Ticket #11105 (closed: fixed)
IPeakFunction should have error estimates for its special parameters
| Reported by: | Michael Wedel | Owned by: | Roman Tolchenov |
|---|---|---|---|
| Priority: | major | Milestone: | Release 3.4 |
| Component: | Framework | Keywords: | Fitting |
| Cc: | michael.wedel@… | Blocked By: | |
| Blocking: | #11384 | Tester: | Michael Wedel |
Description
Currently it is not possible to get error estimates on the special parameters of IPeakFunction (centre, fwhm, height), except when there is a 1:1 correspondence to one of the function parameters.
There should be a way to take the covariance matrix after a fit and calculate error estimates for these special parameters, storing them in IPeakFunction. Roman suggested to put that into a separate algorithm.
The algorithm should take the necessary input from the output of Fit (the fit function and the covariance matrix) and return a function that has the error estimates. In addition it could produce a TableWorkspace where the parameter values are given with errors, similar to what Fit outputs for arbitrary function parameters.
Change History
comment:2 Changed 6 years ago by Roman Tolchenov
Re #11105. Added one more unit test.
Changeset: e330b175eb56d257c87e41fa64cec38563f279ed
comment:3 Changed 6 years ago by Roman Tolchenov
Re #11105. Added the doc page.
Changeset: e745044f572254d996366de4550687d682e0ecce
comment:4 Changed 6 years ago by Roman Tolchenov
- Status changed from inprogress to verify
- Resolution set to fixed
This is being verified as pull request #383.
comment:5 Changed 6 years ago by Michael Wedel
- Status changed from verify to verifying
- Tester set to Michael Wedel
comment:8 Changed 6 years ago by Roman Tolchenov
Re #11105. Fixed a couple of small issues.
Changeset: a971ea15cf8fde7d8eb2306f6291ef05ee2e273c
comment:9 Changed 6 years ago by Roman Tolchenov
Re #11105. Fixed a typo.
Changeset: cba461414073c8cacdfad3447edb50d65e3e23d7
comment:11 Changed 6 years ago by Michael Wedel
- Status changed from verifying to closed
Merge pull request #383 from mantidproject/11105_peak_parameters_error_estimates
Everything works as described. I have created a ticket to use the algorithm in PoldiFitPeaks2D (#11384(http://trac.mantidproject.org/mantid/ticket/11384)) where it will be quite helpful.
Full changeset: 5a9d9596ff332628b871113a25a0278dd5d652cc
comment:12 Changed 5 years ago by Nick Draper
Somehow these slipped through without a resolution. Set to Fixed.
comment:13 Changed 5 years ago by Stuart Campbell
This ticket has been transferred to github issue 11944
Re #11105. Added the algorithm.
Changeset: 5bc2a27ee12b927b7ccfb4725bda3ce8065afaa3