Reasoning

NOTES

Here I fit the data with a simple line. It looks like it fits pretty well. Notice that what we've done is implicitly assume that the variation around the curve is that useless data called "noise".