Minor changes

21 files changed, 102 insertions, 11 deletions

diff --git a/analysis/days.R b/analysis/days.R
index 6ad721e..601f210 100644
--- a/analysis/days.R
+++ b/analysis/days.R
@@ -1,3 +1,8 @@
+########## INFORMATIONS ##########
+# This file is made to study online classification
+# So, each pair (wireless,wakeupfor) has its classification models (knn and decision tree)
+##################################
+
 library("tidyverse")
 options(dplyr.summarise.inform = FALSE)
 library("class")
diff --git a/analysis/figures/combined.pdf b/analysis/figures/combined.pdf
index d9cbbb8..80f32bd 100644
--- a/analysis/figures/combined.pdf
+++ b/analysis/figures/combined.pdf
diff --git a/analysis/figures/dimension_coverage.pdf b/analysis/figures/dimension_coverage.pdf
index a108e34..6761219 100644
--- a/analysis/figures/dimension_coverage.pdf
+++ b/analysis/figures/dimension_coverage.pdf
diff --git a/analysis/figures/dimension_efficiency.pdf b/analysis/figures/dimension_efficiency.pdf
index df0eae8..cc4ac9b 100644
--- a/analysis/figures/dimension_efficiency.pdf
+++ b/analysis/figures/dimension_efficiency.pdf
diff --git a/analysis/figures/dimension_energy-coverage-policy.pdf b/analysis/figures/dimension_energy-coverage-policy.pdf
index d0c73e8..1bcb5cc 100644
--- a/analysis/figures/dimension_energy-coverage-policy.pdf
+++ b/analysis/figures/dimension_energy-coverage-policy.pdf
diff --git a/analysis/figures/dimension_energy-coverage-wakeupfor.pdf b/analysis/figures/dimension_energy-coverage-wakeupfor.pdf
index beacaa9..22d7539 100644
--- a/analysis/figures/dimension_energy-coverage-wakeupfor.pdf
+++ b/analysis/figures/dimension_energy-coverage-wakeupfor.pdf
diff --git a/analysis/figures/dimension_energy-coverage.pdf b/analysis/figures/dimension_energy-coverage.pdf
index 9bd624a..02d04ec 100644
--- a/analysis/figures/dimension_energy-coverage.pdf
+++ b/analysis/figures/dimension_energy-coverage.pdf
diff --git a/analysis/figures/dimension_energy.pdf b/analysis/figures/dimension_energy.pdf
index 379af96..a2f0c51 100644
--- a/analysis/figures/dimension_energy.pdf
+++ b/analysis/figures/dimension_energy.pdf
diff --git a/analysis/figures/f1_scores_offline.csv b/analysis/figures/f1_scores_offline.csv
new file mode 100644
index 0000000..108b671
--- /dev/null
+++ b/analysis/figures/f1_scores_offline.csv
@@ -0,0 +1,3 @@
+model,f1_baseline,f1_hint,f1_extended,f1_hintandextended
+knn,0.88,NA,0.89,0.91
+tree,0.93,NA,0.86,0.92
diff --git a/analysis/figures/knn.csv b/analysis/figures/knn.csv
new file mode 100644
index 0000000..3b5cc10
--- /dev/null
+++ b/analysis/figures/knn.csv
@@ -0,0 +1,2 @@
+f1_baseline
+0.905147752632288
diff --git a/analysis/figures/knn_elbow_NoHintIsFALSE.pdf b/analysis/figures/knn_elbow_NoHintIsFALSE.pdf
new file mode 100644
index 0000000..31eb996
--- /dev/null
+++ b/analysis/figures/knn_elbow_NoHintIsFALSE.pdf
diff --git a/analysis/figures/knn_elbow_NoHintIsTRUE.pdf b/analysis/figures/knn_elbow_NoHintIsTRUE.pdf
new file mode 100644
index 0000000..96072cc
--- /dev/null
+++ b/analysis/figures/knn_elbow_NoHintIsTRUE.pdf
diff --git a/analysis/figures/random_inputs_NoHintIsFALSE.pdf b/analysis/figures/random_inputs_NoHintIsFALSE.pdf
new file mode 100644
index 0000000..3795adc
--- /dev/null
+++ b/analysis/figures/random_inputs_NoHintIsFALSE.pdf
diff --git a/analysis/figures/random_inputs_NoHintIsTRUE.pdf b/analysis/figures/random_inputs_NoHintIsTRUE.pdf
new file mode 100644
index 0000000..c5a7506
--- /dev/null
+++ b/analysis/figures/random_inputs_NoHintIsTRUE.pdf
diff --git a/analysis/figures/sim_dimension_coverage_NO_HINT.pdf b/analysis/figures/sim_dimension_coverage_NO_HINT.pdf
index a946b86..b9b9519 100644
--- a/analysis/figures/sim_dimension_coverage_NO_HINT.pdf
+++ b/analysis/figures/sim_dimension_coverage_NO_HINT.pdf
diff --git a/analysis/figures/sim_dimension_coverage_WITH_HINT.pdf b/analysis/figures/sim_dimension_coverage_WITH_HINT.pdf
index 0f42067..a226cf0 100644
--- a/analysis/figures/sim_dimension_coverage_WITH_HINT.pdf
+++ b/analysis/figures/sim_dimension_coverage_WITH_HINT.pdf
diff --git a/analysis/figures/sim_dimension_energy_NO_HINT.pdf b/analysis/figures/sim_dimension_energy_NO_HINT.pdf
index 0f32541..d5ab543 100644
--- a/analysis/figures/sim_dimension_energy_NO_HINT.pdf
+++ b/analysis/figures/sim_dimension_energy_NO_HINT.pdf
diff --git a/analysis/figures/sim_dimension_energy_WITH_HINT.pdf b/analysis/figures/sim_dimension_energy_WITH_HINT.pdf
index 7e411e5..590bc80 100644
--- a/analysis/figures/sim_dimension_energy_WITH_HINT.pdf
+++ b/analysis/figures/sim_dimension_energy_WITH_HINT.pdf
diff --git a/analysis/figures/tree_FALSE.pdf b/analysis/figures/tree_FALSE.pdf
new file mode 100644
index 0000000..fd9f08d
--- /dev/null
+++ b/analysis/figures/tree_FALSE.pdf
diff --git a/analysis/figures/tree_TRUE.pdf b/analysis/figures/tree_TRUE.pdf
new file mode 100644
index 0000000..56466fc
--- /dev/null
+++ b/analysis/figures/tree_TRUE.pdf
diff --git a/analysis/learning.R b/analysis/learning.R
index d21b8cd..ff31d33 100644
--- a/analysis/learning.R
+++ b/analysis/learning.R
@@ -1,8 +1,10 @@
 library("tidyverse")
+options(dplyr.summarise.inform = FALSE)
 library("class")
 library("rpart")
 library("rpart.plot")
 library("viridis")
+library("MLmetrics")
 
