Question: Plotting ROC curve using R

0

0

Irsan • **6.8k** wrote:

```
Below is the sample code i have found. I am not clear how to interpret with my dataset. kindly help
```

`what did ROCR.simple$predictions contains in my way of understanding it is the predictions done. Correct me if am wrong. Else help me wit some example.`

```
library(ROCR)
data(ROCR.simple)
pred <- prediction( ROCR.simple$predictions, ROCR.simple$labels )
pred2 <- prediction(abs(ROCR.simple$predictions +
rnorm(length(ROCR.simple$predictions), 0, 0.1)),
ROCR.simple$labels)
perf <- performance( pred, "tpr", "fpr" )
perf2 <- performance(pred2, "tpr", "fpr")
plot( perf, colorize = TRUE)
plot(perf2, add = TRUE, colorize = TRUE)
```

0

rk2153 • **0** wrote:

try this -- and adapt for your dataset

titanic<-read.csv("http://christianherta.de/lehre/dataScience/machineLearning/data/titanic-train.csv",header=T) head(titanic) dim(titanic) sm_titanic<-complete.cases(titanic[c(2,3,4,5,6,10),]) dim(sm_titanic) head(sm_titanic) sm_titantic_3<-titanic[,c(2,3,5,6,10)] sm_titanic_3<-sm_titantic_3[complete.cases(sm_titantic_3),] head(sm_titanic_3) dim(sm_titanic_3) tst_idx<-sample(714,200,replace=FALSE) length(tst_idx) tstdata<-sm_titanic_3[tst_idx,] trdata<-sm_titanic_3[-tst_idx,] length(trdata) dim(tstdata) dim(trdata) glm_sm_titanic_3<-glm(Survived~.,data=trdata,family=binomial()) predicted<-predict(glm_sm_titanic_3,tstdata[,-Survived],type="response"); predicted<-predict(glm_sm_titanic_3,tstdata[,2:5],type="response"); require(ROCR) auc_1<-prediction(predicted,tstdata$Survived) auc_1 prf<-performance(auc_1, measure="tpr", x.measure="fpr") slot_fp<-slot(auc_1,"fp") slot_tp<-slot(auc_1,"tp") table(tstdata$Survived) xtpcount<-table(tstdata$Survived) tpcount<-unlist(xtpcount) fpr<-unlist(slot_fp)/tpcount[[1]] tpr<-unlist(slot_tp)/tpcount[[2]] plot(fpr,tpr, main="ROC Curve from first principles -- raw counts")

Please log in to add an answer.

Use of this site constitutes acceptance of our User
Agreement
and Privacy
Policy.

Powered by Biostar
version 2.3.0

Traffic: 1814 users visited in the last hour

try this -- and adapt for your dataset I have posted a redable/formatted version below

0