Having a key error when using group and sum in a dataframe - pandas

Question

I would like to use groupby and sum a csv file
a b c d
1111 0.1 1 1
1111 0 1 0
2222 0.2 1 1
1111 0.2 2 1
2222 1 1
1111 0.3 2 0
3333 0.4 1 1
3333 0.5 2 1
1111 0.6 2 1
e: # if b < 0.2, group column a and sum of column c
f: # If b >= 0.2 group column a and sum of column c
g: # If d = 1, and b >= 0.2, g is sum of c
h: # If d = 0 and b < 0.2, h is sum of c
expected output:
e f g h
1111 2 6 4 1
2222 1 1
3333 3 3
I try:
df1 = df[(df['d'] == 1) & (df['b'] >= 0.2)]
df1.groupby('a')['c'].sum()
However, I got key error in a large file:
pandas.index.IndexEngine.get_loc, pandas.hastable.PobjectHashTable.get_item in column a.

Answer 1

Maybe you can try different approach:
First create conditions columns - e to h and then use mul for filling this mask with values of column c. Last use GroupBy.sum:
df['e'] = df['b'] < 0.2
df['f'] = df['b'] >= 0.2
df['g'] = (df['d'] == 1) & (df['b'] >= 0.2)
df['h'] = (df['d'] == 0) & (df['b'] < 0.2)
print df
a b c d e f g h
0 1111 0.1 1 1 True False False False
1 1111 0.0 1 0 True False False True
2 2222 0.2 1 1 False True True False
3 1111 0.2 2 1 False True True False
4 2222 NaN 1 1 False False False False
5 1111 0.3 2 0 False True False False
6 3333 0.4 1 1 False True True False
7 3333 0.5 2 1 False True True False
8 1111 0.6 2 1 False True True False
df.loc[:, ['e','f','g','h']]= df.loc[:, ['e','f','g','h']].mul(df.c, axis=0)
print df
a b c d e f g h
0 1111 0.1 1 1 1 0 0 0
1 1111 0.0 1 0 1 0 0 1
2 2222 0.2 1 1 0 1 1 0
3 1111 0.2 2 1 0 2 2 0
4 2222 NaN 1 1 0 0 0 0
5 1111 0.3 2 0 0 2 0 0
6 3333 0.4 1 1 0 1 1 0
7 3333 0.5 2 1 0 2 2 0
8 1111 0.6 2 1 0 2 2 0
df1 = df.groupby('a').sum()
print df1[['e','f','g','h']]
e f g h
a
1111 2 6 4 1
2222 0 1 1 0
3333 0 3 3 0