Random Forest (Obesity Status 2013) GColab.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"name":"Random Forest 2013 GColab.ipynb","provenance":[{"file_id":"1GqZu6zmCy2vMNMZqO78DDuHvwv_9vJcp","timestamp":1629336269906},{"file_id":"1IzPbxNFQmmSYk9s14L4YjBfUgACn9mW2","timestamp":1593978024316},{"file_id":"1G7NKeneJNyRtcRxLVbbF9jYtRyuTOa-R","timestamp":1592749700622},{"file_id":"https://github.com/satyajitghana/TSAI-DeepVision-EVA4.0/blob/master/Utils/Colab_25GBRAM_GPU.ipynb","timestamp":1592043804148}],"collapsed_sections":[],"machine_shape":"hm"},"kernelspec":{"name":"python3","display_name":"Python 3"}},"cells":[{"cell_type":"code","metadata":{"id":"6DW42d9W02P9"},"source":["%load_ext rpy2.ipython\n","from google.colab import drive\n","drive.mount('/content/gdrive')"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"fJd4sgJP06A8"},"source":["%%R\n","install.packages(\"/content/gdrive/My Drive/Packages/memisc_0.99.27.3.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/prodlim_2019.11.13.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/ipred_0.9-11.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/recipes_0.1.16.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/bitops_1.0-7.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/gtools_3.9.2.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/caTools_1.18.2.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/gplots_3.1.1.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/ROCR_1.0-11.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/dplyr_1.0.7.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/proxy_0.4-26.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/e1071_1.7-8.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/randomForest_4.6-14.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/haven_2.4.3.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/ggplot2_3.3.5.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/abind_1.4-5.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/zoo_1.8-9.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/xts_0.12.1.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/TTR_0.24.2.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/quantmod_0.4.18.tar.gz\", repos = NULL, type = \"source\")\n","install.packages(\"/content/gdrive/My Drive/Packages/DMwR_0.4.1.tar.gz\", repos = NULL, type = \"source\")"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"G9vwTaTw08xu"},"source":["%%R\n","#Load packages\n","library(memisc)\n","library(ROCR)\n","library(dplyr)\n","library(e1071)\n","library(randomForest)\n","library(haven)\n","library(ggplot2)\n","library(DMwR)"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"9VZ2zXBr5IzK"},"source":["%%R\n","install.packages('caret')\n","library('pROC')\n","library('caret')"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"EMd0xC883du2"},"source":["%%R\n","#Load data\n","Test  <- read.csv('/content/gdrive/My Drive/Data/TEST DATA.csv')\n","Train <- read.csv('/content/gdrive/My Drive/Data/TRAIN PERCOBAAN 1.csv')"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"GSvP1rbv09a3"},"source":["%%R\n","#Load data\n","Test  <- read.csv('/content/gdrive/My Drive/Data/DATA Agustus 2021/1.TEST DATA.csv')\n","Train <- read.csv('/content/gdrive/My Drive/Data/DATA Agustus 2021/2.OR TRAIN DATA.csv')"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"VoFrlhoA2U3A"},"source":["%%R\n","names(Test)[12] <- \"Y\"\n","names(Train)[12] <- \"Y\"\n","auc_graphF <- function(TPR,FPR,AUC,name,code) {\n","\n","      H <- data.frame(FPR = FPR, TPR = TPR)\n","      X <- ggplot() + \n","        geom_line(data = H, aes(x = FPR, y = TPR)) + \n","        theme(plot.title = element_text(hjust = 0.5), plot.subtitle = element_text(hjust = 0.5), panel.grid.major = element_line(color  = \"lightgray\", \n","              size = 0.25), panel.background = element_rect(fill = \"white\", color = \"black\", size = 2)) + \n","        labs(x = \"False Positive Rate\", y = \"True Positive Rate\", title = paste(\"Area Under the ROC Curve =\",round(AUC*100,digits = 2),\"%\"), subtitle = paste(\"Random Forest with\",name,\"&\",code))\n","      \n","      ggsave(plot = X, filename = paste(\"/content/gdrive/My Drive/Result/\",name,\" \",code,\" AUC.jpg\",sep = \"\"))\n","    \n","}\n","\n","RandomForestX <- function(train, test, ntree, mtry){\n","  for (i in 1:dim(test)[2]) {\n","    train[,i] <- as.factor(train[,i])\n","    test[,i] <- as.factor(test[,i])\n","  }\n","  \n","  #Declaration metrics matrix\n","  METRIC <- matrix(0,length(ntree)*length(mtry),20)\n","  colnames(METRIC) <- c(\"Accuracy\",\"Kappa\",\"AccuracyLower\",\"AccuracyUpper\",\"AccuracyNull\",\"AccuracyPValue\",\"McnemarPValue\",\n","                        \"Sensitivity\",\"Specificity\",\"Pos Pred Value\",\"Neg Pred Value\",\"Precision\",\"Recall\",\"F1\",\"Prevalence\",\n","                        \"Detection Rate\",\"Detection Prevalence\",\"Balanced Accuracy\",\"AUC\",\"OOB\")\n","  CFM <- list()\n","  nameX <- paste(ntree,\"Trees\")\n","  mtryX <- paste(mtry,\"Mtry\")\n","  \n","  #Random Forest Process\n","  i <- 0\n","  k <- 0\n","  for (ntrees in ntree) { #Looping with different trees\n","    i <- i + 1\n","    j <- 0\n","    for (mtrys in mtry) { #Looping with different m value (selected random variable)\n","      j <- j + 1\n","      k <- k + 1\n","      #Random Forest function\n","      rf <- randomForest(Y ~., data = train, ntree = ntrees, mtry = mtrys, replace = T, samplesize = nrow(train))\n","      \n","      #Y-prediction (label)\n","      pred <- predict(rf, newdata = test[-dim(train)[2]])\n","      \n","      #Y-prediction probability\n","      pred_prob <- predict(rf, newdata = test[-dim(train)[2]], type = \"prob\")\n","      \n","      #Confusion Matrix\n","      cfmRF <- confusionMatrix(pred,test$Y)\n","      \n","      #Save the metric value\n","      METRIC[k,] <- c(cfmRF$overall,cfmRF$byClass,auc(roc(test$Y,pred_prob[,2])),mean(rf$err.rate[,1]))\n","\n","      #ROC Value for ROC & AUC Graph\n","      X <- prediction(pred_prob[,2],test$Y)\n","      R <- performance(X,\"tpr\",\"fpr\")\n","\n","      auc_graphF(R@y.values[[1]],R@x.values[[1]],METRIC[k,19],nameX[i],mtryX[j])\n","      \n","      #Save Confusion Matrix\n","      CFM[[paste(nameX[i],mtryX[j])]] <- cfmRF$table\n","      \n","      #Erase the value for save memory\n","      rf <- 0\n","      cfmRF <- 0\n","      pred <- 0\n","      pred_prob <- 0\n","      X <- 0\n","      R <- 0\n","    }\n","  }\n","\n","  write.csv(METRIC, file = \"/content/gdrive/My Drive/Result/Metric.csv\", row.names = F)\n","  write.csv(CFM, file = \"/content/gdrive/My Drive/Result/Confusion Matrix.csv\", row.names = F)\n","  return(0)\n","}"],"execution_count":null,"outputs":[]},{"cell_type":"code","metadata":{"id":"blt3tNbs4Uah"},"source":["%%R\n","result <- RandomForestX(train = Train, test = Test, ntree = c(25,50,75,100,125,150), mtry = c(1,3,5,7,9))"],"execution_count":null,"outputs":[]}]}