Add kwargs for EKP to initialize

[nefrathenrici]

Purpose

To-do

Content

---

• I have read and checked the items on the review checklist.

[costachris]

Should this further be exposed and passed through the calibrate function or as additional arguments in the ExperimentConfig?
So if the latter users can do

experiment_config = ExperimentConfig(; n_iterations, ensemble_size, 
                    observations = y_obs, noise = Σ_obs, output_dir, prior, scheduler = EKP.DefaultScheduler(0.5))

slurm_kwargs = kwargs(time = 90, mem="16G")
eki = calibrate(CaltechHPC, experiment_config; slurm_kwargs, verbose=true)

and if the former:

experiment_config = ExperimentConfig(; n_iterations, ensemble_size, 
                    observations = y_obs, noise = Σ_obs, output_dir, prior)

slurm_kwargs = kwargs(time = 90, mem="16G")
eki = calibrate(CaltechHPC, experiment_config; slurm_kwargs, verbose=true, scheduler = EKP.DefaultScheduler(0.5))

[nefrathenrici]

Thanks @costachris, that would be a great addition. I will expose the EKP constructor to calibrate

[costachris]

Looks good to me, thanks!

[codecov]

Codecov Report

Attention: Patch coverage is 28.57143% with 5 lines in your changes missing coverage. Please review.

Project coverage is 39.66%. Comparing base (c56500f) to head (da29045).

Files	Patch %	Lines
src/backends.jl	28.57%	5 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #92   +/-   ##
=======================================
  Coverage   39.66%   39.66%           
=======================================
  Files           7        7           
  Lines         242      242           
=======================================
  Hits           96       96           
  Misses        146      146           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.