Fixed function for subtyping and staging for new data
Added functionality to allow different numbers of z-scores per biomarker using Z_vals
Altered the SuStaIn algorithm and cross-validation so they don't output the results from the final subtypes model only to avoid confusion
Outputted additional probabilities for staging to enable advanced staging methods in future
Corrected CVIC calculation and relabeled previous CVIC_matrix to loglikelihood_matrix
Saved mean_likelihood_subj_test in cross-validation rather than all samples of subject likelihoods to avoid generating unnecesarily large output files with redundant information
Updated simfuncs to take as input Z_vals and Z_max
Made several changes