output3modClass.txt

R version 4.0.2 (2020-06-22) -- "Taking Off Again"
Copyright (C) 2020 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

  Natural language support but running in an English locale

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> load("/research/rgs01/home/clusterHome/qtran/Semisupervised_Learning/label_data_x.RData")
> load("/research/rgs01/home/clusterHome/qtran/Semisupervised_Learning/class_label_data_y.RData")
> source("/research/rgs01/home/clusterHome/qtran/Semisupervised_Learning/Rscripts/supporting_functions.R")
> source("/research/rgs01/home/clusterHome/qtran/Semisupervised_Learning/Rscripts/predictOneNN.R")
> ###Use 70% of the Capper Reference for training
> ##first get the index of the labeled data
> acc = data.frame("Train_function" = "", "Learner"="", "Seed" = "", "Ind_ACC"="", "Trans_ACC" = "")
> main_dir = "/research/rgs01/home/clusterHome/qtran/Semisupervised_Learning_big_private_data"
> #main_dir = "/Volumes/qtran/Semisupervised_Learning_big_private_data/"
> 
> 
> library(caret)
> library(ssc)
> library(kernlab)
> create_class_score_output = function(pred_label, pred_score, truth_label){
+   output = data.frame(pred_label)
+   output$Sample = names(truth_label)
+   output$pred_score = apply(pred_score, 1, max)
+   output$truth_label = truth_label
+   
+   return(output)
+ }
> 
> seeds = c(1, 2, 20, 40, 80, 160, 320)
> 
> for (s in seeds){
+   print(paste0("Start with seed = ", s))
+   
+   sub_dir = paste0("seed", s, "_all")
+   output_tbl_dir <- file.path(main_dir, "processed_data/class_results", sub_dir)
+   output_fig_dir <- file.path(main_dir, "figures/class_results", sub_dir)
+   if (!dir.exists(output_tbl_dir)){
+     dir.create(output_tbl_dir)
+   } else {
+     print("Table Dir already exists!")
+   }
+   if (! dir.exists(output_fig_dir)){
+     dir.create(output_fig_dir)
+   } else {
+     print("Figure Dir already exists!")
+   }
+   
+   
+   load(paste0(main_dir, "/processed_data/class_results/seed", s, "_alldata_sets.RData"))
+   
+   tra.idx = x$tra.idx
+   
+   #tra.idx = get_sample_n_rows_idx(label_data_x, label_data_y, n=nums$min_class, replace=FALSE)
+   ####make train, test data sets#######
+   xtrain = NULL
+   ytrain = NULL
+   xtrain <- x$xtrain
+   ytrain <- x$ytrain
+   
+   ###Use 50% of train instances as unlabeled set
+   tra.na.idx <- x$tra.na.idx
+   ytrain[tra.na.idx] <- NA
+   
+   #Use the other 30% of instances for inductive test
+  
+   xitest <- x$xitest # test instances
+   yitest <- x$yitest # classes of instances in xitest
+   
+   # Use the unlabeled examples for transductive test
+   xttest <- x$xttest # transductive test instances 
+   yttest <- x$yttest # classes of instances in xttest
+   
+   #############base line supervised learning classifier SVM##########
+ 
+   ################ computing distance and kernel matrices
+   dtrain <- as.matrix(proxy::dist(x = xtrain, method = "euclidean", by_rows = TRUE)) 
+   ditest <- as.matrix(proxy::dist(x = xitest, y = xtrain, method = "euclidean", by_rows = TRUE))
+   dttest <- as.matrix(proxy::dist(x = xttest, y = xtrain, method = "euclidean", by_rows = TRUE))
+   ###Set up learner parameters 
+   learner <- e1071::svm
+   learner.pars <- list(type = "C-classification", kernel="radial", 
+                        probability = TRUE, scale = TRUE)
+   pred <- function(m, x){
+     r <- predict(m, x, probability = TRUE)
+     prob <- attr(r, "probabilities")
+     return(prob)
+   }
+   
+   #Training from a set of distances with 1-NN as base classifier
+   set.seed(3)
+   m.selft1 <- selfTraining(x = dtrain, y = ytrain, x.inst = FALSE, 
+                            learner = oneNN, 
+                            pred = "predict", 
+                            pred.pars = list(type = "prob", distance.weighting = "none"))
+ 
+   #Training with learning based on nearest neighbor rule and cut edges
+   set.seed(3)
+   m.snnrce <- snnrce(x = xtrain, y = ytrain, dist = "Euclidean")
+   
+   ###Training with SElf-TRaining with EDiting. SETRED uses an amending scheme to avoid the introduction of noisy examples into the enlarged labeled set. 
+   #For each iteration, the mislabeled examples are identified using the local information provided by the neighborhood graph.
+   set.seed(3)
+   m.setred <- setred(x = dtrain, y = ytrain, x.inst=FALSE, 
+                      learner = ssc::oneNN, pred = "predict",
+                      pred.pars = list(type = "prob", distance.weighting = "none"))
+ 
+ ####Prediction###3
+   m = list(m.selft1,  m.setred,  m.snnrce)
+   names(m) = c("m.selft1", "m.setred",  "m.snnrce")
