Hardcoded xmin in fragmentation analysis to 1 and do not scale with epoch length.

[vincentvanhees]

Following comment in Google group:
https://groups.google.com/g/RpackageGGIR/c/rGhumdrlNrk/m/nzeGkBfiAAAJ

Change g.par5_analyseSegment.R call to g.fragmentation.R in
https://github.com/wadpac/GGIR/blob/master/R/g.part5_analyseSegment.R#L398
such that object xmin is always 1, regardless of epoch length.

[vincentvanhees]

... actually this needs more careful attention.
Alternative cause may be to do epoch length / 60 instead of 60 / epoch length.