Computes the out-of-bag (OOB) predictions for the validation subset of the training data. This allows the user to try many different models while still leaving test data for true out-of-sample performance estimation.
Computes the out-of-bag (OOB) predictions for the test subset of the training data. This function should only be run once. If it run again you risk adapting your model to the test data which will lead to overfitting and hence a biased out-of-sample performance estimate.
Computes information about the "proximity" of observations within the YARF model. Given two datasets, information is computed for all pairs of observations. Information returned is the prediction nodes (for all trees) for both objects and their common node in the tree structure plus much information about the nodes.
Computes information about the "proximity" of observations within the YARF model. Given two datasets, information is computed for all pairs of observations. Information is averaged over the trees in the model.
Computes information about the "proximity" of observations within the YARF model. Given two datasets, information is computed for all pairs of observation.
Function which queries the YARF model about the proportion of times the variables are used when splitting. This is a convenience method which merely normalizes the result of query_variable_counts.
Tests the effect of H0: one, or two, ... or all covariates are not predictive (out of sample) via a permutation test. We permute the covariates column(s) and build YARF models num_permutation_samples times. The p-val is determined by a permutation-like test.
Convenience method for testing a model's fit. The strategy is simple: one builds a YARF model from the residuals of the model you wish to fit and performs an omnibus test of all covariates against those residuals. This effecticely answers the question: "is there out-of-sample predictive information left over after this model was fit?" The p-val is determined by a permutation-like test.
