pandas - how to retrieve the index of the minimum and maximum from two dataframes-CodePudding

I'm looking for a way to always retrieve the row with index and value out of two dataframes.

These two dataframes have a timeseries index of the same length, moreover they both have 1 single column that contains the value of interest.

So far my approach gives me the min index (could be applied to max as well)

pd.concat([df1.col1, df2.col1], axis=1).idxmin()

which gives me

col1   2022-06-04 22:40:00
col1   2022-06-05 01:50:00
dtype: datetime64[ns]

Then I subset the dataframe which has the minimum:

df1.col1['2022-06-04 22:40:00']

But this process is manual and I'd like to always retrieve the minimum, and what if the second dataframe contains the lowest value out of the two? How could I make sure I'm getting the lowest other than manually checking.

CodePudding user response：

you need to add max() command its available in pandas.

PD.COLUMNS.MAX(column name) same for minimum value.

it will help.

CodePudding user response：

Use DataFrame.agg with 4 aggregate functions for get indices and values by min and max:

df1 = pd.DataFrame({'col1':[1,2,6]}, index=pd.date_range('2010-01-02',periods=3))
df2 = pd.DataFrame({'col1':[10,20,6]}, index=pd.date_range('2010-01-02',periods=3))



out = (pd.concat([df1.col1, df2.col1], axis=1, keys=['df1','df2'])
         .agg(['idxmin', 'idxmax', 'min', 'max']))
print (out)
                        df1                  df2
idxmin  2010-01-02 00:00:00  2010-01-04 00:00:00
idxmax  2010-01-04 00:00:00  2010-01-03 00:00:00
min                       1                    6
max                       6                   20


print (out.loc['idxmin', 'df2'])
2010-01-04 00:00:00

print (out.loc['min', 'df2'])
6