Getting date index value in a multi index pandas

Question

I have the following dataframe:

a = df['Low'].groupby(['week']).min()
week
1     1699
2     1721
3     1741
4     1809
...
28    2592
29    2570
30    2696
31    2683
53    1735
Name: Low, dtype: int32

To get a date in this i use this function:

df[df['Low'].isin([1735])]['Date1']

Date        week  day
2020-12-31  53    31    2020-12-31
Name: Date1, dtype: datetime64[ns]

To get all the dates when these values occured i used a for loop:

for stuff in low:
    df[df['Low'].isin([stuff])]['Date1']

Which produces :

MultiIndex([('2021-01-06', 1, 6)],
           names=['Date', 'week', 'day'])
MultiIndex([('2021-01-15', 2, 15)],
           names=['Date', 'week', 'day'])
MultiIndex([('2021-01-19', 3, 19)],
           names=['Date', 'week', 'day'])
MultiIndex([('2021-01-27', 4, 27)],
           names=['Date', 'week', 'day'])
MultiIndex([('2021-02-01', 5, 1)],
           names=['Date', 'week', 'day'])
MultiIndex([('2021-02-10', 6, 10)],
           names=['Date', 'week', 'day'])
MultiIndex([('2021-02-19', 7, 19)],
           names=['Date', 'week', 'day'])

I there anyway i could just get the dates? The df looks like this:

            High    Low     Open    Close   Volume  Adj Close   Date1   month
Date    week    day                                 
2020-12-31  53  31  1758.930054     1735    1735.420044     1751.880005     1011900     1751.880005     2020-12-31  12
2021-01-04  1   4   1760.650024     1707    1757.540039     1728.239990     1901900     1728.239990     2021-01-04  1
2021-01-05  1   5   1747.670044     1718    1725.000000     1740.920044     1145300     1740.920044     2021-01-05  1
2021-01-06  1   6   1748.000000     1699    1702.630005     1735.290039     2602100     1735.290039     2021-01-06  1
2021-01-07  1   7   1788.400024     1737    1740.060059     1787.250000     2265000     1787.250000     2021-01-07  1
...     ...     ...     ...     ...     ...     ...     ...     ...     ...     ...
2021-08-02  31  2   2720.409912     2693    2709.689941     2719.790039     1007000     2719.790039     2021-08-02  8
2021-08-03  31  3   2726.709961     2683    2720.000000     2725.600098     953200  2725.600098     2021-08-03  8
2021-08-04  31  4   2730.760010     2708    2724.989990     2720.570068     826400  2720.570068     2021-08-04  8
2021-08-05  31  5   2739.000000     2712    2720.570068     2738.800049     593300  2738.800049     2021-08-05  8
2021-08-06  31  6   2741.601074     2720    2725.899902     2740.719971     678000  2740.719971     2021-08-06  8

151 rows × 8 columns

Answer 1

To get a certain value of a multi level index you can simply use:

df.index.get_level_values('Date')

or

df.index.get_level_values(0)

You could apply this to your for loop like this:

for stuff in low:
    df[df['Low'].isin([stuff])]['Date1'].index.get_level_values('Date')