Skip to content

Default dimension names of LearnerND.to_dataframe assume 3D space. #468

@serhiy-yevtushenko

Description

@serhiy-yevtushenko

When trying to use LearnerND on 2D data (see issue #466), when calling learner.to_dataframe() function, one gets the exception:
ValueError: point_names (('x', 'y', 'z')) should have the same length as learner.ndims (2)

The issue is caused by the fact, that default values of points_name for LearnerND are point_names: tuple[str, ...] = ("x", "y", "z"),

I believe the similar issue will be observed, if one tries to call to_dataframe with 4D or higher dimensional data, as point names will not be matching the dimension.

Fazit - the default for point_names in to_dataframe make sense only if LearnerND is used with 3D data

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions