From c2affb00ff404613f45b51cd97b50773982fde5f Mon Sep 17 00:00:00 2001 From: Loic Guegan Date: Fri, 11 Nov 2022 15:47:19 +0100 Subject: Minor changes --- analysis/figures/random_inputs.pdf | Bin 158419 -> 150895 bytes analysis/figures/tree.pdf | Bin 9593 -> 9593 bytes analysis/inputs.csv | 81 ---------------------------- analysis/knn.R | 107 ------------------------------------- analysis/learning.R | 107 +++++++++++++++++++++++++++++++++++++ 5 files changed, 107 insertions(+), 188 deletions(-) delete mode 100644 analysis/inputs.csv delete mode 100644 analysis/knn.R create mode 100644 analysis/learning.R (limited to 'analysis') diff --git a/analysis/figures/random_inputs.pdf b/analysis/figures/random_inputs.pdf index 8270126..817d22b 100644 Binary files a/analysis/figures/random_inputs.pdf and b/analysis/figures/random_inputs.pdf differ diff --git a/analysis/figures/tree.pdf b/analysis/figures/tree.pdf index 3ed995b..e526c80 100644 Binary files a/analysis/figures/tree.pdf and b/analysis/figures/tree.pdf differ diff --git a/analysis/inputs.csv b/analysis/inputs.csv deleted file mode 100644 index c144b6e..0000000 --- a/analysis/inputs.csv +++ /dev/null @@ -1,81 +0,0 @@ -"wireless","wakeupfor","energy_constraint","coverage_constraint","simkey","model" -"lora",60,8732.48785137262,10,"hintandextended","knn" -"lora",60,7760.02858805553,4,"baseline","knn" -"lora",60,7647.55841422485,11,"baseline","knn" -"lora",60,7849.4323163005,8,"extended","knn" -"lora",60,7531.65704064597,10,"baseline","knn" -"lora",60,8524.20092750562,9,"hintandextended","knn" -"lora",60,8744.00524232863,1,"hintandextended","knn" -"lora",60,8426.54290125995,1,"hintandextended","knn" -"lora",60,7631.27743269689,1,"baseline","knn" -"lora",60,7661.19496849046,5,"hint","knn" -"lora",60,8732.48785137262,10,"hintandextended","tree" -"lora",60,7760.02858805553,4,"extended","tree" -"lora",60,7647.55841422485,11,"extended","tree" -"lora",60,7849.4323163005,8,"extended","tree" -"lora",60,7531.65704064597,10,"extended","tree" -"lora",60,8524.20092750562,9,"hintandextended","tree" -"lora",60,8744.00524232863,1,"baseline","tree" -"lora",60,8426.54290125995,1,"baseline","tree" -"lora",60,7631.27743269689,1,"hint","tree" -"lora",60,7661.19496849046,5,"extended","tree" -"lora",180,29123.7166634847,4,"hint","knn" -"lora",180,29655.7737664565,3,"hint","knn" -"lora",180,23645.1436280268,11,"extended","knn" -"lora",180,24938.4341620644,0,"hintandextended","knn" -"lora",180,29862.9781404826,5,"hint","knn" -"lora",180,27307.1859899517,6,"hint","knn" -"lora",180,23232.7369937696,3,"baseline","knn" -"lora",180,25404.8205478179,7,"hintandextended","knn" -"lora",180,26553.2369356217,8,"hintandextended","knn" -"lora",180,24432.1396842401,8,"hintandextended","knn" -"lora",180,29123.7166634847,4,"hint","tree" -"lora",180,29655.7737664565,3,"baseline","tree" -"lora",180,23645.1436280268,11,"extended","tree" -"lora",180,24938.4341620644,0,"baseline","tree" -"lora",180,29862.9781404826,5,"hint","tree" -"lora",180,27307.1859899517,6,"hint","tree" -"lora",180,23232.7369937696,3,"baseline","tree" -"lora",180,25404.8205478179,7,"hintandextended","tree" -"lora",180,26553.2369356217,8,"hint","tree" -"lora",180,24432.1396842401,8,"hintandextended","tree" -"nbiot",60,9452.16352643334,9,"hint","knn" -"nbiot",60,9465.21074816174,3,"hint","knn" -"nbiot",60,8456.21622429176,1,"hintandextended","knn" -"nbiot",60,9570.38342273607,3,"hint","knn" -"nbiot",60,8855.86319130787,12,"hintandextended","knn" -"nbiot",60,8444.57516620697,3,"hintandextended","knn" -"nbiot",60,8527.76385865066,4,"hintandextended","knn" -"nbiot",60,8685.2330654384,8,"hintandextended","knn" -"nbiot",60,8744.42417828556,8,"hintandextended","knn" -"nbiot",60,8664.97570233729,0,"hintandextended","knn" -"nbiot",60,9452.16352643334,9,"hintandextended","tree" -"nbiot",60,9465.21074816174,3,"baseline","tree" -"nbiot",60,8456.21622429176,1,"baseline","tree" -"nbiot",60,9570.38342273607,3,"baseline","tree" -"nbiot",60,8855.86319130787