Changing dataframe values as columns - columns - as index

Question

meet the problem of changing values and columns names,

for example i have such Pandas dataframe:

 Value 2013 2014 Company Name 1 Employees 10 12 Aviators 2 Computers 20 2 Aviators 3 Employees 4 13 Mouses 4 Computers 1 8 Mouses 5 Monitors 5 1 Mouses

and i need to make values as colums and years as values or indexes (values better) smth like this:

 Employees Computers Monitors Company name 2013 10 20 NaN Aviators 2014 12 2 NaN Aviators 2013 4 1 5.0 Mouses 2014 13 8 1.0 Mouses

So i need only change years and values, company Name column should be placed like on the sample

What's the easiest way to do it?

Erfan · Accepted Answer · 2019-10-30 12:50:04Z

Using pivot_table with stack:

df = df.pivot_table(index=['Company Name'], columns='Value', aggfunc='sum').stack(level=0).reset_index(level=0) Value Company Name Computers Employees Monitors 2013 Aviators 20.0 10.0 NaN 2014 Aviators 2.0 12.0 NaN 2013 Mouses 1.0 4.0 5.0 2014 Mouses 8.0 13.0 1.0

To remove the column index name, use rename_axis:

df = df.rename_axis(None, axis='columns') Company Name Computers Employees Monitors 2013 Aviators 20.0 10.0 NaN 2014 Aviators 2.0 12.0 NaN 2013 Mouses 1.0 4.0 5.0 2014 Mouses 8.0 13.0 1.0

Edited my answer with aggfunc='sum', can you try that @George
Let's see if this works first, he might have datetime columns instead. @jezrael
With aggfunc it works, but thats the difference between sum and first?

Collectives™ on Stack Overflow

Changing dataframe values as columns - columns - as index

1 Answer 1

10 Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

10 Comments

Related