Datatype of Integer changes depending on indexing method on a dataframe with an integer and a float column

[Khris777]

Take a dataframe that has one integer and one float column, then use various methods to get the first integer element. In certain cases the datatype will be given back correctly as int64, in other cases however it's changed to float64. Indexing using ix even does both, depending on how it is used:

import pandas as pd

DD = {'I':[1,2,3,4,5,6,7,8,9,10],'F':[1.1,2.2,3.3,4.4,5.5,6.6,7.7,8.8,9.9,10.1]}
DF = pd.DataFrame(DD)

print "Direct index:"
print DF['I'][0]
print DF['I'][0].dtype
print "==========="
print "ix with name:"
print DF.ix[0,'I']
print DF.ix[0,'I'].dtype
print "==========="
print "loc of column:"
print DF['I'].loc[0]
print DF['I'].loc[0].dtype
print "==========="
print "iloc of column:"
print DF['I'].iloc[0]
print DF['I'].iloc[0].dtype
print "==========="
print "ix of column:"
print DF['I'].ix[0]
print DF['I'].ix[0].dtype
print "====================="
print "Values:"
print DF.values[0,1]
print DF.values[0,1].dtype
print "==========="
print "ix with index:"
print DF.ix[0,1]
print DF.ix[0,1].dtype
print "==========="
print "loc:"
print DF.loc[0,'I']
print DF.loc[0,'I'].dtype
print "==========="
print "iloc:"
print DF.iloc[0,1]
print DF.iloc[0,1].dtype

Expected Output

Direct index:
1
int64
===========
ix with name:
1
int64
===========
loc of column:
1
int64
===========
iloc of column:
1
int64
===========
ix of column:
1
int64
=====================
Values:
1.0
float64
===========
ix with index:
1.0
float64
===========
loc:
1.0
float64
===========
iloc:
1.0
float64

output of pd.show_versions()

INSTALLED VERSIONS
------------------
commit: None
python: 2.7.12.final.0
python-bits: 64
OS: Windows
OS-release: 7
machine: AMD64
processor: Intel64 Family 6 Model 62 Stepping 4, GenuineIntel
byteorder: little
LC_ALL: None
LANG: de_DE

pandas: 0.18.1
nose: 1.3.7
pip: 8.1.2
setuptools: 25.1.6
Cython: 0.24
numpy: 1.11.1
scipy: 0.17.1
statsmodels: None
xarray: None
IPython: 4.2.0
sphinx: 1.4.1
patsy: 0.4.1
dateutil: 2.5.3
pytz: 2016.4
blosc: None
bottleneck: 1.1.0
tables: 3.2.2
numexpr: 2.6.1
matplotlib: 1.5.1
openpyxl: 2.3.2
xlrd: 1.0.0
xlwt: 1.1.2
xlsxwriter: 0.9.2
lxml: 3.6.0
bs4: 4.4.1
html5lib: None
httplib2: None
apiclient: None
sqlalchemy: 1.0.13
pymysql: None
psycopg2: None
jinja2: 2.8
boto: 2.40.0
pandas_datareader: 0.2.1

[jreback]

dupe of #11617

pull-request to fix are welcome