New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

wide_to_long does not convert integer suffixes to int #17627

Closed
tdpetrou opened this Issue Sep 22, 2017 · 3 comments

Comments

Projects
None yet
2 participants
@tdpetrou
Contributor

tdpetrou commented Sep 22, 2017

Code Sample, a copy-pastable example if possible

>>> df = pd.DataFrame({
                'A_1': [1.0, 2.0],
                'A_2': [3.0, 4.0],
                'X': ['X1', 'X2']})

   A_1  A_2   X
0  1.0  3.0  X1
1  2.0  4.0  X2

>>> df2 = pd.wide_to_long(df, stubnames='A', i='X', j='num', sep='_')
          A
X  num     
X1 1    1.0
X2 1    2.0
X1 2    3.0
X2 2    4.0

>>> df2.reset_index().dtypes
X       object
num     object
A      float64
dtype: object

Problem description

wide_to_long should attempt to coerce integer strings to ints. I have a pull request ready for this.

Additionally, I'd think it might be good to deprecate wide_to_long and lreshape in favor of expanded melt functionality.

Expected Output

>>> df2.reset_index().dtypes
X       object
num     int64
A      float64
dtype: object

Output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.6.1.final.0
python-bits: 64
OS: Darwin
OS-release: 15.6.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8
LOCALE: en_US.UTF-8

pandas: 0.20.3
pytest: 3.0.7
pip: 9.0.1
setuptools: 35.0.2
Cython: 0.25.2
numpy: 1.13.1
scipy: 0.19.0
xarray: None
IPython: 6.0.0
sphinx: 1.5.5
patsy: 0.4.1
dateutil: 2.6.1
pytz: 2017.2
blosc: None
bottleneck: 1.2.0
tables: 3.4.2
numexpr: 2.6.2
feather: None
matplotlib: 2.0.2
openpyxl: 2.4.7
xlrd: 1.0.0
xlwt: 1.2.0
xlsxwriter: 0.9.6
lxml: 3.7.3
bs4: 4.6.0
html5lib: 0.9999999
sqlalchemy: 1.1.9
pymysql: None
psycopg2: None
jinja2: 2.9.6
s3fs: None
pandas_gbq: None
pandas_datareader: 0.3.0.post

@tdpetrou tdpetrou referenced this issue Sep 22, 2017

Merged

BUG: coerce pd.wide_to_long suffixes to ints #17628

3 of 3 tasks complete
@jreback

This comment has been minimized.

Contributor

jreback commented Sep 22, 2017

xref #15003

@jreback

This comment has been minimized.

Contributor

jreback commented Sep 22, 2017

cc @TomAugspurger

seems reasonble.

@tdpetrou

This comment has been minimized.

Contributor

tdpetrou commented Sep 22, 2017

One other thing...if the suffixes are floats and not ints, we could try to cast them to floats as well.

Edit - I went ahead and added a line to cast to float if it was still an object after attempting to cast to int.

@jreback jreback modified the milestones: 0.21.0, Next Major Release Sep 28, 2017

@jreback jreback modified the milestones: Next Major Release, 0.22.0 Nov 19, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment