Run LIWC word count analysis on Sherlock corpus

>>> import liwca
>>> from liwca.datasets import corpora, dictionaries
>>>
>>> df = corpora.fetch_sherlock()
>>> dx = dictionaries.fetch_sleep()
>>> results = liwca.count(df["text"], dx)
>>> results.head()
Category    WC     sleep
text_id
NN01      1592  0.062814
NN02      1958  0.000000
NN03      2492  0.000000
NN04      1746  0.057274
NN05      1398  0.000000