Repository navigation
Fix hard-coded number of phases #57
Description
The number of phases and the belt_offset describe the winding, so they should be defined in Geometry as n_phases and belt_offset. Currently they are not:
- they are set as
self.phases/self.belt_offsetinComputation.set_oper_params, - they are additionally hard-coded in
phase_groups(Q, p, 3, belt_offset=0)/phase_groups(Q, p, 5, belt_offset=1)in the 60s machines, - the phase count is further hidden in
"ABC"/"ABCDE",range(5),np.zeros((len(time), 5)), etc.
Having the same constants in several places makes the code prone to errors.
Metadata
Metadata
Assignees
Labels
No labels
Activity
haitao-ZCU commentedon Oct 1, 2026
@sadda
I prefer to split this issue into two parts and propose two seperate PRs:
with thses actions, each PR can be as small as possible.
haitao-ZCU commentedon Oct 1, 2026
@sadda
Hi, Lukas, with the parameterizations of n_phase, phase count, and so on, I found we can simplify the process of creating coils in 3f_36s and 5f_40s, e.g we can use general method to create them with loops and parameterizations instead of hard-coding.
if possible, could you please create a new issue about this, and I'll do it.
sadda commentedon Oct 1, 2026
@haitao-ZCU Agreed. Create the issue and then the corresponding PR.
sadda commentedon Oct 2, 2026
This has been performed in #61 and #63.