pandas.MultiIndex.drop

MultiIndex.drop(codes, level=None, errors='raise')

Make a new pandas.MultiIndex with the passed list of codes deleted.

Parameters:

codesarray-like

Must be a list of tuples when level is not specified.

levelint or level name, default None

errorsstr, default ‘raise’

Returns:

MultiIndex

Examples

>>> idx = pd.MultiIndex.from_product(
...     [(0, 1, 2), ("green", "purple")], names=["number", "color"]
... )
>>> idx
MultiIndex([(0,  'green'),
            (0, 'purple'),
            (1,  'green'),
            (1, 'purple'),
            (2,  'green'),
            (2, 'purple')],
           names=['number', 'color'])
>>> idx.drop([(1, "green"), (2, "purple")])
MultiIndex([(0,  'green'),
            (0, 'purple'),
            (1, 'purple'),
            (2,  'green')],
           names=['number', 'color'])

We can also drop from a specific level.

>>> idx.drop("green", level="color")
MultiIndex([(0, 'purple'),
            (1, 'purple'),
            (2, 'purple')],
           names=['number', 'color'])

>>> idx.drop([1, 2], level=0)
MultiIndex([(0,  'green'),
            (0, 'purple')],
           names=['number', 'color'])