pandas diff() giving 0 value for first difference, I want the actual value instead

Question

I have df:

Hour  Energy Wh  
1        4          
2        6           
3        9
4        15

I would like to add a column that shows the per hour difference. I am using this:

df['Energy Wh/h'] = df['Energy Wh'].diff().fillna(0)

df1:

Hour  Energy Wh  Energy Wh/h
1        4          0
2        6          2 
3        9          3
4        15         6

However, the Hour 1 value is showing up as 0 in the Energy Wh/h column, whereas I would like it to show up as 4, like below:

Hour  Energy Wh  Energy Wh/h
1        4          4
2        6          2 
3        9          3
4        15         6

I have tried using np.where:

df['Energy Wh/h']  = np.where(df['Hour'] == 1,df['Energy Wh'].diff().fillna(df['Energy Wh']),df['Energy Wh'].diff().fillna(0))

but I am still getting a 0 value in the hour 1 row (df1), with no errors. How do I get the value in 'Energy Wh' for Hour 1 to be filled, instead of 0?

Answer 1

You can just fillna() with the original column, without using np.where:

>>> df['Energy Wh/h'] = df['Energy Wh'].diff().fillna(df['Energy Wh'])
>>> df
      Energy Wh  Energy Wh/h
Hour
   1          4          4.0
   2          6          2.0
   3          9          3.0
   4         15          6.0