Add files via upload

Author: nakarumanchi

Restaurent Revenue Prediction/! 1 Single Tree Model/EDA - Single Tree Models.R

@@ -0,0 +1,91 @@
+require(ggplot2)
+require(caret)
+
+
+getwd()
+setwd('!!Regression/ML Process/!2 Restaurent Revenue Prediction/! 1 Single Tree Model/')
+getwd()
+training=read.csv('train.csv',header=T,sep=",",na.strings=c("NA",""," "))
+dim(training)
+str(training)
+names(training)
+head(training)
+summary(training$revenue)
+
+summary(head(training))
+str(head(training))
+
+levels(training$City)
+levels(training$City.Group)
+levels(training$Type)
+
+summary(training$P1)
+x11()
+xtabs(revenue~City,training)
+plot(training$City.Group,training$revenue)
+dotplot(revenue~Open.Date,training)
+
+?trainControl
+set.seed(1000)
+control1=trainControl(method="cv",number=30,repeats=10)
+control2=trainControl(method="boot",number=10,repeats=5)
+control3=trainControl(method="repeatedcv",number=20,repeats=5)
+
+set.seed(10001)
+cart_model1=train(training[,-c(1,2,3,4,5,43)],training[,c("revenue")],method="rpart",trControl=control1,tuneLength=20)
+cart_model1
+summary(cart_model1)
+cart_model1$method
+cart_model1$modelInfo
+cart_model1$modelType
+cart_model1$results
+cart_model1$resample
+cart_model1$bestTune
+cart_model1$call
+cart_model1$metric
+cart_model1$control
+cart_model1$finalModel
+varImp(cart_model1)
+X11()
+plot(varImp(cart_model1))
+cart_model1$finalModel$variable.importance
+plot(cart_model1$finalModel$variable.importance,type='both')
+cart_model1$yLimits
+cart_model1$perfNames
+var(cart_model1$results)
+  
+set.seed(10002)
+cart_model2=train(training[,-c(1,2,3,4,5,43)],training[,c("revenue")],method="rpart",trControl=control2,tuneLength=35)
+cart_model2
+cart_model2$results
+cart_model2$resample
+X11()
+varImp(cart_model2) 
+plot(varImp(cart_model2))
+
+set.seed(10003)
+cart_model3=train(training[,-c(1,2,3,4,5,43)],training[,c("revenue")],method="rpart",trControl=control3)
+cart_model3
+x11()
+plot(cart_model3)
+cart_model3$results
+
+
+test=read.csv('test.csv',header=T,sep=",",na.strings=c("NA"," ",""))
+dim(test)
+dim(training)
+names(test)
+names(training)
+
+levels(training$Type)=levels(test$Type)
+
+test$Prediction=predict(cart_model1,test[,-c(1,2,3,4,5,43)])
+summary(test$Prediction)
+x11()
+write.csv(test[,c("Id","Prediction")],file="rpartmodel.csv",row.names=F)
+
+
+
+
+
+