A compressed file that contains the postage stamps for morphology classification can be found here:
morphology_classification.tar.gz
Please also download this (newer version of the view_galaxies.pro) file:
Information for the classifiers (same as in email sent):
In there you can find some postage-stamps of the galaxies and a little idl code to display them and store your classifications. Let me explain these things one by one in detail so that there is no confusion:
Firstly, the galaxies that I selected for classification are:
- galaxies within our colour and magnitude cuts
- plus all the HI galaxies
- minus galaxies with known redshifts outside the desired range
- minus galaxies outside the field of view of WSRT
This selection yielded 1366 galaxies in A963 and 1269 in A2192, which adds up to a total of 2635 galaxies we need to look at!
After you donwload it and extract morphology_classification.tar.gz, you will see there is a parent folder morphology_classification/ that contains 3 other folders: A963/, A2192/, and A963test/. In each of these you will find R and B-band cutouts of the selected galaxies. Notice that the cutouts are fits files. We chose to leave them in that format so that we can display them as we want in ds9 and really see all the information we can.
The folder A963test/ contains the postage stamps of the INT B and R images of those galaxies for which we have better HST imaging. they are just 12 galaxies and I copied them in that folder separately to test our method and allow you to make your first classifications in that small sample.
I am aware that some of you have never used ds9, if you are familiar with ds9 skip this paragraph. ds9 is an astronomical visualization application. If you don't have it installed, you can get it at http://hea-www.harvard.edu/RD/ds9/. Before trying the script to classify galaxies, try opening any of the .fits files in any of the folders by hand in ds9 to see how the interface looks and to check your installation. From the command line try simply:
ds9 nameoffile.fits
or you can also right-click the file and open with ds9 (hopefully you already have it installed in your computer, it generally comes with scisoft). Once displayed, the most important feature of ds9 is the ability to change the contrast, this can be done by simply right-clicking (and maintaining the click pressed down) at the same time as moving the mouse from one side to the other. The best way is to just try it out. Let me know if you have any problems with ds9 or any questions about it!
To classify the galaxies (in the postage stamps) I created a very simple script called view_galaxies.pro. It is writted in IDL (Does everyone have access to IDL?). If you are not an IDL user, it should not be a problem, because you only need to run it and give it your classification, it is nothing too involved… To run it, you need to go (from a shell) to the parent data directory morphology_classification/ and type in the command line the following:
idl
then, from the idl prompt type:
.r view_galaxies.pro
It will ask you which folder you want to classify, so simply enter the name of the folder, e.g. A963test/ (always include the “/” at the end). It will then tell you: “Press enter to start classifying:”, just to make sure you're ready! As soon as you press enter, a ds9 window should display 2 images of the same galaxy, one in the B-band and one in the R-band (left and right respectively). The changing contrast will only work on one of the images at the time, to make it effective on the other just click on it first. At the same time the images are displayed, the command line will say something like:
On screen: B and R-band images of galaxy 40404
Enter the classification for this galaxy here:
So after you've seen the galaxy and you've decided that its an elliptical for example, you will type the number 1 into the command line and enter (to go to the next galaxy). As simple as that!
The classification scheme we've thought of is the following:
- 1 = Elliptical
- 2 = S0 (or disky early-type, as it is very hard to tell)
- 3 = Spiral: a late-type disk galaxy
- 4 = Irregular
- 5 = Unclassifiable (e.g. too compact)
- 6 = There is not a galaxy in this postage stamp!!!! It can happen that the photometry targets were pointing towards a wrong coordinate for example in the vicinity of bright stars.
- Furthermore, if you see a case of interaction/merger, add a 9 at after the classification number, e.g. “39”, meaning interacting spiral. If its a messy interaction then type 49 (irregular galaxy in interaction)
- Comments can be inputted after the classification number, please include them after a space and a colon and in a single string so that my life is easier afterwords. For example, enter “3 :beautiful_ring_galaxy”
If you want to suggest changes to this scheme please do so soon, before we go ahead classifying all the galaxies!
SO…whenever you enter a classification the code will save it in a file called my_classification.txt inside the folder you first inputted. To guarantee you never overwrite this file, please change the name of this file after having classified a lot. If you forget, the code copies this file into my_classification_OLD.txt whenever you run it again, but the best way to keep things tidy is if you rename the file after each session and save it somewhere.
To abort at any time press control+D and to quit idl just enter “exit”
The output (my_classification.txt) will look like this:
40404 1
42495 3
42512 3
42658 6
42986 1
43041 29
43117 3
43143 5
44184 2
44203 3 :this_is_a_wonderful_comment
44289 3
97008 1
(THIS VALUES ARE NOT MY CLASSIFICATION, I MADE THEM UP TO MAKE THIS EXAMPLE!)
The first column is the galaxy ID, the second is your classification and the third is possible comments.
And that is it. I hope is simple enough….
The only problem I see in the script is that it takes all the files inside the folder you input and will not stop until you have classified them all. If you abort with coltrol+D it will save what you have done so far in my_classification.txt, but next time you run it, it will start from the beginning, so you would have to remember and skip until the one you last classified, and modify file names to save the data… and this is messy. To avoid this problem, I HIGHLY recommend that you dived the data into smaller folders (depending on how many galaxies you feel like classifying at the time) so that you can classify lets say 200 galaxies per time. What I mean is that you have a folder tree that looks like:
morphology_classification/
A963/
A963_1/ A963_2/ A963_3/ A963_4/ A963_5/
A2192/
A2192_1/ A2192_2/ A2192_3/ A2192_4/ A2192_5/
And the data divided into these sub folders…. That will make it easier. I didn't do it for you as I don't know how many galaxies you want to do in one go.
Before you start however, I created a folder called A963test/ that contains B and R images for 12 galaxies only. These galaxies are not random, I chose them because they are the only galaxies in our selection for which we also have nice HST imaging for. Please start by classifying these ones and comparing with the classification Bianca and I have made for these same galaxies (copied below) to see if there are any discrepancies.
Please let me know if you have any questions/comments/complaints/suggestions. Preferably before the 21st of September if possible.
Good luck 
Yara
my_classification.tex for A963test/, made by Bianca and me:
40404 3
42495 2
42512 3
42658 1
42986 3
43041 3
43117 1
43143 1
44184 1
44203 1
44289 29
97008 1