photometric redshifts ...
Today was plots and photometric redshifts day ...First, yesterday I had to re-install kcorrect on my Mac. Then, today, I found out that some definition was preventing me from using SuperMongo ... took a while to find it.
Then I tried Txitxo Benitez's BPZ. Then the test failed, some problem with Python. Then you try this and that, install a new version and nothing helps.
Then you give up and say "I'll use kphotz from kcorrect!!!". That should be easy ... and the results were non-sense ... the same for all the galaxies, even for a test line with all magnitudes equal to 0.00 (hey! it's Vega!!! :-)) And it's at z=0.0301).
Then you give up again and go to another computer. I went to my desktop at ESO. Unpacked BPZ and it works perfectly, while kphotz still tells me all the galaxies are at the same redshift (it's a hell of a cluster at z=0.0301, but Vega is not there anymore).
Now learning some stuff about BPZ. That's what happens when your photo-z guru is on vacation hehehehe
Tomorrow more and I gotta go faster, but I can't stay awake any longer ... small kid at home does that for you.
Labels: hi-z morph
posted by Cris Da Rocha @ 11:56 PM
2 Comments:
You have defined something wrongly. kphotz can give non-sense results, but not that nonsense. I tested it once and could not see my clusters anymore in the redshift distribution. But the redshifts were "coherent".
Now, phot-zs based in broad-band filters give in general very bad redshifts, despite the claims of different groups.
Good luck!
I know phot-zs are bad, but when they're all you can get, I still think they are better than nothing. Let's see.
Post a Comment
<< Home