 ## Simulation Parameters:
 ## simkey         {baseline,extended,hint,hintandextended}
@@ -24,16 +26,95 @@ nsimkey=4
 nsimulations=nseed*nwakeupfor*nwireless*nsimkey # Must be 3200
 
 ## Load data
-data=read_csv("../CCGRID2022.csv")%>%distinct() # Note that in the data experiment wireless=="lora",seed==1,wakeupfor==60,simkey=="baseline" is present 2 times in the CSV file
-tmp_data_coverage=data%>%group_by(simkey,wireless,wakeupfor,seed)%>%mutate(coverage=sum(nDataRcv))%>%ungroup()%>%filter(isSender==1)%>%select(simkey,wireless,wakeupfor,seed,coverage)
-data_seed_isSender=data%>%group_by(simkey,wireless,wakeupfor,seed,isSender)%>%summarize(energy_mean=mean(energy))%>%
-    left_join(tmp_data_coverage,by=c("simkey","wireless","wakeupfor","seed"))%>%
-    mutate(efficiency=energy_mean/coverage)%>%
-    ungroup()
+data=suppressMessages(read_csv("../CCGRID2022.csv"))%>%distinct() # Note that in the data experiment wireless=="lora",seed==1,wakeupfor==60,simkey=="baseline" is present 2 times in the CSV file
 data_seed=data%>%group_by(simkey,wireless,wakeupfor,seed)%>%summarize(energy=sum(energy),coverage=sum(nDataRcv))%>%
     mutate(efficiency=energy/coverage)%>%
     ungroup()
 
