Append multidimensional numpy array to existing pandas dataframe

Asked Dec 30 '17 at 13:57

Active Dec 30 '17 at 13:57

Viewed 2,215 times

I am trying to append a numpy array into an already existing dataframe.

df_train = pd.DataFrame()
sk_train = sklearn_representation_train.toarray()

The shape of both dataframe and sk_train is:

sk_train
-> array([[ 0.,  0.,  0., ...,  0.,  0.,  0.],
       [ 0.,  0.,  0., ...,  0.,  0.,  0.],
       [ 0.,  0.,  0., ...,  0.,  0.,  0.],
       ..., 
       [ 0.,  0.,  0., ...,  0.,  0.,  0.],
       [ 0.,  0.,  0., ...,  0.,  0.,  0.],
       [ 0.,  0.,  0., ...,  0.,  0.,  0.]])

sk_train.shape
-> (95851, 2000)

df_train.shape
-> (95851, 9)

How do convert the sk_train numpy array to dataframe and concatenate it to df_train?

asked Dec 30 '17 at 13:57

bazinga

2,120
4
21
35

Have you looked at `DataFrame.append` or `DataFrame.concat` depending on how you want things to line up? – Jon Clements Dec 30 '17 at 14:00
`x = np.zeros((100,200)); y = np.zeros((100, 90)); pd.concat([pd.DataFrame(x),pd.DataFrame(y)])` – skrubber Dec 30 '17 at 14:05
@skrubber I am facing memory error while performing this operation. – bazinga Dec 30 '17 at 14:18
https://stackoverflow.com/questions/44715393/how-to-concatenate-multiple-pandas-dataframes-without-running-into-memoryerror – skrubber Dec 30 '17 at 14:26

Append multidimensional numpy array to existing pandas dataframe

0 Answers0