Skip to content

min_samples_split is actually min_samples_leaf #8

@kjgm

Description

@kjgm

Both the OCT and BinOCT model have the parameter min_samples_split with default value 2.
However, they are used to constrain the minimum samples in the leaf, not in a split.
See oct.py line 166: m.addConstrs(z.sum('*', t) >= self.min_samples_split * l[t] for t in self.l_index)

Following the naming convention of sklearn's DecisionTreeClassifer, I suggest changing the parameter name to min_samples_leaf and changing its default value to 1

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions