ALPHA-g-Experiment / alpha-g

MIT License
1 stars 1 forks source link

Bug in track fitting when helix has a very small (subnormal) `h`. #100

Closed DJDuque closed 1 year ago

DJDuque commented 1 year ago

An unreasonable (given the precision we care about) small value of h in a helix, causes the eccentricity in closest_t (1/h^2) to blow to infinity. This creates a NaN value for the closest t.

Probably the best solution would be to deal with these (what threshold) values of h together with h == 0.