Slice DataFrame in Row Vector

Asked Jan 13 '20 at 14:37

Active Jan 13 '20 at 14:46

Viewed 34 times

If I have a DataFrame like this:

Day Hour   G
  1  0     1  
  1  1     10
  1  2     20
  1  3      2
  1  4      3
  .. ..
  1  23    10
  2  0      2
  2  1      3
  .. ..     ..

If I would get an array like this (without for loop):

     1 1 10 20 2 3 ...
     2 2 3 .. ........

that is: take groups of 23 values and transform each group in a column vector in such a way I'll have N column vectors each of 23 values. In other words: the row index is the Day (1,2,3...) and the columns for every rows are the values of the column G. In more details: if I have one big column vector of 356x23 values (one value for every hour in the year) I would like to get a 2D vector with 356 rows and 23 columns.

Which is the best technique in pandas/python/numpy?

edited Jan 13 '20 at 14:46

asked Jan 13 '20 at 14:37

gdm

7,647
3
41
71

I'm not sure I get your question. Could you explain the desired output and how it's built in more detail? – dhasson Jan 13 '20 at 14:42
In other words, split the input in group indexed by the Day, then for every group transform it in a row vector. – gdm Jan 13 '20 at 14:47
1

This is pivot: `df.pivot(index='Day', columns='Hour', values='G')`. – Quang Hoang Jan 13 '20 at 15:00

Slice DataFrame in Row Vector

0 Answers0