Dictionary of nested lists to pandas DataFrame

问题

I'm trying to learn how pandas works but I assume I'm missing something obvious.

I have a file looking like this :

dict_spl ={'doc1':[[('word11',1,1),('word12',1,2)]], 'doc2':[[('word21',2,1),('word22',2,2)]]}

And I'm trying to obtain a pandas DataFrame looking like this:

# doc1 word11 1 1
# doc1 word12 1 2
# doc2 word21 2 1
# doc2 word22 2 2

I haven't found a way to create both new columns and new rows while duplicating the common values.

回答1:

You can use:

a = [[(k, *y) for y in v[0]] for k,v in dict_spl.items()]
a = [item for sublist in a for item in sublist]

df = pd.DataFrame(a, columns=list('abcd'))
print (df)
      a       b  c  d
0  doc1  word11  1  1
1  doc1  word12  1  2
2  doc2  word21  2  1
3  doc2  word22  2  2

I feel there is better solution, so I asked here:

#Martijn Pieters♦'s solution
a = [(k, *t) for k, v in dict_spl.items() for t in v[0]]
df = pd.DataFrame(a, columns=list('abcd'))
print (df)
      a       b  c  d
0  doc2  word21  2  1
1  doc2  word22  2  2
2  doc1  word11  1  1
3  doc1  word12  1  2

来源：https://stackoverflow.com/questions/45215820/dictionary-of-nested-lists-to-pandas-dataframe