1 files changed, 31 insertions, 12 deletions
diff --git a/analysis/knn.R b/analysis/knn.R
index 71771f1..3cfa2a9 100644
--- a/analysis/knn.R
+++ b/analysis/knn.R
@@ -1,5 +1,8 @@
 library("tidyverse")
 library("class")
+library("rpart")
+library("rpart.plot")
+library("viridis")
 
 ## Simulation Parameters:
 ## simkey         {baseline,extended,hint,hintandextended}
@@ -32,18 +35,34 @@ data_seed=data%>%group_by(simkey,wireless,wakeupfor,seed)%>%summarize(energy=sum
     ungroup()
 
 
-## Prepare data for knn
+## Prepare data for traning
 set.seed(1) # Reproducibility
-data_seed=data_seed%>%select(-efficiency,-seed)%>%mutate(wireless=as.numeric(as.factor(data_seed$wireless)))
-
-## Train
+data_seed=data_seed%>%select(-efficiency,-seed)%>%mutate(wireless=as.numeric(as.factor(data_seed$wireless)))#%>%filter(simkey!="hint")
 train_set=data_seed%>%sample_frac(0.8)     # 80% of the data
 test_set=data_seed%>%anti_join(train_set)  # 20% of the data
-classifier=knn(train=train_set%>%select(-simkey),test=test_set%>%select(-simkey),cl=train_set$simkey,k=10)
-
-## Analysis
-cont_table=table(classifier,test_set$simkey)
-accuracy=round((sum(diag(cont_table)/sum(rowSums(cont_table))))*100)
-prop_table=round(prop.table(cont_table),digits=2)
-print(prop_table)
-print(paste0("Overall KNN accuracy ",accuracy,"%"))
+
+## KNN training
+knn_predictions=knn(train=train_set%>%select(-simkey),test=test_set%>%select(-simkey),cl=train_set$simkey,k=10)
+## KNN analysis
+knn_cont_table=table(knn_predictions,test_set$simkey)
+knn_accuracy=round((sum(diag(knn_cont_table)/sum(rowSums(knn_cont_table))))*100)
+knn_prop_table=round(prop.table(knn_cont_table),digits=2)
+
+## Decision tree
+tree=rpart(
+    simkey ~ wireless + wakeupfor + energy + coverage,
+    data=train_set,
+    method="class",
+    minsplit=60,
+    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_prop_table=round(prop.table(tree_cont_table),digits=2)
+
+## Prints
+print(paste0("Accuracy: KNN=",knn_accuracy,"% CART=",tree_accuracy,"%"))
+pdf("figures/tree.pdf")
+tree_plot=rpart.plot(tree,box.palette=as.list(viridis::viridis(4,begin=0.48)))
+dev.off()
+## Notes: KNN accuracy jump to 76% and CART to 80% accuracy without the hint policy