+   rlist::list.save(m, file = paste0(output_tbl_dir, 'models_selft_setred_snnrce_1NN.RData'))
+   
+   j = 0
+   i_label_data = as.data.frame(yitest)
+   colnames(i_label_data) = "TRUTH"
+   t_label_data = as.data.frame(yttest)
+   colnames(t_label_data) = "TRUTH"
+   
+   source("/rgs01/home/clusterHome/qtran/Semisupervised_Learning/Rscripts/predict_setred_prob.R")
+   for (i in m){
+     trainer = attr(i, 'class')
+     learner = attr(i$model, 'class')
+     if (is.null(learner)){
+       learner = j + 1
+     }
+     
+     ###perform prediction
+     ipred = get_inductive_pred(i, test_x=xitest, test_y=yitest, train_x=xtrain)
+     transpred = get_trans_pred(i, train_x = xtrain, trans_y=yttest, tra.na.idx)
+     ###create confusion matrices
+     iconM = createConfusionMatrix(ipred, yitest, mode = "everything")
+     transconM = createConfusionMatrix(transpred, yttest, mode="everything")
+     
+     write.csv(iconM$byClass, file=paste0(output_tbl_dir,"/Inductive_metrics_", trainer, "_", learner, ".csv"))
+     write.csv(transconM$byClass, file=paste0(output_tbl_dir, "/Transductive_metrics_", trainer, "_", learner, ".csv"))
+     
+     ####plot confusion matrices
+     iPT = create_plotTable(iconM)
+     plot_CM(iPT, filename = paste0(output_fig_dir, "/Inductive_", trainer, "_", learner, ".pdf"))
+     pPT = create_plotTable(transconM)
+     plot_CM(pPT, filename = paste0(output_fig_dir, "/Transductive_", trainer, "_", learner, ".pdf"))
+     
+     ###accuracy
+     acc = rbind(acc, c(trainer, learner, s, iconM$overall[1], transconM$overall[1]))
+     write.csv(acc, file =paste0(output_tbl_dir, "/ACC_ALL_SELFT_SNNRCE_1NN.csv"))
+     
+     ###get predicted labels
+     j = j + 1
+     i_label_data = cbind(i_label_data, createLabelData(rownames(xitest), ipred)$class)
+     colnames(i_label_data)[j+1] = paste0(trainer, "_", learner) 
+     t_label_data = cbind(t_label_data, createLabelData(rownames(xttest), transpred)$class)
+     colnames(t_label_data)[j+1] = paste0(trainer, "_", learner) 
+     
+     
+     ###get predicted probabilities
+     print(paste("j=",j))
+     #
+     if(trainer =="selfTraining" || trainer =="snnrce" || trainer =="setred"){
+       ipred_score = predProb.OneNN(i, ditest[, i$instances.index])
+       rownames(ipred_score) = rownames(ditest)
+       tpred_score = predProb.OneNN(i, dttest[, i$instances.index])
+       rownames(tpred_score) = rownames(dttest)
+       write.csv(ipred_score, file=paste0(output_tbl_dir,"/Inductive_scores_", trainer, "_", learner, ".csv"), row.names=TRUE)
+       write.csv(tpred_score, file = paste0(output_tbl_dir,"/Transductive_scores_", trainer, "_", learner, ".csv"), row.names=TRUE)
+       
+       ioutscores = create_class_score_output(ipred, ipred_score, yitest)
+       toutscores = create_class_score_output(transpred, tpred_score, yttest)
+       write.csv(ioutscores, file=paste0(output_tbl_dir,"/Inductive_MaxScores_", trainer, "_", learner, ".csv"), row.names=TRUE)
+       write.csv(toutscores, file = paste0(output_tbl_dir,"/Transductive_MaxScores_", trainer, "_", learner, ".csv"), row.names=TRUE)
+     }
+   }
+   #write.csv(acc, file =paste0(output_tbl_dir, "/ACC_SELFT.csv"))
+   write.csv(i_label_data, file=paste0(output_tbl_dir, "/Inductive_predicted_SELFT_SNNRCE_SETRED_labels.csv"), row.names=TRUE)
+   write.csv(t_label_data, file=paste0(output_tbl_dir, "/Transductive_predicted_SELFT_SNNRCE_SETRED_labels.csv"), row.names=TRUE)
+   
+   print(paste0("Complete run for seed = ", s))
+ }
[1] "Start with seed = 1"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 1"
[1] "Start with seed = 2"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 2"
[1] "Start with seed = 20"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 20"
[1] "Start with seed = 40"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 40"
[1] "Start with seed = 80"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 80"
[1] "Start with seed = 160"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 160"
[1] "Start with seed = 320"
[1] "Table Dir already exists!"
[1] "Figure Dir already exists!"
[1] "Inductive testing for selfTraining using OneNN"
[1] "Transductive testing for selfTraining using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 1"
[1] "Inductive testing for setred using OneNN"
[1] "Transductive testing for setred using OneNN"
[1] "dtrain"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 2"
[1] "Inductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Transductive testing for snnrce using OneNN"
[1] "SNNRCE"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "Create table for plotting"
[1] "Plot CM"
[1] "j= 3"
[1] "Complete run for seed = 320"
> 

