dividing selected columns in pandas - pandas

Question

This is the dataframe:
bins year binA binB binC binD binE binF binG binH
0 1998 4.0 5.0 1.0 1.0 2.0 0.0 1.0 0.0
1 1999 4.0 2.0 1.0 0.0 0.0 4.0 1.0 2.0
2 2000 4.0 1.0 1.0 0.0 4.0 1.0 1.0 2.0
3 2001 2.0 1.0 4.0 1.0 1.0 0.0 2.0 3.0
My goal is to divide binA through binH by sum of binA:binH or for row 1998, divide by the sum of the row excluding the year number.
Sum of desired columns:
newdfdd.loc[:,'binA':'binH'].sum(axis=1)
To get the desired value this is what I have tried:
newdfdd[['binA','binB','binC','binD','binE',
'binF','binG' ,'binH']].div(newdfdd.loc[:,'binA':'binH'].sum(axis=1))
But, I get NaN and four extra columns as following:
0 1 2 3 binA binB binC binD binE binF binG binH
0 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
1 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
2 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
3 NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
I want results in the following format:
bins year binA binB binC binD binE binF binG binH
0 1998 0.285 0.357 ... .... .... .... ... ...
1 1999 .. .. .. .. .. .. .. ..
.... means some number from calculation.
What do I need to edit in my code for the desired output?

Answer 1

In the div statement you need to provide the axis='index' and it should get the result you are looking for.
So your code above should look like:
newdfdd.update(newdfdd.loc[:,'binA':'binH'].div(newdfdd.loc[:,'binA':'binH'].sum(axis=1),
axis='index'))
This will compute your percent of row sum as desired and then update the values inplace in the newfdd dataframe.
Here is the entirety of my solution for clarity (I used df, and random variables, but the rest is the same):
df = pd.DataFrame({'bins':[0,1,2,3],
'year':[1998,1999,2000,2001],
'binA':np.random.randint(1,10,4),
'binB':np.random.randint(1,10,4),
'binC':np.random.randint(1,10,4),
'binD':np.random.randint(1,10,4),
'binE':np.random.randint(1,10,4),
'binF':np.random.randint(1,10,4),
'binG':np.random.randint(1,10,4),
'binH':np.random.randint(1,10,4)})
#reodering columns to match your dataframe layout
df = df[['bins','year','binA','binB','binC','binD','binE',
'binF','binG' ,'binH']]
df.update(df.loc[:,'binA':'binH'].div(df.loc[:,'binA':'binH'].sum(axis=1),axis='index'))
print(df)
bins year binA binB binC binD binE binF binG binH
0 0 1998 0.222222 0.037037 0.148148 0.185185 0.037037 0.111111 0.037037 0.222222
1 1 1999 0.264706 0.058824 0.205882 0.058824 0.029412 0.147059 0.176471 0.058824
2 2 2000 0.166667 0.041667 0.145833 0.020833 0.166667 0.166667 0.145833 0.145833
3 3 2001 0.062500 0.187500 0.020833 0.145833 0.083333 0.166667 0.166667 0.166667

Answer 2

I think this is the result you are looking for:
df['rowSum'] = df[df.columns[2:]].apply(sum, axis=1)
df[df.columns[2:]].apply(lambda x: (x / x['rowSum']), axis=1).drop(columns=['rowSum'])
binA binB binC binD binE binF binG binH
0 0.285714 0.357143 0.071429 0.071429 0.142857 0.000000 0.071429 0.000000
1 0.285714 0.142857 0.071429 0.000000 0.000000 0.285714 0.071429 0.142857
2 0.285714 0.071429 0.071429 0.000000 0.285714 0.071429 0.071429 0.142857
3 0.142857 0.071429 0.285714 0.071429 0.071429 0.000000 0.142857 0.214286