Re: fitting, step

On Jul 6, 2010, at 5:45 PM, Maurice Leutenegger wrote:

> I am importing a histogram I made in another program. I use "define_counts" to make it into a spectrum I can fit with isis, and I assume a perfect response, fitting individual lines as Gaussians. I can define a Gaussian model with "fit_fun", preset parameter values to be close to the best fit values using "set_par", and use "eval_counts" or "renorm_counts" to make sure that the function evaluates, and renormalize the line strength before fitting the centroid and width. So far so good.
> 
> However, when I do "fit_counts", it doesn't fit the data, or it barely changes the fit parameters. The one clue that something may be wrong is that the "step" field is set to zero for all model parameters I want to fit, even if I explicitly set it using "set_par". I have checked this with XMM and Chandra data, and there, when I load data, define a model, and fit, the fitting algorithm proceeds normally, and the "step" field of each thawed parameter is non-zero. So somehow this seems to be a problem with this data format.

Could you send me a copy of the data and the script you use to load it?  That might help.

One thing that I might worry about is if whether your minimum statistical error were set properly (ISIS defaults that to 1, so it's possible that it's loads bigger than your actual data, depending on how you normalized things.)

Thanks,


Mike



----
You received this message because you are
subscribed to the isis-users list.
To unsubscribe, send a message to
isis-users-request_at_email.domain.hiddenwith the first line of the message as:
unsubscribe