Error in cbind2(1, newx) %*% nbeta : invalid class 'na' to dup_mmatrix_as_dgematrix

Question

Problem :

I want to use the glmnet package on a dataset. I am trying to use cv.glmnet() to get the lambda value for the glmnet().

Please find below the dataset and error message:

> head(t2)

  X1 X2        X3 X4 X5         X6    X7 X8 X9 X10 X11 X12

1  1  1 1.7661266 45  2 0.80298213  9120 13  0   6   1   2

2  2  0 1.9571510 40  0 0.12187620  2600  4  0   0   1   1

3  3  0 1.6581801 38  1 0.08511338  3042  2  1   0   1   0

4  4  0 1.2338098 30  0 0.03604968  3300  5  0   0   1   0

5  5  0 1.9072394 49  1 0.02492570 63588  7  0   1   1   0

6  6  0 1.2131787 74  0 0.37560697  3500  3  0   1   1   1

> cv1 <- cv.glmnet(as.matrix(t2[,-c(1,2,7,12)]), t2[,2], family="binomial")

Error in as.matrix(cbind2(1, newx) %*% nbeta) : 

error in evaluating the argument 'x' in selecting a method for function 'as.matrix': Error in t(.Call(Csparse_dense_crossprod, y, t(x))) : 

error in evaluating the argument 'x' in selecting a method for function 't': Error: invalid class 'NA' to dup_mMatrix_as_dgeMatrix

> cv1 <- cv.glmnet(as.matrix(t2[,-c(1,2,7,12)]), t2[,2], family="multinomial")

Error in t(.Call(Csparse_dense_crossprod, y, t(x))) : 

error in evaluating the argument 'x' in selecting a method for function 't': Error: invalid class 'NA' to dup_mMatrix_as_dgeMatrix

Any suggestions regarding the solution?

Answer 1

For some reason the glmnet prefers data.matrix() over as.matrix() so

cv1 <- cv.glmnet(data.matrix(t2[,-c(1,2,7,12)]), t2[,2], family="multinomial")

will solve your problem.