Why I get float when I changed it to integer

Question

I am writing a function to change from float to integer. Right now, for example, if the value is 1.34, the code returns 1.0 instead of 1. How to solve this problem?

dataframe.iloc[:,i] = dataframe.iloc[:,i].apply(lambda x: int(x) if str(x) != 'nan' else x)

Your datafame's dtype is probably a float type. You cannot store int type in a float dataframe. — Akaisteph7, Jul 04 '19 at 04:24
If I never add the if else clause, I can get it in int format — Ye Htet, Jul 04 '19 at 04:25
It probably casts to the most flexible type. the else leaves some floats (with `nan`) value in the dataframe along with the ints. — flakes, Jul 04 '19 at 04:30

U13-Forward · Accepted Answer · 2019-07-04T04:32:47.673

3

It's because NaN is a float:

>>> type(np.nan)
<class 'float'>
>>>

Even 1 out of a million values is a float, than the whole thing is a float.

So you could use:

df.iloc[:, i] = df.iloc[:, i].astype('Int64')

You can do Int64, Int32 or Int16.

edited Jul 04 '19 at 04:32

answered Jul 04 '19 at 04:30

U13-Forward

69,221
14
89
114

Why I get float when I changed it to integer

1 Answers1