Skip to content

DEPR: read_csv header with non-exisiting rows behaving differently #47400

Closed
@phofl

Description

@phofl

Pandas version checks

  • I have checked that this issue has not already been reported.

  • I have confirmed this bug exists on the latest version of pandas.

  • I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

data = """a,b
1,2"""

pd.read_csv(StringIO(data), header=[0, 1, 2], engine="python")

This creates a MultiIndex with 2 levels

While the c engine raises


pandas.errors.ParserError: Passed header=[0,1,2], len of 3, but only 2 lines in file

Issue Description

They should be consistent, since the c engine is more widely used, I would propose adapting the behavior to raise for the python engine. We've deprecated this behavior for usecols and will raise starting with 2.0. We should do something similar here.

Expected Behavior

Both engines should be consistent

Installed Versions

Replace this line with the output of pd.show_versions()

Metadata

Metadata

Assignees

No one assigned

    Labels

    DeprecateFunctionality to remove in pandasIO CSVread_csv, to_csvNeeds TriageIssue that has not been reviewed by a pandas team member

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions