带R.的随机森林不能分配大小为7.5 Gb的向量

i have a dataset (df) with 10000 tweet with 2 observations (body and label); The first 4000 tweets are labelled(positive or negative) and i want to use it as training set to predict the label of the rest of the tweets ,testing, according to the text body. I want to use Random Forest algorithm in order to predict(caret package) and cross validation to determine the accuracy. i used this script:i have a dataset (df) with 10000 tweet with 2 o