This data is a few years old, so you can add a little to both:
http://www.newsweek.com/right-wing-extremism-islamist-terrorism-donald-trump-steve-bannon-628381
There are 198,000,000 white people in the US and roughly 3,000,000 Muslims in the US, thus you have you have a population 1.5% the size of the larger population committing half the number of attacks which the larger population committed.
If you want to look at the rates within the extremist populations themselves, then you have to use a model:
115 attacks = e_w * x_w * 198,000,000 moderate white people
63 attacks = e_m * x_m * 3,000,000 moderate Muslims
where
e_w = average number of attacks per white extremist
x_w = average number of white extremists per moderate white person
e_m = average number of attacks per Muslim extremist
x_m = average number of white extremists per moderate Muslim person
thus
5.8*10^-7 = e_w * x_w
2.1*10^-5 = e_m * x_m
Obviously e_m and/or x_m must be relatively larger than e_w and/or x_w to make up the difference. You could call e*x an indicator of relative contribution to terrorist attacks, with a higher value indicating more contribution.