Search code examples
rrandom-forestpredict

Error in UseMethod "predict" when running Random Forest model

Looking for feedback on the error below. I built a Random Forest classification model a couple years ago, and now I'm simply trying to run it again in rStudio on a new set of data. Hoping someone can educate me on the error.

    #load libraries
    library(readr)
    library(randomForest)

    #load Random Forest model
    rf_model <- readRDS("my_model.rds")

    #import files
    set1 <- read_csv("set_1.csv")
    set2 <- read_csv("set_2.csv")
    set3 <- read_csv("set_3.csv")

    #combine sets into single df
    total <- rbind(set1, set2, set3)

    #remove non_variables
    total_to_score <- total[-c(1)]

    #run model
    scored_output <- predict(rf_model, total_to_score)

And here's the error:

    Error in UseMethod("predict") : 
      no applicable method for 'predict' applied to an object of class "c('train', 'train.formula')"

I tried loading 'caret' too, but that didn't help.

Solution

  • I was able to make that command work after reinstalling rtools and rlang. I also installed caret but I'm not sure if that necessary.

    At any rate, I think rtools was lacking or not found.