Python Forum
Conditional Cumsum in pandas data-frame - Printable Version

+- Python Forum (https://python-forum.io)
+-- Forum: Python Coding (https://python-forum.io/forum-7.html)
+--- Forum: General Coding Help (https://python-forum.io/forum-8.html)
+--- Thread: Conditional Cumsum in pandas data-frame (/thread-26410.html)

Conditional Cumsum in pandas data-frame - klllmmm - Apr-30-2020

I'm trying to get a cumulative total if the code is equal to 100. 

df = pd.DataFrame(data = {'Code':[100,100,200,300,100],
                             'Value':[1000,1000,500,750,1000],
                             'Expected_Cum_100':[1000,2000,2000,2000,3000],})
This was my try, but I couldn't get the expected result 

c = df["Value"].cumsum()
df['Cum_100'] = c.mask(df["Code"] != 100).ffill()
Output:
df
Out[81]: 
   Code  Expected_Cum_100  Value  Cum_100
0   100              1000   1000   1000.0
1   100              2000   1000   2000.0
2   200              2000    500   2000.0
3   300              2000    750   2000.0
4   100              3000   1000   4250.0
Appreciate it if anyone can suggest a way to get the expected cumulative total.

RE: Conditional Cumsum in pandas data-frame - klllmmm - May-01-2020

I managed to do this by creating an intermediary column as a new value.

df['New_Value'] = df.apply(lambda x : x['Value'] if (x['Code'] == 100) else np.nan, axis=1)
c = df['New_Value'].cumsum()
df['Cum_100'] = c.mask(df["Code"] != 100).ffill()
del ['New_Value']
Output:
print(df)
   Code  Expected_Cum_100  Value  New_Value  Cum_100
0   100              1000   1000     1000.0   1000.0
1   100              2000   1000     1000.0   2000.0
2   200              2000    500        NaN   2000.0
3   300              2000    750        NaN   2000.0
4   100              3000   1000     1000.0   3000.0