从multiindex过滤列表中的特定日期

Question

I have a multi index data in which i would like to filter out a list of specific dates eg :

  date_list=[Timestamp('2018-05-19 00:00:00'),
 Timestamp('2018-06-24 00:00:00'),
 Timestamp('2014-11-12 00:00:00'),
 Timestamp('2015-11-11 00:00:00'),
 Timestamp('2012-05-28 00:00:00'),
 Timestamp('2012-06-23 00:00:00')] 

I try to filter out these dates by the following but it does not work:

df.iloc[df.index.get_level_values('Date') != date_list] 

Can anyone please help.

Answer 1

Use Index.isin with inverted boolean mask by ~ , iloc should be removed because filter by boolean indexing :

Notice:

Check if DatetimeIndex before filtering:

print (df.index.get_level_values('Date'))

---

df1 = df[~df.index.get_level_values('Date').isin(date_list)]

Another solution with drop withparameters level and errors :

df1 = df.drop(date_list, level='Date', errors='ignore')

Sample :

df = pd.DataFrame({'Date':['2018-05-19','2014-11-10','2018-06-24','2014-11-13'],
                   'ID':[1,1,2,2],
                   'Val':list('abcd')})

df['Date'] = pd.to_datetime(df['Date'])
df = df.set_index(['Date','ID'])
print (df)
              Val
Date       ID    
2018-05-19 1    a
2014-11-10 1    b
2018-06-24 2    c
2014-11-13 2    d

date_list=[pd.Timestamp('2018-05-19 00:00:00'),
 pd.Timestamp('2018-06-24 00:00:00'),
 pd.Timestamp('2014-11-12 00:00:00'),
 pd.Timestamp('2015-11-11 00:00:00'),
 pd.Timestamp('2012-05-28 00:00:00'),
 pd.Timestamp('2012-06-23 00:00:00')] 

---

print (df.index.get_level_values('Date'))
DatetimeIndex(['2018-05-19', '2014-11-10', '2018-06-24', '2014-11-13'], 
              dtype='datetime64[ns]', name='Date', freq=None)

df1 = df[~df.index.get_level_values('Date').isin(date_list)]
print (df1)
              Val
Date       ID    
2014-11-10 1    b
2014-11-13 2    d

df1 = df.drop(date_list, level='Date', errors='ignore')
print (df1)
              Val
Date       ID    
2014-11-10 1    b
2014-11-13 2    d