Plot the data without making a light curve

[earlbellinger]

Right now we always try to make a light curve.

I propose that if we specify max fourier degree <1 then we simply phase and plot the data with the specified period, or search for the period if none is specified, or plot the unphased data if period<0.

[dwysocki]

This sounds like a good idea to me.

[dwysocki]

These features will be very useful, but they're a little harder to implement than it'd seem.

One issue is the output table. If max_fourier_degree < 0, then the only information we have for the output table is star names and periods. And that's assuming we don't make period < 0, in which case there will be no output table. So we need to make the output table more flexible than it is now. Perhaps now's the time to think of a more configurable output table.

There are a lot of possible combinations of allowed inputs with this new approach, and we want to handle all of the ones that make sense. We need to do this by making the options as independent as possible, and not by checking for specific combinations, as that complects the program.

[earlbellinger]

We should also have some sort of configurable subplotting. So optionally we will have raw data on the left and each phased period on the right, or something like that.

[dwysocki]

The subplotting argument should allow the user to specify the number of columns, and rows. So if you are expecting 4 subplots, you could do 2 columns and 2 rows, or 1 column and 4 rows, etc. You could also just specify the number of columns, or the number of rows, but not both, and the other parameter would be determined automatically. Or even both could be left unspecified, and it would try to do the same number of both.

[dwysocki]

You said in the original post that we should do this for fourier_degree<1, but actually I think for fourier_degree=0 we should be fitting the A_0 term. So really we should use some other convention, maybe negatives?

[earlbellinger]

Sure, negatives is fine.

Earl Bellinger

On Mon, Nov 2, 2015 at 6:29 AM, Dan Wysocki [email protected]
wrote:

You said in the original post that we should do this for fourier_degree<1,
but actually I think for fourier_degree=0 we should be fitting the A_0
term. So really we should use some other convention, maybe negatives?

—
Reply to this email directly or view it on GitHub
#40 (comment).