the only use is to make comparison easier since it is just a constant
offset in time

this avoids using "l" which is hard to distinguish from 1, | and I
depending on the font in use

these occur du to the sum of l' not contribution when l' < 0 would be
required

this seems more robust than aligning (the noise) at the beginning to 2pi

whether or not glob results are sorted depends on the file system

Cleaned up and most recent version. Nakano method uses psi4ToStrain, still working on replacing scipy.integrate