dtnguyen2
dtnguyen2
```Minuit``` gives the covariance matrix for any kind of statistics without having to call sherpa's unreliable ```covar``` function. I think this is useful especially what @hamogu said at our last...
That's great @DougBurke, comparing it with: ``` (testing-PR) [dtn@devel12 sherpa]$ pytest sherpa/optmethods/tests/test_optmethods.py ============================= test session starts ============================== platform linux -- Python 3.9.7, pytest-6.2.5, py-1.11.0, pluggy-1.0.0 rootdir: /export/sherpa, configfile: pytest.ini collected...
There was a couple of discussions about requiring ```scipy``` (I think from @DougBurke and @hamogu, but I'm too pressed for time to find it but I'm sure you guys can...
All of Sherpa's test functions are unbounded, so I made sure that when testing ```Minuit``` the limits are set accor ```Python def init(name, npar): x0, _, _, fmin = _tstoptfct.init(name,...
hmm... just realized that I had assumed, incorrectly, that all free parameters are around unity. Will have to update to take into account that they may have large magnitudes.
I'm sure there is a fix for the _the sum of models_ if they have the _same_ high resolution grid; However, if they happen to have different high resolution grids...
My branch is all mess up, I will ASAP once I get @Marie-Terrell to help me.
Just found out that there's a bug with one of the transformations (hi limit only). I'm glad I had said that this is still a work in progress!
Here's a question for you, currently the interface is: ```Python def lmdif(fcn, x0, xmin, xmax, ftol=EPSILON, xtol=EPSILON, gtol=EPSILON, maxfev=None, epsfcn=EPSILON, factor=100.0, numcores=1, verbose=0, transform=False): ``` However, I would like to...