Series.count level is only valid with a MultiIndex

Package:

pandas

30911

Exception Class:

ValueError

Raise code

                "Using the level keyword in DataFrame and Series aggregations is "
                "deprecated and will be removed in a future version. Use groupby "
                "instead. ser.count(level=1) should use ser.groupby(level=1).count().",
                FutureWarning,
                stacklevel=2,
            )
            if not isinstance(self.index, MultiIndex):
                raise ValueError("Series.count level is only valid with a MultiIndex")

        index = self.index
        assert isinstance(index, MultiIndex)  # for mypy

        if isinstance(level, str):
            level = index._get_level_number(level)

Links to the raise (1)

https://github.com/pandas-dev/pandas/blob/b3e335254f46a526ee3ce9bb757eac4011d9d1fe/pandas/core/series.py#L1907

Ways to fix

Error code:

import pandas as pd
import numpy as np

index = [1,2,3,4,6,7,8]
s = pd.Series(np.random.randn(8),index)
s.count(level=1)

Pandas.Series.count returns number of non-NA/null observations in the Series.

If the index of Series is not instance of MultiIndex , it will not possible to count the number of non-Na/null observations.

Fix code:

import pandas as pd
import numpy as np

iterables = [["bar", "baz", "foo", "qux"], ["one", "two"]]
index = pd.MultiIndex.from_product(iterables,names=["first", "second"])
# We created MultiIndex and initial it as a index of Series
s = pd.Series(np.random.randn(8),index)
s.count(level=1)

Sep 30, 2021

anonim 13.0k

Add a possible fix

Please authorize to post fix