CODEWITHSUNDEEP
pandasrows
TylerNG
TylerNG

Reputation: 941

Adding rows in pandas by matching

I need help with adding 2 rows together for a specific match. Given df:

ID Name1 Name2 Value 
1  xy      a     1
1  yz      b     2
1  zz      c     3
2  xy      a     4
2  yz      b     5
3  zz      c     6

Result: Value of (Name1 =yz,Name2=b) = Value of (Name1 =yz,Name2=b) + Value of (Name1 =xy,Name2=a)

ID Name1 Name2 Value 
1  xy      a     1
1  yz      b     3    (2+1)
1  zz      c     3
2  xy      a     4
2  yz      b     9    (5+4)
3  zz      c     6

I believe we need to group them by id first then do the addition. I usually use np.where to operate on columns but not sure how it works for rows in this case.

Many thanks!

Upvotes: 0

Views: 60

Answers (1)

wazo
wazo

Reputation: 311

I suggest you add a new column: 

df["new_column"] = np.where(df["Name1"] == "yz", df["Value"] + df["Value"].shift(-1),0)

If you want the current value in df["Value"] if the condition isn't met, then just do the following:

df["new_column"] = np.where(df["Name1"] == "yz", df["Value"] + df["Value"].shift(-1),df["Value"])

Upvotes: 1

Related Questions