I ran a random forest classification using the randomForest package. When it is finished, I typed summary() on my classifier and it appeared that the ntree parameter was left equal to 1, when I was told that the default value was 500, and it could be changed manually in the argument of randomForest, which I tried unsuccessfully.
I also tried it with another dataset and I had the same issue. Does anyone has any idea of what might be going on?
TL;DR: To get a summary of the model, just type the name of the model object. For example, if the model object is rf1 type rf1, not summary(rf1).
Most packages have a summary "method" that gets dispatched when you run summary on an object produced by the package. But in the case of randomForest there doesn't seem to be a summary method. The output of randomForest is a list containing a bunch of model output. When you run summary on it, it just runs the default summary function, which returns the length of each list element, which is not very useful here.
Thus, in this case, when you run summary on your randomForest model object, you're seeing a value of 1 for ntree, because ntree is an element of the list returned by randomForest and it is a vector of length 1. (Note that the column name of the summary output is Length.)
To see a summary of model results, just type the name of your model object and this will cause an actual summary to be printed to the console. For example, if your model object is called rf1, just type rf1, not summary(rf1). Typing the object name causes the print.randomForest method to be dispatched, and this does provide a summary of the randomForest results, including ntree.
If you want to extract the value of ntree or other results from your model, run str(rf1) to see the structure of the list returned by randomForest and also look at the help for randomForest for additional information on what's in this list. For example, rf1$ntree would return the number of trees in the model.
