Classification and regression tree on the flattened representation

One tree per response, over every bin-by-channel column of the representation, grown under rpart's rules: the Gini index under a presence-absence head, the sum of squares under a head with a squared-error loss and the Poisson deviance under a count head, a split only between two distinct values of a column, and the cost-complexity bookkeeping that keeps a split only where it lowers the risk by at least cp of the root's. On the same columns, weights and folds the tree is the one rpart grows, split for split, and its complexity table the one rpart reports; the tree is grown by the core the Python package calls, so the two languages grow it identically.

Usage

tree(
  data = NULL,
  min_split = NULL,
  min_leaf = NULL,
  cp = NULL,
  max_depth = NULL,
  prune = c("se_sum", "one_se", "min", "none"),
  n_inner = NULL,
  preset = c("default", "bigboss"),
  shrink = 1,
  seed = 1L,
  threads = 1L
)

Arguments

data

A representation the learner is pinned to, or NULL to run across every representation of the run.

min_split

Observations a node needs before a split of it is tried.

min_leaf

Observations each child of a split keeps.

cp

The share of the root's risk a split has to remove to be kept.

max_depth

Depth of the deepest node, the root at depth 0; 30 at most.

prune

How the grown tree is pruned: "se_sum", "one_se", "min" or "none".

n_inner

Folds of the inner cross-validation the pruning reads.

preset

Whose defaults the settings left NULL take: "default" or "bigboss".

shrink

Under a count head, the coefficient of variation of the gamma prior a leaf's rate is shrunk by; 0 for no shrinkage. rpart's default is 1.

seed

Seed for the inner cross-validation's fold draw.

threads

Responses grown at once. What comes back does not depend on it.

Value

A learner().

Details

The grown tree is pruned back by an inner cross-validation. Its folds are dealt for each response and stratified on it, as the elastic net's are, and prune names the rule that reads the complexity table: "se_sum" takes the row of least cross-validated error plus its standard error among the rows that keep a split, the last of them where several tie, which is how biomod2 prunes its classification tree; "one_se" takes the smallest tree within one standard error of the least cross-validated error; "min" the first row reaching the least error; and "none" keeps the tree as grown.

preset says whose defaults the settings left NULL take. "default" is rpart's own, which is what biomod2's default option set fits: min_split = 20, min_leaf = round(min_split / 3) (or min_split = 3 * min_leaf where only min_leaf is given), cp = 0.01, max_depth = 30 and ten inner folds. "bigboss" is biomod2's tuned option set: min_split = 5, min_leaf = 5, cp = 0.001, max_depth = 10 and five inner folds. A setting given explicitly beats either.

The case weights are the response head's, positive_weights() under presence-absence. They weigh every class count, sum of squares and event count the tree is grown on; min_split and min_leaf count observations, as rpart's do. Under the shipped presence-absence head those weights are on, so a default tree() is rpart's specification fitted under them; a head registered without weights fits it unweighted.

Under a count head a leaf predicts a rate, and the rate is shrunk towards the rate of the units the tree is grown on, as rpart's method = "poisson" shrinks it: the posterior mean of a gamma prior whose coefficient of variation is shrink. A split is chosen on the deviance of the unshrunk rates, and a subtree's risk, its complexity and the pruning's cross-validated error are read on the shrunk ones.

Examples

tree()
tree(preset = "bigboss", prune = "one_se")