Hi
I followed the procedure of parallel processing of different models in the tidymodels package (Chapter 16. Dimensionality Reduction), but got stuck at this step:
pls_rec <-
data_rec %>%
step_pls(all_numeric_predictors(), outcome = "Group", num_comp = tune())
I receive the error :
Calling 'tune.spls'
Error: 'X' must be a numeric matrix (possibly including NA's) with finite values
My data looks like this:
> str (data)
'data.frame': 123 obs. of 30 variables:
$ Group: chr "Islands" "Islands" "Islands" "Islands" ...
$ Al : num 0.517 2.974 0.13 0.13 0.13 ...
$ Ca : num 13.1 96.1 120.1 17.6 11.6 ...
$ Fe : num 0.05 1.743 0.304 0.141 0.254 ...
$ K : num 4.27 7.28 68.42 243.33 233.05 ...
$ Mg : num 7.97 28.12 57.37 26.65 23.62 ...
$ Na : num 3.02 1.89 2.95 7.76 5.64 ...
$ P : num 1.19 4.04 6.6 23.53 20.56 ...
$ S : num 32.3 32.2 38.5 18 26.4 ...
$ Ba : num 0.0106 0.104 0.01 0.01 0.01 ...
$ Cd : num 0.0009 0.00195 0.000936 0.000944 0.0005 ...
$ Co : num 0.00101 0.00202 0.00202 0.00203 0.00101 ...
$ Cr : num 0.0215 0.0112 0.0025 0.0025 0.0025 ...
$ Cs : num 0.000581 0.000892 0.0005 0.0005 0.0005 ...
$ Cu : num 0.463 0.192 0.472 0.138 0.142 ...
$ Mn : num 0.0127 0.2067 0.061 0.0839 0.1031 ...
$ Mo : num 0.00811 0.00715 0.00415 0.00348 0.0033 ...
$ Ni : num 0.025 0.025 0.135 0.0702 0.025 ...
$ Pb : num 0.028 0.0533 0.0271 0.01 0.01 ...
$ Rb : num 0.00198 0.00602 0.08361 0.06402 0.18307 ...
$ Sb : num 0.00401 0.00398 0.00302 0.00203 0.00302 ...
$ Sn : num 0.0352 0.0104 0.01 0.0231 0.01 ...
$ Sr : num 0.0159 0.0501 0.039 0.005 0.005 ...
$ Ti : num 0.005 0.05247 0.005 0.00615 0.00689 ...
$ V : num 0.00203 0.00505 0.001 0.001 0.001 ...
$ Zn : num 0.399 0.342 0.456 0.825 0.279 ...
$ Ce : num 0.00406 0.0091 0.00202 0.00102 0.00101 ...
$ La : num 0.00101 0.00303 0.00101 0.0005 0.0005 ...
$ Nd : num 0.000602 0.006446 0.001281 0.0005 0.001056 ...
$ Y : num 0.0006 0.001734 0.001577 0.00035 0.000476 ...
Can anybody help me?
Thanks!
Mavro