storing priors in 'sol'

[jeffjennings]

storing priors in 'sol'

[rbooth200]

That's a good idea. To be flexible I'd probably add an optional "info" dictionary to the sol object. You can store alpha, etc, plus its easily adaptable to add other notes at a later date

[jeffjennings]

that sounds good yeah; I'd think the best point to add this would be in FrankFitter, in the __init__, as this takes in all the hyperparameters. Or do you think it should be deeper, like in FrankGaussianFit and FrankLogNormalFit?

[rbooth200]

I think you're only storing the sol object, not the fitter so this would need to be deeper. I would probably make FrankRadialFit store the info object, with it being passed through the constructors of FrankGaussianFit and FrankLogNormalFit, in a similar way to the DHT or geometry.

[jeffjennings]

ok I did this in c49fc18 in the lognormal branch; let me know / feel free to change if you think it should be done differently. but that PR will close this issue.

[rbooth200]

Looks good, see the comment I made on the PR