Slice pandas multiindex dataframe using list of index values [duplicate]

Question

I have a multi-index dataframe that looks like

uid tid text

abc x t1

bcd y t2

uid and tid are the indexes. I have a list of uids, and want to get the rows corresponding to the uids in that list, but keeping the 2nd level index values (tid). I want to do it without running any explicit loop. Is that possible?

jezrael · Accepted Answer · 2017-09-30 17:58:22Z

3

Data:

L = ['abc', 'bcd']

print (df)
         text
uid  tid     
abc  x     t1
abc1 x     t1
bcd  y     t2

1.slicers

idx = pd.IndexSlice
df1 = df.loc[idx[L,:],:]

2.boolean indexing + mask with get_level_values + isin:

df1 = df[df.index.get_level_values(0).isin(L)]

3.query, docs:

df1 = df.query('@L in uid')

print (df1)
        text
uid tid     
abc x     t1
bcd y     t2

edited Sep 30, 2017 at 17:58

answered Sep 30, 2017 at 17:46

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Rakib Over a year ago

thanks @jezrael. With approach one, I get 'MultiIndex Slicing requires the index to be fully lexsorted tuple len (2), lexsort depth (0)' error. But 2 works. Which is the most efficient? I have a huge data frame.

jezrael Over a year ago

the fastest is sorting first by df = df.sort_index() and then use first method. sorting is explained 'In 97'

Collectives™ on Stack Overflow

Slice pandas multiindex dataframe using list of index values [duplicate]

1 Answer 1

2 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Linked

Related