If this is a real world example, I would recommend looking at what the data represents.

If I have lots of values n1, n2 and r, I'm just fumbling about trying to find a relation but... if I know that n1 is the number of sunny days I operated my ice-cream truck in a month, n2 is the number of cloudy days I operated my ice-cream truck in a month and r is the amount of revenue I generated in that month then I could try to fit an equation of the form:

(n1)(r1) + (n2)(r2) = (r)

where r1 is the typical revenue I generate on a sunny day and r2 the typical revenue I generate on a cloudy day. and all that before looking at a single data point.