Categories
pandas python

Sorting Multi-Index to full depth (Pandas)

I have a dataframe which Im loading from a csv file and then setting the index to few of its columns (usually two or three) by the set_index method. The idea is to then access parts of the dataframe using several key combination, as such:

df.set_index(['fileName','phrase'])
df.ix['somePath','somePhrase']

Apparently, this type of selection with multiple keys is only possible if the MultiIndex of the dataframe is sorted to sufficient depth. In this case, since im supplying two keys, the .ix operation will not fail only if the dataframe MultiIndex is sorted to depth of at least 2.

for some reason, when Im setting the index as shown, while to me it seems both layers are sorted, calling df.index.lexsort_depth command returns 1 , and I get the following error when trying to access with two keys:

MultiIndex lexsort depth 1, key was length 2

Any help?