fitting advice?

[maddiecain]

When I press "Fit Model" when working on synthesis, almost every time I press it the abundance updates to a new number, even when I haven't changed any of the [X/Fe] values. Is this how it's supposed to be working? If so, should I just press "Fit Model" a few times until it stops updating and stays at a constant value?

[andycasey]

I see this behaviour sometimes too. It is definitely not expected behaviour. If you can get it to converge near the same value, do so, but me and Alex will be working towards a permanent fix as soon as possible. It could be a tricky problem to solve, so if you can write down what session this is and which spectral model it is, it may be useful for testing later.

Sent from my Commodore 64

On 5 Aug 2016, at 22:16, maddiecain notifications@github.com wrote:

When I press "Fit Model" when working on synthesis, almost every time I press it the abundance updates to a new number, even when I haven't changed any of the [X/Fe] values. Is this how it's supposed to be working? If so, should I just press "Fit Model" a few times until it stops updating and stays at a constant value?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[maddiecain]

sure, i noticed it first when working on MP2318 on line CH4312 but I also noticed it happening on the Eu lines

[alexji]

This is a generic problem on all syntheses. I'm closing this for now because indeed the problem just tends to be that the fit has not converged in the number of iterations allowed. Since the real solution is hard to solve and the practical solution is just to fit multiple times, I think this is in the "not-worth-fixing" category.

Note that we don't really want the fit to go to completion necessarily, because e.g. it autofits when you mask the spectrum, and you don't want that to take too long.