We have the following Pandas Dataframe:
# Stackoverflow question
data = {'category':[1, 2, 3, 1, 2, 3, 1, 2, 3], 'date':['2000-01-01', '2000-01-01', '2000-01-01', '2000-01-02', '2000-01-02', '2000-01-02', '2000-01-03', '2000-01-03', '2000-01-03']}
df = pd.DataFrame(data=data)
df['date'] = pd.to_datetime(df['date'])
df
category date
0 1 2000-01-01
1 2 2000-01-01
2 3 2000-01-01
3 1 2000-01-02
4 2 2000-01-02
5 3 2000-01-02
6 1 2000-01-03
7 2 2000-01-03
8 3 2000-01-03
How can we query this dataframe to find the date 2000-01-02 with category 3? So we are looking for the row with index 5.
It should be accomplished without set_index(‘date’).
The reason for this is as follows, when setting the index on the actual data rather than the example data I receive the following error:
InvalidIndexError: Reindexing only valid with uniquely valued Index objects
Answers:
Thank you for visiting the Q&A section on Magenaut. Please note that all the answers may not help you solve the issue immediately. So please treat them as advisements. If you found the post helpful (or not), leave a comment & I’ll get back to you as soon as possible.
Method 1
Take a subset of relevant category, subtract the target date, and get idxmin
tmp = df.loc[df.category.eq(3)]
(tmp.date - pd.to_datetime("2000-01-02")).abs().idxmin()
# 5
Method 2
To get the (first) closest index date with category 3 you could use:
m = df['category'].eq(3)
d = df['date'].sub(pd.Timestamp('2000-01-02')).abs()
d.loc[m].idxmin()
output: 5
Method 3
df[(df['category']==3) & (df['date']==pd.Timestamp(2000,1,2))]
To get the list of all indices:
df.index[(df['category']==3) & (df['date']==pd.Timestamp(2000,1,2))].tolist()
Okay 🙂
All methods was sourced from stackoverflow.com or stackexchange.com, is licensed under cc by-sa 2.5, cc by-sa 3.0 and cc by-sa 4.0