,12,"hintandextended","tree" -"nbiot",60,8444.57516620697,3,"baseline","tree" -"nbiot",60,8527.76385865066,4,"hintandextended","tree" -"nbiot",60,8685.2330654384,8,"hintandextended","tree" -"nbiot",60,8744.42417828556,8,"hintandextended","tree" -"nbiot",60,8664.97570233729,0,"baseline","tree" -"nbiot",180,23875.3493146438,7,"hintandextended","knn" -"nbiot",180,26217.4338214212,12,"hint","knn" -"nbiot",180,23369.7590769622,9,"extended","knn" -"nbiot",180,28487.7830938115,8,"hint","knn" -"nbiot",180,27871.6472379192,11,"hint","knn" -"nbiot",180,22901.9086687423,7,"extended","knn" -"nbiot",180,27745.8252319951,8,"hint","knn" -"nbiot",180,26146.6891755201,10,"hintandextended","knn" -"nbiot",180,23831.1879103171,8,"extended","knn" -"nbiot",180,27030.7096498314,10,"hint","knn" -"nbiot",180,23875.3493146438,7,"hintandextended","tree" -"nbiot",180,26217.4338214212,12,"hint","tree" -"nbiot",180,23369.7590769622,9,"baseline","tree" -"nbiot",180,28487.7830938115,8,"hint","tree" -"nbiot",180,27871.6472379192,11,"hint","tree" -"nbiot",180,22901.9086687423,7,"extended","tree" -"nbiot",180,27745.8252319951,8,"hint","tree" -"nbiot",180,26146.6891755201,10,"hint","tree" -"nbiot",180,23831.1879103171,8,"baseline","tree" -"nbiot",180,27030.7096498314,10,"hint","tree" diff --git a/analysis/knn.R b/analysis/knn.R deleted file mode 100644 index d8a6ce1..0000000 --- a/analysis/knn.R +++ /dev/null @@ -1,107 +0,0 @@ -library("tidyverse") -library("class") -library("rpart") -library("rpart.plot") -library("viridis") - -## Simulation Parameters: -## simkey {baseline,extended,hint,hintandextended} -## wireless {lora,nbiot} -## wakeupfor {60s,180s} -## seed [1,200] -## node on[0,12] -## isSender {0,1} -## dataSize {1MB} - -## Metrics: -## energy [0,+inf) -## nDataRcv [0,+inf) - -nseed=200 -nwakeupfor=2 -nwireless=2 -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_seed=data%>%group_by(simkey,wireless,wakeupfor,seed)%>%summarize(energy=sum(energy),coverage=sum(nDataRcv))%>% - mutate(efficiency=energy/coverage)%>% - ungroup() - -## 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$wireless])#%>%filter(simkey!="hint") -train_set=data_ml%>%sample_frac(0.8) # 80% of the data -test_set=data_ml%>%anti_join(train_set) # 20% of the data - -## KNN training -knn_predictions=knn(train=train_set%>%select(-simkey),test=test_set%>%select(-simkey),cl=train_set$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))) -silent_call=dev.off() -## Notes: KNN accuracy jump to 76% and CART to 80% accuracy without the hint policy - -## Generate simulation inputs -inputs=tibble( -wakeupfor = c(60,180,60,180), -wireless = c("lora", "lora", "nbiot", "nbiot")) -constraints=apply(inputs,1,function(row){ - wi=row["wireless"] - wa=as.numeric(row["wakeupfor"]) - ## First extract energy/coverage boundaries - min_energy=min((data_seed%>%filter(wireless==wi,wakeupfor==wa))$energy) - max_energy=max((data_seed%>%filter(wireless==wi,wakeupfor==wa))$energy) - min_coverage=min((data_seed%>%filter(wireless==wi,wakeupfor==wa))$coverage) - max_coverage=max((data_seed%>%filter(wireless==wi,wakeupfor==wa))$coverage) - ## Generate random points (10 per scenarios) - n=10 - current_inputs=tibble( - wireless=rep(wi,n), - wakeupfor=rep(wa,n), - energy_constraint=runif(n,min_energy,max_energy), - coverage_constraint=round(runif(n,min_coverage,max_coverage))) - predictions_knn=knn(train=train_set%>%select(-simkey),test=current_inputs%>% - rename(energy=energy_constraint,coverage=coverage_constraint)%>% - mutate(wireless=wireless_map[wireless]),cl=train_set$simkey,k=10) - predictions_tree=predict(tree,newdata=current_inputs%>% - rename(energy=energy_constraint,coverage=coverage_constraint)%>% - mutate(wireless=wireless_map[wireless]),type="class") - knn_final=tibble(cbind(current_inputs,tibble(simkey=predictions_knn,model="knn"))) - tree_final=tibble(cbind(current_inputs,tibble(simkey=predictions_tree,model="tree"))) - rbind(knn_final,tree_final) -}) -inputs=do.call("rbind",constraints) -## Dimension