I have a json line file, where each line has some structure which I am trying to (mostly) flatten, thus:
with open("/home/igor/data/feed.jsonl") as json_file:
thelist2 = []
for line in json_file:
thelist2.append(json_normalize(json.loads(line)))
Followed by pd.concat(thelist2) The semantics of the above is correct, but what is not so good is that this is horrifically slow, while running the above without json_normalize is quite speedy (but does the wrong thing). Is there a way to normalize the dataframe after the fact, or some other speedier scheme?
