DTree

DTree(root=None, name=None, **data)

A sampleable, serializable decision tree.

Attributes

Name	Description
model_config	dict() -> new empty dictionary

Methods

Name	Description
expected_value	Return the rolled-back expected value of the tree.
plot	Plot sampled tree outcomes as an empirical CDF with a histogram.
plot_tree	Plot the decision tree structure from top to bottom.
rollback	Return a rollback table indexed by node.
rvs	Alias for :meth:sample.
sample	Sample outcomes under the rollback-selected policy, without perfect information.
summary	Summarize simulated outcomes under the rollback-selected policy.

expected_value

DTree.expected_value(size=None, seed=None)

Return the rolled-back expected value of the tree.

plot

DTree.plot(ax=None, *, size=10000, seed=None, bins=80, show=False, **kwargs)

Plot sampled tree outcomes as an empirical CDF with a histogram.

plot_tree

DTree.plot_tree(
    ax=None,
    *,
    size=None,
    seed=None,
    show_expected_values=True,
    show_probabilities=True,
    show_selected=True,
    precision=2,
    show=False,
)

Plot the decision tree structure from top to bottom.

rollback

DTree.rollback(size=None, seed=None, precision=2, node_types='decision')

Return a rollback table indexed by node.

By default, only decision-node branches are shown. Pass node_types=None to include decision, chance, and outcome nodes, or pass a node type / tuple of node types to filter the table.

rvs

DTree.rvs(size=1, *, seed=None)

Alias for :meth:sample.

sample

DTree.sample(size=1, *, seed=None)

Sample outcomes under the rollback-selected policy, without perfect information.

summary

DTree.summary(
    size=10000,
    seed=None,
    percentiles=DEFAULT_PERCENTILES,
    precision=2,
)

Summarize simulated outcomes under the rollback-selected policy.