testRF: Test a random forest classifier
In kkatellis/TLBC: Two-level Behavior Classification

Function to apply a random forest classifier to some data.

1	testRF(featDirs, modelName, saveDir, testNames)

`featDirs`	Path to a directory (or list of directories) containing features, i.e., computed by the function sensorsToFeatures.
`modelName`	Path to pre-trained model.
`saveDir`	Path to a directory where predictions will be saved. Predictions will be saved in files named <identifier>.csv with two fields: timestamp, prediction.
`testNames`	List of participant identifiers to use.