say I have a categorical column in a DataFrame (for example weekday). And I want to encode it to one-hot-encoding.
I am using pandas.get_dummies()
to do this. But I can't see a way how to make the order deterministic. For example, I have these two dataframes
df1 = pd.DataFrame({'visitors':[220, 240, 180, 210, 220, 260, 270], 'weekday': ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']})
df2 = pd.DataFrame({'visitors':[240, 180, 210, 220, 260, 270, 220], 'weekday': ['Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun', 'Mon']})
And if I call get_dummies()
on both, I get two different encodings. I can see that, as the order is different. But is there a way, how I can have any DataFrame have the same encoding, regardless of how the values come in? So for example: Mon=1000000
, Tue=0100000
, etc?
pd.get_dummies(df1['weekday'])
pd.get_dummies(df2['weekday'])
The behavior of get_dummies
is consistent.
You're seeing the difference because the position of visitors
has changed.
print(df1)
print(pd.get_dummies(df1['weekday']))
print()
print(df2)
print(pd.get_dummies(df2['weekday']))
Here's the output:
visitors weekday
0 220 Mon
1 240 Tue
2 180 Wed
3 210 Thu
4 220 Fri
5 260 Sat
6 270 Sun
Fri Mon Sat Sun Thu Tue Wed
0 0 1 0 0 0 0 0
1 0 0 0 0 0 1 0
2 0 0 0 0 0 0 1
3 0 0 0 0 1 0 0
4 1 0 0 0 0 0 0
5 0 0 1 0 0 0 0
6 0 0 0 1 0 0 0
visitors weekday
0 240 Tue
1 180 Wed
2 210 Thu
3 220 Fri
4 260 Sat
5 270 Sun
6 220 Mon
Fri Mon Sat Sun Thu Tue Wed
0 0 0 0 0 0 1 0
1 0 0 0 0 0 0 1
2 0 0 0 0 1 0 0
3 1 0 0 0 0 0 0
4 0 0 1 0 0 0 0
5 0 0 0 1 0 0 0
6 0 1 0 0 0 0 0