CODEWITHSUNDEEP
r
screechOwl
screechOwl

Reputation: 28169

Save a random Forest object

I have a randomForest object that I want to save for later use. I've tried some of the following but with no luck.

save(topDawg , file="myRFobject.RData")

This just saves a string "topDawg"

> formula(topDawg)
Error in formula.default(topDawg) : invalid formula


> save(getTree(topDawg))
Error in save(getTree(topDawg)) : object ‘getTree(topDawg)’ not found

Any suggestions?

Upvotes: 15

Views: 19717

Answers (4)

Rijin
Rijin

Reputation: 91

Once a random forest model is created save the model as a file so that you can reuse it later.

#model
  model_rf <- randomForest(y~ ., data = dataset, mtry=7,ntree=500)

#write model
  saveRDS(model_rf , "model_rf.RDS")

#load model
 readRDS("model_rf.RDS")

Upvotes: 0

Andrii
Andrii

Reputation: 3043

Here is a solution if you would like to load the model under another name

  library(randomForest)

  # 1. Create data set
  set.seed(100)
  df_iris <- randomForest(Species ~ ., data = iris, importance = TRUE,  proximity = TRUE)

  # 2. Save model
  file_name <- "model_iris.rds"
  saveRDS(df_iris, file_name)

  # 2.3. Load model under another name
  df_iris_loaded <- readRDS(file_name)
  df_iris_loaded

  # 2.4. Test two models
  identical(df_iris, df_iris_loaded, ignore.environment = TRUE)

Upvotes: 3

Matija Černe
Matija Černe

Reputation: 51

I had the same problem (loading RandomForest object resulted in character string) and something like this seemed to have worked for me:

forest = get(load("forestGOOG.RData"))

(I have a random forest object 'forestGOOG' saved in the working directory)

Upvotes: 5

joran
joran

Reputation: 173677

I'm not sure exactly what you're trying to do here, since normally you save an object and then load it later, like this:

set.seed(71)
> irisrf <- randomForest(Species ~ ., data=iris, importance=TRUE,
+                         proximity=TRUE)
> save(irisrf,file = "irisrf.RData")
> 
> rm(irisrf)
> print(irisrf)
Error in print(irisrf) : object 'irisrf' not found
> 
> load("irisrf.RData")
> print(irisrf)

Call:
 randomForest(formula = Species ~ ., data = iris, importance = TRUE,      proximity = TRUE) 
               Type of random forest: classification
                     Number of trees: 500
No. of variables tried at each split: 2

        OOB estimate of  error rate: 4.67%
Confusion matrix:
           setosa versicolor virginica class.error
setosa         50          0         0        0.00
versicolor      0         47         3        0.06
virginica       0          4        46        0.08

Upvotes: 27

Related Questions