+
+F1_Score2=function(truth, pred){
+    result=sapply(c("baseline","extended","hint","hintandextended"),function(c){
+        cur_truth=truth[truth==c]
+        cur_pred=pred[truth==c]
+        col=paste0("f1_",c)
+        score=F1_Score(cur_truth,cur_pred)
+        if(is.nan(score)){score=0}
+        list(tibble(!!col:=score))
+    })
+    do.call("cbind",result)
+}
+
+build_models=function(ignore_hint=TRUE){
+    ## Prepare data for traning
+    set.seed(1) # Reproducibility
+    wireless_map=c("lora"=1,"nbiot"=2)
+    data_ml=data_seed%>%select(-efficiency,-seed)%>%mutate(wireless=wireless_map[data_seed$wireless])
+    if(ignore_hint){
+        data_ml=data_ml%>%filter(simkey!="hint")
+    }
+    train_set=data_ml%>%sample_frac(0.8)     # 80% of the data
+    test_set=data_ml%>%suppressMessages(anti_join(train_set))  # 20% of the data
+
+    ## KNN predict function
+    knn_fn=function(inputs){
+        as.vector(knn(train=train_set%>%select(-simkey),test=inputs%>%select(-simkey),cl=train_set$simkey,k=10))
+    }
+    
+    ## Decision tree
+    tree=rpart(
+        simkey ~ wireless + wakeupfor + energy + coverage,
+        data=train_set,
+        method="class",
+        minsplit=60,
+        minbucket=1)
+    ## Tree predict function
+    tree_fn=function(inputs){
+        as.vector(predict(tree,newdata=inputs%>%select(-simkey),type="class"))
+    }
+
+    ## Build models
+    models=list(predict_knn=knn_fn,predict_tree=tree_fn)
+
+    ## Computer performances
+    perfs=sapply(seq(1,20),function(i){
+        ## Prepare data for traning
+        set.seed(1) # Reproducibility
+        wireless_map=c("lora"=1,"nbiot"=2)
+        data_ml=data_seed%>%select(-efficiency,-seed)%>%mutate(wireless=wireless_map[data_seed$wireless])
+        if(ignore_hint){
+            data_ml=data_ml%>%filter(simkey!="hint")
+        }
+        train_set=data_ml%>%sample_frac(0.8)     # 80% of the data
+        test_set=data_ml%>%suppressMessages(anti_join(train_set))  # 20% of the data
+
+        ## KNN
+        knn_predictions=as.vector(knn(train=train_set%>%select(-simkey),test=test_set%>%select(-simkey),cl=train_set$simkey,k=10))
+        ## Decision tree
+        tree=rpart(
+            simkey ~ wireless + wakeupfor + energy + coverage,
+            data=train_set,
+            method="class",
+            minsplit=60,
+            minbucket=1)
+        tree_predictions=as.vector(predict(tree,newdata=test_set%>%select(-simkey),type="class"))
+
+        ## Prefs
+        f1_knn=F1_Score2(test_set$simkey,knn_predictions)
+        f1_tree=F1_Score2(test_set$simkey,tree_predictions)
+        list(cbind(tibble(model=c("knn","tree")),rbind(f1_knn,f1_tree)))
+    })
+    perfs=do.call("rbind",perfs)%>%mutate_if(is.numeric, ~round(.,digits=2))
+    perfs=perfs%>%group_by(model)%>%summarize(
+        f1_baseline=mean(f1_baseline),
+        f1_hint=mean(f1_hint),
+        f1_extended=mean(f1_extended),
+        f1_hintandextended=mean(f1_hintandextended))
+    write.csv(perfs,"figures/f1_scores_offline.csv",quote=FALSE,row.names=FALSE)
+              
+    ## Return models
+    models
+}
+
 generate_inputs=function(ignore_hint=FALSE) {
     ## Prepare data for traning
     set.seed(1) # Reproducibility
@@ -61,7 +142,7 @@ generate_inputs=function(ignore_hint=FALSE) {
         minbucket=1)
     tree_predictions=predict(tree,newdata=test_set%>%select(-simkey),type="class")
     tree_cont_table=table(tree_predictions,test_set$simkey)
-    tree_accuracy=round((sum(diag(tree_cont_table)/sum(rowSums(tree_cont_table))))*100)
+    tree_accuracy=(sum(diag(tree_cont_table)/sum(rowSums(tree_cont_table))))
     tree_prop_table=round(prop.table(tree_cont_table),digits=2)
 
     ## Elbow plot
@@ -69,12 +150,12 @@ generate_inputs=function(ignore_hint=FALSE) {
         knn_predictions=knn(train=train_set%>%select(-simkey),test=test_set%>%select(-simkey),cl=train_set$simkey,k=kvalue)
         ## KNN analysis
         knn_cont_table=table(knn_predictions,test_set$simkey)
-        knn_accuracy=round((sum(diag(knn_cont_table)/sum(rowSums(knn_cont_table))))*100)
+        knn_accuracy=(sum(diag(knn_cont_table)/sum(rowSums(knn_cont_table))))
         knn_prop_table=round(prop.table(knn_cont_table),digits=2)
         tibble(k=kvalue,accuracy=knn_accuracy)
     })
     elbow_data=do.call("rbind",elbow_data)
-    ggplot(data=elbow_data,aes(k,accuracy))+geom_point()+geom_line()+ggtitle(paste("K-elbow for with NoHint to",as.character(ignore_hint)))
+    ggplot(data=elbow_data,aes(k,accuracy))+geom_point()+geom_line()+ggtitle(paste("K-elbow for with NoHint to",as.character(ignore_hint)))+ylim(c(0,1))
     ggsave(paste0("figures/knn_elbow_NoHintIs",as.character(ignore_hint),".pdf"))
     
     ## Prints
@@ -126,5 +207,5 @@ generate_inputs=function(ignore_hint=FALSE) {
 }
 
 ## Generate inputs
-generate_inputs(FALSE)
-generate_inputs(TRUE)
+#generate_inputs(FALSE)
+#generate_inputs(TRUE)