modin.pandas.Series.to_dict¶

Series.to_dict(into: type[dict] = <class 'dict'>) → dict[source]¶

Convert Series to {label -> value} dict or dict-like object. Note that this method will pull the data to the client side.

Parameters:

into (class, default dict) – The collections.abc.Mapping subclass to use as the return object. Can be the actual class or an empty instance of the mapping type you want. If you want a collections.defaultdict, you must pass it initialized.

Returns:

Key-value representation of Series.

Return type:

collections.abc.Mapping

Examples

>>> s = pd.Series([1, 2, 3, 4])
>>> s.to_dict()
{0: 1, 1: 2, 2: 3, 3: 4}
>>> from collections import OrderedDict, defaultdict
>>> s.to_dict(OrderedDict)
OrderedDict([(0, 1), (1, 2), (2, 3), (3, 4)])
>>> dd = defaultdict(list)
>>> s.to_dict(dd)
defaultdict(<class 'list'>, {0: 1, 1: 2, 2: 3, 3: 4})
Copy