CODEWITHSUNDEEP
pythonpandasdataframe
dvsn
dvsn

Reputation: 17

Lists to columns of dataframe

df1 = pd.DataFrame(np.column_stack([CIK, period, data]), columns=['CIK','Period','Text'])

I have 3 lists which I want to be columns of my dataframe. Above code worked fine when my data was small. Now this gives me memory error. Am I missing something? Is there a different way to do this?

Upvotes: 0

Views: 66

Answers (1)

cs95
cs95

Reputation: 402353

You could build a dataframe by passing a dict to it. 

i = ['CIK','Period','Text']
j = [CIK, period, data]

df = pd.DataFrame(dict(zip(i, j))

This is cheap as it doesn't result in creating copies of your data. The dict simply generates key-value pairs around the references (there's no need to create any data copies, only references are being moved around). Unfortunately, with your column_stack call, the arrays must be stacked into a freshly allocated array and a new result returned, which is wasteful.

Upvotes: 2

Related Questions