CODEWITHSUNDEEP
pythonpandasdataframe
Thustra
Thustra

Reputation: 329

Pandas: creating values in a column based on the previous value in that column

Quick example:

Before:

In  Out    
1    5    
10   0    
2    3

After

In  Out  Value    
1    5   -4    
10   0    6    
2    3    5

So the formula here is Value(rowx) = Value (rowx - 1) + In(rowx) - Out(rowx).

I started with adding a Value column where each cell is 0. I then have looked a shift() but that uses the value in the previous row from the start of the command/function. So it will always use 0 as the value for Value. Is there a way of doing this without using something like iterrows() or a for loop ?

Upvotes: 0

Views: 310

Answers (1)

furas
furas

Reputation: 142641

It seems in your calculations you could first calculate In - Out and later use cumsum()

import pandas as pd

data = {
    'In': [1,10,2],
    'Out': [5,0,3]
}

df = pd.DataFrame(data)

df['Value'] = df['In'] - df['Out']
df['Value'] = df['Value'].cumsum()

print(df)

or even shorter

df['Value'] = (df['In'] - df['Out']).cumsum()

Upvotes: 4

Related Questions