spark.sql.types.StructType
import org.apache.spark.sql.{Column, DataFrame}
import scala.util.matching.Regex

val FirstAtRe: Regex = "^_".r
val AliasRe: Regex = "[\\s_.:@]+".r

def getFieldAlias(field_name: String): String = {
    FirstAtRe.replaceAllIn(AliasRe.replaceAllIn(field_name, "_"), "")
}

def selectFieldsNormalized(columns: List[String])(df: DataFrame): DataFrame = {
    val fieldsToSelect: List[Column] = columns.map(field =>
        col(field).as(getFieldAlias(field))
    )
    df.select(fieldsToSelect: _*)
}

def normalizeSchema(df: DataFrame): DataFrame = {
    val schema = df.columns.toList
    df.transform(selectFieldsNormalized(schema))
}

FirstAtRe = ^_
AliasRe = [\s_.:@]+




getFieldAlias: (field_name: String)String
selectFieldsNormalized: (columns: List[String])(df: org.apache.spark.sql.DataFrame)org.apache.spark.sql.DataFrame
normalizeSchema: (df: org.apache.spark.sql.DataFrame)org.apache.spark.sql.DataFrame






[\s_.:@]+

val winePath = "~/Research/mlflow-workshop/examples/wine_quality/data/winequality-red.csv"
val modelPath = "/tmp/mlflow/artifactStore/0/96cba14c6e4b452e937eb5072467bf79/artifacts/model"

winePath = ~/Research/mlflow-workshop/examples/wine_quality/data/winequality-red.csv
modelPath = /tmp/mlflow/artifactStore/0/96cba14c6e4b452e937eb5072467bf79/artifacts/model






/tmp/mlflow/artifactStore/0/96cba14c6e4b452e937eb5072467bf79/artifacts/model

val df = spark.read
              .format("csv")
              .option("header", "true")
              .option("delimiter", ";")
              .load(winePath)
              .transform(normalizeSchema)

df = [fixed_acidity: string, volatile_acidity: string ... 10 more fields]






[fixed_acidity: string, volatile_acidity: string ... 10 more fields]

%%PySpark
import mlflow
from mlflow import pyfunc

model_path = "/tmp/mlflow/artifactStore/0/96cba14c6e4b452e937eb5072467bf79/artifacts/model"
wine_quality_udf = mlflow.pyfunc.spark_udf(spark, model_path)

spark.udf.register("wineQuality", wine_quality_udf)

<function spark_udf.<locals>.predict at 0x1116a98c8>

df.createOrReplaceTempView("wines")

%%SQL
SELECT
    quality,
    wineQuality(
        fixed_acidity,
        volatile_acidity,
        citric_acid,
        residual_sugar,
        chlorides,
        free_sulfur_dioxide,
        total_sulfur_dioxide,
        density,
        pH,
        sulphates,
        alcohol
    ) AS prediction
FROM wines
LIMIT 10

+-------+------------------+
|quality|        prediction|
+-------+------------------+
|      5| 5.576883967129615|
|      5|  5.50664776916154|
|      5| 5.525504822954496|
|      6| 5.504311247097457|
|      5| 5.576883967129615|
|      5|5.5556903912725755|
|      5| 5.467882654744997|
|      7| 5.710602976324739|
|      7| 5.657319539336507|
|      5| 5.345098606538708|
+-------+------------------+

spark.catalog.listFunctions.filter('name like "%wineQuality%").show(20, false)

+-----------+--------+-----------+---------+-----------+
|name       |database|description|className|isTemporary|
+-----------+--------+-----------+---------+-----------+
|wineQuality|null    |null       |null     |true       |
+-----------+--------+-----------+---------+-----------+