wish list: frequency search grid

Suggestions for improvements, new features, etc.
Post Reply
ccameron
Posts: 12
Joined: Sun Apr 07, 2013 11:35 am

wish list: frequency search grid

Post by ccameron » Wed May 22, 2013 9:07 am

Hi Rich,

I'm thinking it may be a useful feature to read in either a user defined frequency grid to search for modes or to read in a previously generated summary file and make a search grid from that.

The usefulness in the first case would be in 'quick' searches for observed frequencies for a given model.

For large grids of stellar models, when there is a small change in \Delta\nu between time steps, the previous model's frequencies will be close enough to the
next model's frequencies that it is more efficient to start from those. That's where the second case may be the most useful.

Thanks, and I'm more than happy to help out with testing if you think this would be a useful feature.

Chris

User avatar
rhtownsend
Site Admin
Posts: 284
Joined: Sun Mar 31, 2013 4:22 pm

Re: wish list: frequency search grid

Post by rhtownsend » Thu May 23, 2013 10:45 am

ccameron wrote:
> Hi Rich,
>
> I'm thinking it may be a useful feature to read in either a user defined
> frequency grid to search for modes or to read in a previously generated
> summary file and make a search grid from that.
>

Preference for one or the other? Or both?

For the previously-generated case, ideally the grid should interlace the frequencies for the previous summary file, right? That is, we should place a grid point between each pair of summary-file eigenfrequencies?

> The usefulness in the first case would be in 'quick' searches for observed
> frequencies for a given model.
>
> For large grids of stellar models, when there is a small change in
> \Delta\nu between time steps, the previous model's frequencies will be
> close enough to the
> next model's frequencies that it is more efficient to start from those.
> That's where the second case may be the most useful.
>
> Thanks, and I'm more than happy to help out with testing if you think this
> would be a useful feature.

ccameron
Posts: 12
Joined: Sun Apr 07, 2013 11:35 am

Re: wish list: frequency search grid

Post by ccameron » Thu May 23, 2013 3:14 pm

Hi Rich,

>Preference for one or the other? Or both?
>For the previously-generated case, ideally the grid should interlace the frequencies for the previous summary file, right? That is, we should place a grid >point between each pair of summary-file eigenfrequencies?

Reading a frequency list from a summary file would be best as it's a standard output from gyre. In principle a user can generate a file in the summary format that has any frequency list they want. I'll probably write a python script to do that.

Let's say the routine would read from the summary file:

a) The list of modes. I'm not sure what would be best for the nad case: nu_ad, real(nu_nad), or just nu_nad ... coder's choice :)
b) The \ell value (ignore \ell from &osc).
c) The frequency unit (ignore freq_units from &scan).

The routine could use the frequency bounds from the list of modes from the summary file and augment that grid by placing n_freq (from &scan) points between modes, spaced as grid_type (from &scan).

I think that would cover most cases ...?

Thanks,

Chris

Post Reply