Re: ITP: iminuit -- Robust Python minimisation library based around MINUIT2

[Drew Parsons <dparsons@debian.org>]

On 2020-08-08 13:53, Jeremy Sanders wrote:
> Sorry. I forgot to CC the science list while filing the ITP.
>
> You can follow progress on this Bug here: 968075:
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=968075.
>
>
> Package: wnpp
> * Package name    : iminuit
> * URL             : https://github.com/scikit-hep/iminuit
>   Description     : Robust Python minimisation library based around 
> MINUIT2
>
> iminuit is a Jupyter-friendly Python frontend to the MINUIT2 C++ 
> library.
> It can be used as a general robust function minimisation method, but is
> most commonly used for likelihood fits of models to data, and to get 
> model
> parameter error estimates from likelihood profile analysis.
>
> Further comments
...
>  - Other packages which provide similar minimisers are scipy. A 
> comparison
>    of its performance is here:
>    https://iminuit.readthedocs.io/en/stable/benchmark.html


One question: those benchmarks indicate that scipy CG or Powell 
generally performs better.  What is the use-case for minuit? Why not 
just use scipy?