Skip to content

DataFrameTypeDetector can't work with missing values #66

@olekscode

Description

@olekscode

If a single value of a column is missing, the all other values in that column will be interpreted as strings. This can be fixed, but it should be done together with other issues related to missing values

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