Code Sample, a copy-pastable example if possible
# Your code here
In [61]: df_null = pd.DataFrame(columns=['a','b','c'])
In [62]: df_null
Out[62]:
Empty DataFrame
Columns: [a, b, c]
Index: []
In [63]: ind1 = df_null.a == '0'
In [65]: ind2 = df_null.a.apply(lambda x: x == '0')
In [66]: df_null[ind1]
Out[66]:
Empty DataFrame
Columns: [a, b, c]
Index: []
In [67]: df_null[ind2]
Out[67]:
Empty DataFrame
Columns: []
Index: []
Problem description
Would expect both indices to result in the same empty dataframe with the structure of the columns preserved.
Expected Output
Empty DataFrame
Columns: [a, b, c]
Index: []
for both indices
Output of pd.show_versions()
[paste the output of ``pd.show_versions()`` here below this line]
INSTALLED VERSIONS
------------------
commit: None
python: 3.6.3.final.0
python-bits: 64
OS: Windows
OS-release: 10
machine: AMD64
processor: Intel64 Family 6 Model 79 Stepping 1, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None
LOCALE: None.None
pandas: 0.20.3
pytest: 3.2.1
pip: 9.0.1
setuptools: 36.5.0.post20170921
Cython: 0.26.1
numpy: 1.13.3
scipy: 0.19.1
xarray: None
IPython: 6.1.0
sphinx: 1.6.3
patsy: 0.4.1
dateutil: 2.6.1
pytz: 2017.2
blosc: None
bottleneck: 1.2.1
tables: 3.4.2
numexpr: 2.6.2
feather: None
matplotlib: 2.1.0
openpyxl: 2.4.8
xlrd: 1.1.0
xlwt: 1.3.0
xlsxwriter: 1.0.2
lxml: 4.1.0
bs4: 4.6.0
html5lib: 0.999999999
sqlalchemy: 1.1.13
pymysql: None
psycopg2: None
jinja2: 2.9.6
s3fs: None
pandas_gbq: None
pandas_datareader: None
Comment From: jreback
These are 2 different results. Using .apply for this purpose is quite non-idiomatic. [4] is a boolean indexer, while [3] is just an empty selection.
In [3]: df_null.a.apply(lambda x: x == '0')
Out[3]: Series([], Name: a, dtype: object)
In [4]: df_null.a == '0'
Out[4]: Series([], Name: a, dtype: bool)
Since .loc works correctly I guess fixing [] would be ok.
In [8]: df_null.loc[indexer]
Out[8]:
Empty DataFrame
Columns: [a, b, c]
Index: []
Comment From: phofl
As discussed in #37772 this is expected