AttributeError: 'DataFrame' object has no attribute 'to_sparse'

AttributeError: 'DataFrame' object has no attribute 'to_sparse'

sdf = df.to_sparse() 已弃用。转换为稀疏 DataFrame 的更新方法是什么?

您可以使用scipy创建稀疏矩阵:

scipy.sparse.csr_matrix(df.values)

这些是 pandas 1.0.0+ 中的 updated sparse conversions


如何将密集转换为稀疏

使用DataFrame.astype() with the appropriate SparseDtype()(例如int):

>>> df = pd.DataFrame({'A': [1, 0, 0, 0, 1, 0]})
>>> df.dtypes
# A    int64
# dtype: object

>>> sdf = df.astype(pd.SparseDtype(int, fill_value=0))
>>> sdf.dtypes
# A    Sparse[int64, 0]
# dtype: object

或为简洁起见使用字符串别名:

>>> sdf = df.astype('Sparse[int64, 0]')

如何将稀疏转换为密集

使用DataFrame.sparse.to_dense():

>>> from scipy import sparse
>>> sdf = pd.DataFrame.sparse.from_spmatrix(sparse.eye(3), columns=list('ABC'))
>>> sdf.dtypes
# A    Sparse[float64, 0]
# B    Sparse[float64, 0]
# C    Sparse[float64, 0]
# dtype: object

>>> df = sdf.sparse.to_dense()
>>> df.dtypes
# A    float64
# B    float64
# C    float64
# dtype: object

如何将稀疏转换为 COO

使用DataFrame.sparse.to_coo():

>>> from scipy import sparse
>>> sdf = pd.DataFrame.sparse.from_spmatrix(sparse.eye(3), columns=list('ABC'))
>>> sdf.dtypes
# A    Sparse[float64, 0]
# B    Sparse[float64, 0]
# C    Sparse[float64, 0]
# dtype: object

>>> df = sdf.sparse.to_coo()
# <3x3 sparse matrix of type '<class 'numpy.float64'>'
#         with 3 stored elements in COOrdinate format>
# (0, 0)    1.0
# (1, 1)    1.0
# (2, 2)    1.0