Linear regression (R2) fails when boostrapping on very few data points

[IAlibay]

What happens?

1. When plotting dGs, the statistics calculations attempt to generate 1000 boostraps (with replacements) of the input dataset.

2. If one of the boostraps yields the same number for all the values, the linear regression fails (because you can analyze a flat line).

3. If you have 4 data points, this means that there is a 0.39% chance that you will pick the same number 4 times in a row. I.e. ~ 3 boostraps will have the same value for all the data points.

Should this even happen?

Probably not, calculating correlation metrics on so few data points is probably not sensible. It may be that the answer here is to just not report correlation metrics when there are fewer than a set number of data points (at least 10?).