Opened 9 years ago
Closed 8 years ago
#54 closed defect (fixed)
Potential problem when parameters are fixed in the second stage of the fit
| Reported by: | Daniel Craik | Owned by: | Thomas Latham |
|---|---|---|---|
| Priority: | major | Milestone: | v3r1 |
| Version: | Keywords: | ||
| Cc: |
Description
As mentioned in 46, the finalisation performed in LauAbsFitModel::finaliseFitResults assumes that the covariance matrix returned by Minuit contains all of the fit parameters.
If a parameter is fixed in the second stage of the fit then the covariance matrix will be too small. This could be resolved in such cases, by performing a special third stage with all parameters floated.
Change History (2)
comment:1 Changed 8 years ago by
| Owner: | set to Thomas Latham |
|---|---|
| Status: | new → assigned |
comment:2 Changed 8 years ago by
| Resolution: | → fixed |
|---|---|
| Status: | assigned → closed |
Note: See
TracTickets for help on using
tickets.
(In [322]) Remove the feature that allows parameters to float in first stage of a two-stage fit but to be fixed in second, closes #54