------------------------------------------------------------
Sender: LSF System <lsfadmin@noderome159>
Subject: Job 157310123: <threemod_class> in cluster <hpcf_research_cluster> Done

Job <threemod_class> was submitted from host <splprhpc05> by user <qtran> in cluster <hpcf_research_cluster> at Thu Apr 14 13:29:54 2022
Job was executed on host(s) <noderome159>, in queue <standard>, as user <qtran> in cluster <hpcf_research_cluster> at Thu Apr 14 13:29:56 2022
</home/qtran> was used as the home directory.
</home/qtran> was used as the working directory.
Started at Thu Apr 14 13:29:56 2022
Terminated at Thu Apr 14 16:04:28 2022
Results reported at Thu Apr 14 16:04:28 2022

Your job looked like:

------------------------------------------------------------
# LSBATCH: User input
R --vanilla < /research/rgs01/home/clusterHome/qtran/Semisupervised_Learning/Rscripts/run_SELFT_SNNRCE_SETRED_1NN_subclass.R
------------------------------------------------------------

Successfully completed.

Resource usage summary:

    CPU time :                                   9245.85 sec.
    Max Memory :                                 1241 MB
    Average Memory :                             1032.11 MB
    Total Requested Memory :                     16382.00 MB
    Delta Memory :                               15141.00 MB
    Max Swap :                                   -
    Max Processes :                              6
    Max Threads :                                7
    Run time :                                   9268 sec.
    Turnaround time :                            9274 sec.

The output (if any) is above this job summary.



PS:

Read file </research/rgs01/home/clusterHome/qtran/Semisupervised_Learning/error3modsClass.txt> for stderr output of this job.