IO Conflict Between index_col and dtype

[mbkupfer]

Code Sample, a copy-pastable example if possible

I'm reading in an excel file using the code below. Assume that column 0 is my zip codes and column 2 is my adjustment factor that I'm mapping zips to.

# Your code here
 zip_adjustments_factors = pd.read_excel(fp, sheet_name='zip_mapping', 
    dtype={'zip': object}, index_col=0, usecols=[0,2])

In [1]: zip_adjustments_factors.index.dtype
Out[1]: dtype('int64')

Problem description

In my situation, I need zipcodes to be dtype of object since zipcodes can start with 0's, but the converter under-the-hood is converting these to int's and dropping the 0's.

This requires an extra step as a workaround:

zip_adjustments_factors = pd.read_excel(fp, sheet_name='zip_mapping', 
    dtype={'zip': object}, usecols=[0,2])
zip_adjustments_factors.set_index('zip')

In [1]: zip_adjustments_factors.set_index('zip').index.dtype
Out[1]: dtype('O')

But couldn't this be done within pd.read_excel by checking whether the index column parameter has a specified dtype parameter?

Output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.6.4.final.0
python-bits: 32
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 78 Stepping 3, GenuineIntel
byteorder: little
LC_ALL: None
LANG: None
LOCALE: None.None

pandas: 0.23.4
pytest: 3.7.1
pip: 18.1
setuptools: 40.5.0
Cython: None
numpy: 1.15.2
scipy: 1.2.0
pyarrow: None
xarray: None
IPython: 6.2.1
sphinx: None
patsy: None
dateutil: 2.7.3
pytz: 2018.5
blosc: None
bottleneck: None
tables: None
numexpr: None
feather: None
matplotlib: None
openpyxl: 2.5.5
xlrd: 1.1.0
xlwt: None
xlsxwriter: 1.1.2
lxml: 4.1.1
bs4: 4.6.1
html5lib: 1.0.1
sqlalchemy: 1.2.12
pymysql: None
psycopg2: None
jinja2: 2.10
s3fs: None
fastparquet: None
pandas_gbq: None
pandas_datareader: None

[WillAyd]

Can you provide a sample excel file for this?

[mbkupfer]

Of course. You can recreate with this:

zipcode-example.xlsx

[WillAyd]

The problem here is really the combination of dtype and index_col. As a workaround if you remove the latter can get the result you want:

df = pd.read_excel(fp, sheet_name='zip_mapping', dtype={'zip': object})
df = df.set_index('area_name')

Note that this seems to be a general IO issue not just excel:

pd.read_csv(io.StringIO("a,b\n01,02"), dtype=object, index_col=0)
    b
a
1  02

Investigation and PRs would certainly be welcome

[WillAyd]

Closing as a duplicate of #9435