While working through the 2nd book edition, I went to run R code 7.17 in the Overthinking box n p 210 (Chapter 7). The book code runs as is, but I was interested in computing the LOOCV value to replicate Figure 7.10, so I ran r <- sim_train_test( N=N, k=k, LOOCV=TRUE). I got an error which I traced back to the following line of cv_data_slice:
The call class(dlo[["mm"]]) returns "matrix" "array", so the comparison to "numeric" returns FALSE FALSE, which has length > 1, causing the error. I'm not sure the best way to correct this multiple class type issue, but will work on a PR unless someone beats me to it.
While working through the 2nd book edition, I went to run R code 7.17 in the Overthinking box n p 210 (Chapter 7). The book code runs as is, but I was interested in computing the LOOCV value to replicate Figure 7.10, so I ran
r <- sim_train_test( N=N, k=k, LOOCV=TRUE). I got an error which I traced back to the following line ofcv_data_slice:if (class(dlo[[vname]]) == "numeric")link to sourceThe call
class(dlo[["mm"]])returns"matrix" "array", so the comparison to "numeric" returnsFALSE FALSE, which has length > 1, causing the error. I'm not sure the best way to correct this multiple class type issue, but will work on a PR unless someone beats me to it.