Dividing the sum of one DataFrame column with the sum of another DataFrame column yields 0

Question

I have a DataFrame through which I would like to calculate percentages; however, when attempting to do so, python spits back either 0 or 1.

For example, columns OIR.a1.d.bull and OIR.a1r1.d are arrays filled with 1s and 0s.

>>> data['OIR.a1.d.bull'].sum()
653
>>> data['OIR.a1r1.d'].sum()
320
>>> den = data['OIR.a1.d.bull'].sum()
>>> num = data['OIR.a1r1.d'].sum()
>>> den
653
>>> num
320
>>> num / den
0

I expect to see - roughly - the following.

>>> num / den
0.49004594

works. why? Is it converting to an integer and rounding? – Brian Mar 25 '13 at 18:52 — Brian, Mar 25 '13 at 18:52
See [PEP 238](http://www.python.org/dev/peps/pep-0238/). – Steven Rumbalski Mar 25 '13 at 18:53 — Steven Rumbalski, Mar 25 '13 at 18:53

halex · Answer 1 · 2013-03-25T18:56:36.343

0

You do integer division so your result will be an integer, which means that the fractional part of the division's result is truncated. To get a float you have to convert one of your values to a float.

float(num)/den

edited Mar 25 '13 at 18:56

answered Mar 25 '13 at 18:51

halex

16,253
5
58
67

Since OP is using Python 2.7 he also has the option of doing `from __future__ import division`. – Steven Rumbalski Mar 25 '13 at 19:01
2

@StevenRumbalski You are right and I considered to add this information, but maybe OP uses `/` for strict int division in other places of his code that will break when done as float division. – halex Mar 25 '13 at 19:03

Dividing the sum of one DataFrame column with the sum of another DataFrame column yields 0

1 Answers1