dividing dataframe by a total column

Question

I have a dataframe of the following shape:

 country    date    col1    col2    col3    col4    col5   total

The columns that start with col are floats.

I want to replace (or create a new dataframe) each value with % of total, as in, for each date and country,

new_col1[row0]=col1[row0]/total[row0]

I tried the following:

df2=df.copy()

df2.select_dtypes(include=['float64']).div(df2.total)

I get all NaNs, and there are no missing values in the dataframe.

If I do each column individually, it works just fine.

Any help appreciated!

Answer 1

Is this waht you are after?

df = pd.DataFrame({'Col1': [1.0, 2.0, 3.0], 'Col2': [4.0, 5.0, 6.0]}) #Data
df.dtypes#Check data format

#Outcome of check Col1 float64 Col2 float64 dtype: object

#solution

df[['diva', 'divb']] = df[['A','B']].transform(lambda x: x / x.sum(), axis=1)

Or are you after sum of rows divided by count?

df['div']=df.apply(lambda x: (x.sum()/(len(x))), axis=1)