Pandas dataframe row operation within a group-CodePudding

I have a dataframe with information about a stock that looks like this:

Product ID	Initial stock	Quantity
a1	5	-2
a1	5	-1
a1	5	7
a2	10	2
a2	10	-6
b1	14	-9
b1	14	11

Each row represents a purchase/sale of a certain product. I want to add a new column to know the current stock of the product. The first row of every product should be: Current stock = Inital stock Quantity and the next ones: Current stock[i] = Current stock[i-1] Quantity[i]

Product ID	Initial stock	Quantity	Current stock
a1	5	-2	3
a1	5	-1	2
a1	5	7	9
a2	10	2	12
a2	10	-6	6
b1	14	-9	5
b1	14	11	16

In Excel is quite easy, but how would you do it using Pandas? I've tried grouping by Product ID, but I don't know how to introduce the condition of "the first row of every product"

CodePudding user response：

Compute a groupby.cumsum of the quantity and subtract it from the Initial stock :

df['Current stock'] = (df['Initial stock']
                         .add(df.groupby('Product ID')['Quantity'].cumsum())
                       )

output:

  Product ID  Initial stock  Quantity  Current stock
0         a1              5        -2              3
1         a1              5        -1              2
2         a1              5         7              9
3         a2             10         2             12
4         a2             10        -6              6
5         b1             14        -9              5
6         b1             14        11             16