User talk:Hans

Hej Hans! -- I've made a link to your sub-project (the implementation of VWA-SE) for you to start filling out. Don't worry if you don't have a clear idea yet. The important thing is to get started discussing how to do it. So take a look at the other project pages, and start filling in ideas on "your" page.

The link is at Automatic spectral and variability classification in the soon section. -- Tdall 17:30, 13 December 2006 (CLST)

V349 Vel
Hi Hans. I just saw your comments on the page of V349 Vel. I don't understand perfectly: did you perform 2 analysis, one with VASP and one with VWA? How can we make the information you provided clearer? Shouldn't we add another "Analysis" box for your new results? I may look like very picky on the details. That's not completely untrue. :-) But I try to catch up potential problems as soon as possible, with in mind the possible future of VSOP, i.e. managing thousands of stars. You can add your answer just below this text, I'm "following" the page. Cheers. -- 04:03, 16 May 2007 (CLT)


 * Hi Cedric. I just entered the values for the paper for one of the stars (V349 Vel) just to get a bit more used to the wiki editing. In fact we have results for 5-6 stars in paper one. It would certainly make sense to add a format for the VWA-SE results. Perhaps you should think about a general format. What if somebody decides to use a completely different code (say in 2 years) ? Cheers.


 * Hi Hans. Ok, I see a bit clearer. VWA will do basically the same job as VASP. Personally, I don't quite measure the difference and the advantages of one over the other. But anyway, this is very good to have two tools validating the results. I will create a similar box for VWA, but I realize that your software is also a "VASP", although we have coined this term to designate the software of Carlos, based on his work for STELLA. To keep things consistent, I'll change the name of Carlos' VASP to VASP-STELLA and yours to VASP-VWA (I find VWA-SE ugly...). If somebody decides to use a completely different code, we'll create very easily a new template for that! (I often talk about "templates" in the wiki pages. These are special pages containing a bit of "wiki code" to help formating things, and that can be filled with just specifying a few parameters. To understand, check this: Template:AnalysisVASP, and edit the page, without making a modification.) . One last thing: To extensively test the wiki syntax, use the Sandbox page, made just for that! Cheers. -- 07:03, 21 May 2007 (CLT)


 * Hans, for simplicity, I'm simply adding a "code name" parameter in the Template:AnalysisVASP, assuming that you can provide basically the same output. Is that true? I guess yess for logg, Teff and [Fe/H], but what bout "Fit goodness" and a picture? Let me know (on that page, it's fine). -- 07:13, 21 May 2007 (CLT)


 * Ok, Cedric. Additionally, VWA will provide abundances of individual elements. As long as the spectrum is good (S/N around 80-100 or more) and vsini < 35 km/s, VWA can find very precise estimates of the "main" elements like Fe, Ti, Cr, Ca, Ni, etc. VWA does not calculate a goodness-of-fit parameter, but the uncertainties on Teff/logg/[Fe/H] will give you a more detailed idea about the "goodness". As I understand it the VASP-STELLA is basically some fortran/c-code that will be run on the raw data and provide reasonably reliable result quickle. However, VASP-VWA will require that an IDL session is started and at the moment (and a for good long while) it requires some interaction from the user. I could certainly provide a screen shot that will give the user a good idea about the quality of the abundance analysis (e.g. plotting Fe I+II abundance vs line strength or excitation potential). --Hans 08:58, 21 May 2007 (CLT)
 * Hi Hans. This is great news that your code can give us directly abundances. That was one of the reason we wanted Tim to join us: getting abundances and do statistics abundances vs variability type. Even if it's only estimates, since we have everything we need to derive real abundances ( spectras, stellar parameters, and stellar models) this will be of tremendous help. Great great great! 10:20, 21 May 2007 (CLT)