a prediction function for ulam objects, as comes from rethinking::ulam

# S3 method for ulam
predict(ulamobj, new_data, n = 1000, reduce = T, conf = 0.95)

Arguments

ulamobj

the out of rethinking::ulam

new_data

the dataset on which to draw samples and make predictions

n

the number of samples to draw

reduce

defuaults to TRUE; a boolean that indicates if the samples should be returned or a mean and high credibility interval

conf

defaults to .95, indicates the desired converage of credibility interval when reduce=T

Value

if reduce=T a tibble with pred, pred_lower_conf, pred_upper_conf columns else if reduce =F the output of link (a matrix with sampled predictions)