For instance:
polarity(data_dictionary_LSD2015) <-
list(simple = list(pos = "negative", neg = "negative"),
expanded = list(pos = c("positive", "neg_negative"), neg = c("negative", "neg_positive")))
then the output would be:
doc_id simple expanded
d1 1.3 1.6
d2 -2.1 -2.3
instead of a single comment called sentiment.