0 votes
50 views

i have calculated the median absolute deviation of the g-band time-series data (with at least 5 valid measurements) for my list of targets in the smc region from nsc dr1. however, i found out that there are some patterns existed in the data, where the obvious difference of the mad values appears in the different parts of the region as can be seen below. i guess this is some kind of systematic effect?

https://www.dropbox.com/s/ol8b7c908p3lc5u/nsc_g_mad.jpg?dl=0

asked Mar 19 by deepeyesym (240 points)
edited Mar 19 by deepeyesym | 50 views

2 Answers

0 votes
We've asked David Nidever, who constructed the catalog, to take a look.  He's on travel, however, so it may take a bit of time.
answered Mar 19 by kolsen (2,660 points)
0 votes
That's a very interesting figure.  I haven't seen that before.  Could you send the code that you used to construct it so I can take a closer look?  Thanks.

David
answered Mar 27 by anonymous
Hi David,
      The process of generating this figure is a bit complicated including retrieve large amount of time-series data from NSC DR1. However, there is a much simple way to have a similar figure. You can just load the data from nsc_dr1.object table (my figure shows the data for the SMC region with 3<RA<25 and -75.5<Dec<-70; note that I have also deblened the catalog by excluding any targets having neighbors within 3 arcsec, but I don't think it has any major influence on this figure), choose targets with (u,g,r,i,z)rms<10 to avoid the 999999.0 value, and plot the spatial distribution with z-axis color coded with rms (you may need to adjust the range of z-axis to make it more obvious, e.g., 0~0.1 or 0~0.2). I have checked my data and found the patterns in g,r,i,z bands.

130 questions

118 answers

90 comments

714 users

Welcome to Data Lab Help Desk, where you can ask questions and receive answers from other members of the community.