Created
October 6, 2021 11:34
-
-
Save rzykov/847f97d48e14da2b0e432af442af16e7 to your computer and use it in GitHub Desktop.
DataAnalysisIntro6.scala
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
//CODE of MIC | |
import data.VarPairData | |
import mine.core.MineParameters | |
import analysis.Analysis | |
import analysis.results.BriefResult | |
import scala.util.Random | |
//Code a discrete value by randomly changing the order of the "codes | |
def encode(col: Array[String]): Array[Double] = { | |
val ns = scala.util.Random.shuffle(1 to col.toSet.size) | |
val encMap = col.toSet.zip(ns).toMap | |
col.map{encMap(_).toDouble} | |
} | |
// function to calculate MIC | |
def mic(x: Array[Double], y: Array[Double]) = { | |
val data = new VarPairData(x.map(_.toFloat), y.map(_.toFloat)) | |
val params = new MineParameters(0.6.toFloat, 15, 0, null) | |
val res = Analysis.getResult(classOf[BriefResult], data, params) | |
res.getMIC | |
} | |
//in case of a discrete value do a lot of iterations and take the maximum | |
def micMax(x: Array[Double], y: Array[Double], n: Int = 100) = | |
(for{ i <- 1 to 100} yield mic(x, y)).max | |
//Now we're close to the final result, let's perform the calculation: | |
val aov = dataAov.filter(x => interestedBrowsers.contains(x.osFamily)) //we leave only the OSes we want | |
.filter(_.categoryId == 128) // filter categories | |
//osFamily | |
var aovMic = aov.map(x => (x.osFamily, x.aov)).collect() | |
println("osFamily MIC =" + micMax(encode(aovMic.map(_._1)), aovMic.map(_._2)) | |
//orderId | |
aovMic = aov.map(x => (x.orderId, x.aov)).collect() | |
println("orderId MIC =" + micMax(encode(aovMic.map(_._1)), aovMic.map(_._2)) | |
//cityId | |
aovMic = aov.map(x => (x.cityId, x.aov)).collect() | |
println("cityId MIC =" + micMax(encode(aovMic.map(_._1)), aovMic.map(_._2)) | |
//uaName | |
aovMic = aov.map(x => (x.uaName, x.aov)).collect() | |
println("uaName MIC =" + mic(encode(aovMic.map(_._1)), aovMic.map(_._2)) | |
//aov | |
println("aov MIC =" + micMax(aovMic.map(_._2), aovMic.map(_._2)) | |
//random | |
println("random MIC =" + mic(aovMic.map(_ => math.random*100.0), aovMic.map(_._2)) | |
//OUTPUT | |
// osFamily MIC =0.06658 | |
// orderId MIC =0.10074 | |
// cityId MIC =0.07281 | |
// aov MIC =0.99999 | |
// uaName MIC =0.05297 | |
// random MIC =0.10599 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment