What's needed
When a saved fit is loaded back in to adjust its plot styling, the drawn fitting
line does not match the line that was shown when the fit was originally run.
A loaded fit should reproduce exactly what the user saw at fit time.
This points at a broader need to review the result export/import pipeline for
validity — not just patch the visible symptom. In particular:
- the discrepancy may stem from older saved results (a format/version mismatch),
so loading legacy exports should be checked
- do exported results carry physical units, and are they applied correctly on
re-import (no double conversion, no unit loss)?
- are exports self-contained — do they include the underlying measurement data
needed to faithfully redraw the fit, or do they rely on state that isn't saved?
Context
Reported via the fitting-app feedback channel. Screenshots were attached showing
the discrepancy between the original and reloaded fit curves.
Depends on #31.
Done when
- Loading a saved fit reproduces the same fit curve that was displayed at fit time
- Loading older/legacy saved results either works correctly or fails loudly with a clear message
- Exported results carry physical units that round-trip correctly on re-import
- Exports are self-contained enough to faithfully reproduce the original plot
What's needed
When a saved fit is loaded back in to adjust its plot styling, the drawn fitting
line does not match the line that was shown when the fit was originally run.
A loaded fit should reproduce exactly what the user saw at fit time.
This points at a broader need to review the result export/import pipeline for
validity — not just patch the visible symptom. In particular:
so loading legacy exports should be checked
re-import (no double conversion, no unit loss)?
needed to faithfully redraw the fit, or do they rely on state that isn't saved?
Context
Reported via the fitting-app feedback channel. Screenshots were attached showing
the discrepancy between the original and reloaded fit curves.
Depends on #31.
Done when