"Error: Must subset rows with a valid subscript vector" in preProcess() when using knnImpute
I'm using kaggle's pokemon data to practice KNN imputation via preProcess(), but when I did I encountered this following message after the predict() step. I am wondering if I use the incorrect data format or if some columns have inappropriate "class." Below is my code.
and I received this error message
Error: Must subset rows with a valid subscript vector.
x Subscript `nn$nn.idx` must be a simple vector, not a matrix.
Run `rlang::last_error()` to see where the error occurred.
Answer
The input for preProcess needs to be a data.frame. This works:
pre_process_missing_data <- preProcess(as.data.frame(df), method="knnImpute")
classify_legendary <- predict(pre_process_missing_data, newdata = df)
classify_legendary
> classify_legendary
# A tibble: 801 x 6
hp weight_kg height_m sp_attack sp_defense capture_rate
<dbl> <dbl> <dbl> <dbl> <dbl> <chr>
1 -0.902 -0.498 -0.429 -0.195 -0.212 45
2 -0.337 -0.442 -0.152 0.269 0.325 45
3 0.415 0.353 0.774 1.57 1.76 45
4 -1.13 -0.484 -0.522 -0.349 -0.748 45
5 -0.412 -0.388 -0.0591 0.269 -0.212 45
6 0.340 0.266 0.496 2.71 1.58 45
7 -0.939 -0.479 -0.615 -0.659 -0.247 45
8 -0.375 -0.356 -0.152 -0.195 0.325 45
9 0.378 0.221 0.404 1.97 1.58 45
10 -0.902 -0.535 -0.800 -1.59 -1.82 255
# ... with 791 more rows