在sorted（）中的Python lambda函数

Question

This woking code gives me the 5 most relevant documents for a topic out of my corpus.

most_relevant_docs = sorted(bow_corpus, reverse=True, key=lambda doc: abs(dict(doc).get(topic_number, 0.0))) 
print most_relevant_docs[ :5]

But since the corpus is not readable by human I want to zip an index to the corpus so I can recover the depending documents.

corpus_ids = range(0,len(corpus))
most_relevant_docs = sorted(zip(corpus_ids, bow_corpus), reverse=True, key=lambda my_id, doc : abs(dict(doc).get(topic_number, 0.0)))
print most_relevant_docs[ :5]

Where do I have to adapt the lambda function so it returns the id together with the document?

Answer 1

Try this

sortingFunc = lambda doc: abs(dict(doc).get(topic_number, 0.0))
corpus_ids = range(0,len(corpus))
most_relevant_docs = sorted(zip(corpus_ids, bow_corpus), reverse=True, key=lambda pair: sortingFunc(pair[1]))

When you zip it, each element becomes like (index, value) , so the original sorting key wouldn't work. You'd need to modify it so it sorts by the value as opposed to the pair