Dear Gijs, et al.,<br><br><div class="gmail_quote">On Thu, Nov 22, 2012 at 6:44 AM, Gijs Verdoes Kleijn <span dir="ltr"><<a href="mailto:verdoes@astro.rug.nl" target="_blank">verdoes@astro.rug.nl</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Francesco et al,<br>
<br>
Many thanks for the input Francesco. I bring this conversation now into the mailinglist space to ease following the discussion.<div class="im"><br></div></blockquote><div><br></div><div>OK! </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
I ran the procedure for all 52 fields. The procedure produces a single plot, with SNR as a function<br>
of 2'' diameter mag. One example is attached here, for the field KIDS_185.0_1.5.r.<br>
Black points are all good (flag==0) objects in the field. In red you can see the median trend, while<br>
the blue curve is the analytic fit<br>
p_1*flux/(p_2*flux+p_3)<br>
where p_1, p_2, and p_3 are free fitting parameters.<br>
</blockquote></div>
Looks good! Could you share the script you are using? Then Hugo and I will make a python version of it as part of the QC script for sourcelists.<br>
<br></blockquote><div><br></div><div>I'm attaching here the fortran program, which is part of KIDSCAT now. In order to compile it, you</div><div>need the port library+pgplot. Binary versions of these libraries were already included in the first KIDSCAT internal release. You should have them already, otherwise, pls, just lemme know.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
The series of 4 (5?) curves are due to discrete steps in integration times due to the dither pattern, right?<br>
The displacement in magnitude seem in rough agreement with that idea:<br>
>>> -2.5*math.log10(4/5.)<br>
0.24227503252014099<br>
>>> -2.5*math.log10(3/5.)<br>
0.55462187404089092<br>
>>> -2.5*math.log10(1/5.)<br>
1.7474250108400469<br>
>>> -2.5*math.log10(2/5.)<br>
0.99485002168009395<div class="im"><br>
<br></div></blockquote><div>Yup, different patterns correspond to objects in regions with different texp (because of dithering). </div><div>Your figures further confirm this. </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
The analytic fit is used to compute the typical mag corresponding to SNR_LIMIT=5 (see dashed and<br>
solid black lines in the plot). This is our limiting mag.<br>
The procedure also computes the same kind of limit but for different SNR_LIMIT(i.e. 10 and 15). These<br>
values are stored in an ascii file (MLIM.dat), into the subdirectory OUTPUT_RES.<br>
</blockquote></div>
Good. Please store the analytic fit parameters as well. Those will come in handy.<div class="im"><br></div></blockquote><div>Done. In the ascii file MLIM.dat, we also have now a third column with p_i.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
I made some tests according to what we mentioned some time ago:<br>
1) I can confirm that computing SNR from either FLUXERR or MAGERR doesn't make any difference at all (just a sanity check);<br>
</blockquote></div>
Good.<div class="im"><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
2) using different detection tresholds does not change significantly the mag lim estimates; the second attached plot<br>
shows the effect of decreasing DETECT_TRESH from 2.5 to 2.0 for the same fit as in the example. The mag lim goes<br>
from 24.77 to 24.78!!<br>
</blockquote></div>
Good to know MAG_APER at 2" diameter is very insensitive to this change in DETECT_THRESH. That might be different for e..g, MAG_ISO. For now I think it is enough to restrict the lim. mag. analysis to MAG_APER with 2". For those we have predictions we can compare to (in KiDS proposal). KiDS team meeting is a good opportunity if/how we need to extend the lim. mag. analysis.<br>
</blockquote><div><br></div><div>Could you kindly provide us the predictions (the ones in hte porposal) for different wavebands ? We (Nicola+me) are performing a systematic comparison to predicted mag_lim's, and want to be sure to use correct values. </div>
<div> </div><div>Of course, the meeting will be an excellenet opportunity to discuss possible extensions....</div><div><br></div><div>Cheers!</div><div>F.</div><div><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Best wishes, Gijs<div class="HOEnZb"><div class="h5"><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Pls, lemme know whatever comment/suggestion you have. Also, do we want/need compute other limiting mags<br>
(e.g. using MAGAUTO/different apertures)???<br>
<br>
Cheers,<br>
Francesco<br>
<br>
</blockquote>
<br></div></div><span class="HOEnZb"><font color="#888888">
-- <br>
------------------------------<u></u>------------------------------<u></u>--------<br>
|dr Gijs Verdoes Kleijn | astronomer |<br>
|<a href="mailto:e-mail%3Averdoes@astro.rug.nl" target="_blank">e-mail:verdoes@astro.rug.nl</a> | OmegaCEN / Kapteyn Institute / |<br>
|www:<a href="http://www.astro.rug.nl/~verdoes" target="_blank">www.astro.rug.nl/~verdoes</a> | Target |<br>
|tel: <a href="tel:%2B31-50-3638326" value="+31503638326" target="_blank">+31-50-3638326</a> | University of Groningen |<br>
|mobile: <a href="tel:%2B31-654658050" value="+31654658050" target="_blank">+31-654658050</a> | postal address: |<br>
| | Kapteyn Astronomical Institute |<br>
| | Postbus 800, 9700 AV, Groningen|<br>
| | The Netherlands |<br>
------------------------------<u></u>------------------------------<u></u>--------<br>
<br>
</font></span></blockquote></div><br>