Hello, I'm fitting a logistic regression as follows and I have this error "In predict.lm(object, newdata, se.fit, scale = 1, type = ifelse(type == : prediction from a rank-deficient fit may be misleading" Codes: ##Training fit<- glm(Y.Binary.training~., family=binomial(link="logit"), data=training) pred<-predict(fit, training, type="response") pred pred.train<-cbind(pred, Y.Binary.training) preds<-prediction(as.numeric(pred),as.numeric(Y.Binary.training)) perf<- performance(preds, "auc") perf #### I get a Roc Curve of 1 which is surprising! ##Testing pred<-predict(fit, testing, type="response") pred pred.test<-cbind(pred, Y.Binary.testing) preds<-prediction(as.numeric(pred),as.numeric(Y.Binary.testing)) perf<- performance(preds, "auc") perf #### I also get a ROC curve of 1...!!! I'm wondering if there is something wrong in my codes . Thanks a lot and happy holidays! Best,Farnoosh Sheikhi [[alternative HTML version deleted]]