ENH: change get_dummies default dtype to bool

doc/source/user_guide/reshaping.rst

@@ -608,6 +608,7 @@ values, can derive a :class:`DataFrame` containing ``k`` columns of 1s and 0s us
 :func:`~pandas.get_dummies`:
 
 .. ipython:: python
+   :okwarning:
 
    df = pd.DataFrame({"key": list("bbacab"), "data1": range(6)})
 
@@ -617,6 +618,7 @@ Sometimes it's useful to prefix the column names, for example when merging the r
 with the original :class:`DataFrame`:
 
 .. ipython:: python
+   :okwarning:
 
    dummies = pd.get_dummies(df["key"], prefix="key")
    dummies
@@ -626,6 +628,7 @@ with the original :class:`DataFrame`:
 This function is often used along with discretization functions like :func:`~pandas.cut`:
 
 .. ipython:: python
+   :okwarning:
 
    values = np.random.randn(10)
    values
@@ -642,6 +645,7 @@ variables (categorical in the statistical sense, those with ``object`` or
 
 
 .. ipython:: python
+    :okwarning:
 
     df = pd.DataFrame({"A": ["a", "b", "a"], "B": ["c", "c", "b"], "C": [1, 2, 3]})
     pd.get_dummies(df)
@@ -650,6 +654,7 @@ All non-object columns are included untouched in the output. You can control
 the columns that are encoded with the ``columns`` keyword.
 
 .. ipython:: python
+    :okwarning:
 
     pd.get_dummies(df, columns=["A"])
 
@@ -667,6 +672,7 @@ the prefix separator. You can specify ``prefix`` and ``prefix_sep`` in 3 ways:
 * dict: Mapping column name to prefix.
 
 .. ipython:: python
+    :okwarning:
 
     simple = pd.get_dummies(df, prefix="new_prefix")
     simple
@@ -680,6 +686,7 @@ variable to avoid collinearity when feeding the result to statistical models.
 You can switch to this mode by turn on ``drop_first``.
 
 .. ipython:: python
+    :okwarning:
 
     s = pd.Series(list("abcaa"))
 
@@ -690,6 +697,7 @@ You can switch to this mode by turn on ``drop_first``.
 When a column contains only one level, it will be omitted in the result.
 
 .. ipython:: python
+    :okwarning:
 
     df = pd.DataFrame({"A": list("aaaaa"), "B": list("ababc")})
 

doc/source/whatsnew/v0.13.0.rst

@@ -501,6 +501,7 @@ Enhancements
 - ``NaN`` handing in get_dummies (:issue:`4446`) with ``dummy_na``
 
   .. ipython:: python
+     :okwarning:
 
      # previously, nan was erroneously counted as 2 here
      # now it is not counted at all

doc/source/whatsnew/v0.15.0.rst

@@ -1007,6 +1007,7 @@ Other:
   left untouched.
 
   .. ipython:: python
+    :okwarning:
 
     df = pd.DataFrame({'A': ['a', 'b', 'a'], 'B': ['c', 'c', 'b'],
                     'C': [1, 2, 3]})

doc/source/whatsnew/v0.19.0.rst

@@ -431,6 +431,7 @@ The ``pd.get_dummies`` function now returns dummy-encoded columns as small integ
 **New behavior**:
 
 .. ipython:: python
+   :okwarning:
 
    pd.get_dummies(["a", "b", "a", "c"]).dtypes
 

doc/source/whatsnew/v0.23.0.rst

@@ -366,6 +366,7 @@ Function ``get_dummies`` now supports ``dtype`` argument
 The :func:`get_dummies` now accepts a ``dtype`` argument, which specifies a dtype for the new columns. The default remains uint8. (:issue:`18330`)
 
 .. ipython:: python
+   :okwarning:
 
    df = pd.DataFrame({'a': [1, 2], 'b': [3, 4], 'c': [5, 6]})
    pd.get_dummies(df, columns=['c']).dtypes

doc/source/whatsnew/v0.24.0.rst

@@ -833,6 +833,7 @@ then all the columns are dummy-encoded, and a :class:`SparseDataFrame` was retur
 Now, the return type is consistently a :class:`DataFrame`.
 
 .. ipython:: python
+   :okwarning:
 
    type(pd.get_dummies(df, sparse=True))
    type(pd.get_dummies(df[['B', 'C']], sparse=True))

doc/source/whatsnew/v1.5.0.rst

@@ -876,6 +876,7 @@ Other Deprecations
 - Deprecated unused arguments ``encoding`` and ``verbose`` in :meth:`Series.to_excel` and :meth:`DataFrame.to_excel` (:issue:`47912`)
 - Deprecated the ``inplace`` keyword in :meth:`DataFrame.set_axis` and :meth:`Series.set_axis`, use ``obj = obj.set_axis(..., copy=False)`` instead (:issue:`48130`)
 - Deprecated producing a single element when iterating over a :class:`DataFrameGroupBy` or a :class:`SeriesGroupBy` that has been grouped by a list of length 1; A tuple of length one will be returned instead (:issue:`42795`)
+- Deprecated ``np.uint8`` as the default ``dtype`` for :func:`get_dummies` - in a future version, it will be changed to ``bool`` (:issue:`45848`)
 - Fixed up warning message of deprecation of :meth:`MultiIndex.lesort_depth` as public method, as the message previously referred to :meth:`MultiIndex.is_lexsorted` instead (:issue:`38701`)
 - Deprecated the ``inplace`` keyword in :meth:`DataFrame.set_index`, use ``df = df.set_index(..., copy=False)`` instead (:issue:`48115`)
 - Deprecated the ``sort_columns`` argument in :meth:`DataFrame.plot` and :meth:`Series.plot` (:issue:`47563`).

pandas/core/reshape/encoding.py

@@ -1,13 +1,16 @@
 from __future__ import annotations
 
 from collections import defaultdict
+import inspect
 import itertools
 from typing import Hashable
+import warnings
 
 import numpy as np
 
 from pandas._libs.sparse import IntIndex
 from pandas._typing import Dtype
+from pandas.util._exceptions import find_stack_level
 
 from pandas.core.dtypes.common import (
     is_integer_dtype,
@@ -228,6 +231,12 @@ def _get_dummies_1d(
     codes, levels = factorize_from_iterable(Series(data))
 
     if dtype is None:
+        warnings.warn(
+            "In a future version of pandas the default dtype will change from "
+            "'uint8' to 'bool', please specify a dtype to silence this warning",
+            FutureWarning,
+            stacklevel=find_stack_level(inspect.currentframe()),
+        )
         dtype = np.dtype(np.uint8)
     # error: Argument 1 to "dtype" has incompatible type "Union[ExtensionDtype, str,
     # dtype[Any], Type[object]]"; expected "Type[Any]"