Energy/Coverage -ggplot(data_seed%>%mutate(wakeupfor=as.character(wakeupfor)), - aes(coverage,energy,color=simkey))+geom_point()+ - geom_point(data=inputs%>%mutate(wakeupfor=as.character(wakeupfor)),aes(coverage_constraint,energy_constraint),size=3,pch=5)+ - ggtitle("Dimension Energy/Coverage")+xlab("Coverage")+ylab("Sum of nodes energy consumption (J)")+ - facet_wrap(~wakeupfor+wireless,scale="free") -ggsave("figures/random_inputs.pdf") -write.csv(inputs,"inputs.csv",row.names=FALSE) diff --git a/analysis/learning.R b/analysis/learning.R new file mode 100644 index 0000000..c393d9b --- /dev/null +++ b/analysis/learning.R @@ -0,0 +1,107 @@ +library("tidyverse") +library("class") +library("rpart") +library("rpart.plot") +library("viridis") + +## Simulation Parameters: +## simkey {baseline,extended,hint,hintandextended} +## wireless {lora,nbiot} +## wakeupfor {60s,180s} +## seed [1,200] +## node on[0,12] +## isSender {0,1} +## dataSize {1MB} + +## Metrics: +## energy [0,+inf) +## nDataRcv [0,+inf) + +nseed=200 +nwakeupfor=2 +nwireless=2 +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_seed=data%>%group_by(simkey,wireless,wakeupfor,seed)%>%summarize(energy=sum(energy),coverage=sum(nDataRcv))%>% + mutate(efficiency=energy/coverage)%>% + ungroup() + +## 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])#%>%filter(simkey!="hint") +train_set=data_ml%>%sample_frac(0.8) # 80% of the data +test_set=data_ml%>%anti_join(train_set) # 20% of the data + +## 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$simkey) +tree_accuracy=round((sum(diag(tree_cont_table)/sum(rowSums(tree_cont_table))))*100) +tree_prop_table=round(prop.table(tree_cont_table),digits=2) + +## Prints +print(paste0("Accuracy: KNN=",knn_accuracy,"% CART=",tree_accuracy,"%")) +pdf("figures/tree.pdf") +tree_plot=rpart.plot(tree,box.palette=as.list(viridis::viridis(4,begin=0.48))) +silent_call=dev.off() +## Notes: KNN accuracy jump to 76% and CART to 80% accuracy without the hint policy + +## Generate simulation inputs +inputs=tibble( +wakeupfor = c(60,180,60,180), +wireless = c("lora", "lora", "nbiot", "nbiot")) +constraints=apply(inputs,1,function(row){ + wi=row["wireless"] + wa=as.numeric(row["wakeupfor"]) + ## First extract energy/coverage boundaries + min_energy=min((data_seed%>%filter(wireless==wi,wakeupfor==wa))$energy) + max_energy=max((data_seed%>%filter(wireless==wi,wakeupfor==wa))$energy) + min_coverage=min((data_seed%>%filter(wireless==wi,wakeupfor==wa))$coverage) + max_coverage=max((data_seed%>%filter(wireless==wi,wakeupfor==wa))$coverage) + ## Generate random points (10 per scenarios) + n=10 + current_inputs=tibble( + wireless=rep(wi,n), + wakeupfor=rep(wa,n), + energy_constraint=runif(n,min_energy,max_energy), + coverage_constraint=round(runif(n,min_coverage,max_coverage))) + predictions_knn=knn(train=train_set%>%select(-simkey),test=current_inputs%>% + rename(energy=energy_constraint,coverage=coverage_constraint)%>% + mutate(wireless=wireless_map[wireless]),cl=train_set$simkey,k=10) + predictions_tree=predict(tree,newdata=current_inputs%>% + rename(energy=energy_constraint,coverage=coverage_constraint)%>% + mutate(wireless=wireless_map[wireless]),type="class") + knn_final=tibble(cbind(current_inputs,tibble(simkey=predictions_knn,model="knn"))) + tree_final=tibble(cbind(current_inputs,tibble(simkey=predictions_tree,model="tree"))) + rbind(knn_final,tree_final) +}) +inputs=do.call("rbind",constraints) +## Dimension Energy/Coverage +ggplot(data_seed%>%mutate(wakeupfor=as.character(wakeupfor)), + aes(coverage,energy,color=simkey))+geom_point()+ + geom_point(data=inputs%>%mutate(wakeupfor=as.character(wakeupfor)),aes(coverage_constraint,energy_constraint),size=3,pch=5)+ + ggtitle("Dimension Energy/Coverage")+xlab("Coverage")+ylab("Sum of nodes energy consumption (J)")+ + facet_wrap(~wakeupfor+wireless,scale="free") +ggsave("figures/random_inputs.pdf") +write.csv(inputs,"../inputs.csv",row.names=FALSE) -- cgit v1.2.3