Re: [R] predict () for LDA and GLM

2012-03-22 Thread palanski
.testSetDF[,1]) ass4q1.dNBtable.testData[1,2] = (ass4q1.dNB.cTable[2,1] + ass4q1.dNB.cTable [1,2])/(sum(ass4q1.dNB.cTable)) #WORKS! #2 features for LDA ass4q1.dLDA.cTable = table(predict(ass4q1.dLDA, ass4q1.testSetDF[,2:3])$class, ass4q1.testSetDF[,1]) #DOESN'T WORK! ass4q1.dLDAtabl

[R] predict () for LDA and GLM

2012-03-21 Thread palanski
Hi! I'm using GLM, LDA and NaiveBayes for binomial classification. My training set is 70 rows long with 32 features, and my test set is 30 rows long with 32 features. Using Naive Bayes, I can train a model, and then predict the test set with it like so: ass4q1.dLDA = lda(ass4q1.trainSet[,1]~ass4

[R] ANOVA for glmnet

2012-02-16 Thread palanski
Simply: The R2 value we obtain at an optimized lambda using glmnet: how do we state whether that's significant or not? Using the standard lm() function, we are able to run an ANOVA and test for significance. We have no such output with glmnet. Thanks! -- View this message in context: http://r.