I'm currently trying to use an unsupervised random forest to generate a proximity matrix for clustering. Unfortunately, my dataset is too large for the R randomforest package. The other methods of running unsupervised random forests (e.g., scikit in python), are tailored towards feature selection and don't seem able to output the distance matrix I am looking for. Any suggestions of different R packages, python tools, etc. would be much appreciated. Thanks!
I've tried the R packages Randomforest, rfunsupervised from rutils, and the clustering tools on scikit.