When I unstack() one level of a DataFrame having only int64 column dtypes (and a MultiIndex created only from int64 values), the result will have float64 columns. This happens even when the unstacked DataFrame does not have any missing data.
Is this expected? Is there any way to prevent unstack() from changing the dtypes?
Behavior confirmed with pandas 0.8.1 and 0.10.1.
In [1]: import pandas as pd
In [2]: rows = [[1, 1, 3, 4],
...: [1, 2, 3, 4],
...: [2, 1, 3, 4],
...: [2, 2, 3, 4]]
In [3]: df = pd.DataFrame(rows, columns=list('ABCD'))
In [4]: print df.dtypes
A int64
B int64
C int64
D int64
In [5]: print df
A B C D
0 1 1 3 4
1 1 2 3 4
2 2 1 3 4
3 2 2 3 4
In [6]: df2 = df.set_index(['A', 'B'])
In [7]: print df2.dtypes
C int64
D int64
In [8]: print df2
C D
A B
1 1 3 4
2 3 4
2 1 3 4
2 3 4
In [9]: df3 = df2.unstack('B')
In [10]: print df3.dtypes
B
C 1 float64
2 float64
D 1 float64
2 float64
In [11]: print df3
C D
B 1 2 1 2
A
1 3 3 4 4
2 3 3 4 4
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4