TRS Data Analysis

From CoolWiki
Jump to navigationJump to search

Abajian 17:09, 13 December 2011 (PST) -- There was some discussion about our methods, with APT, Python, Excel, etc. Here is my version.

  • Please also see Tim's reminders below (15 August 2011).
  • Feel free to update this list, correct it, etc. Use as much detail or as little, for your poster(s) and report(s).
  • Also, you can view the APT preferences file, Python scripts, etc., from the TRS Data and Tools page.
Steps We Took To Perform Our Analysis
  1. Audrey and Varoujan collected the IRAC1 and IRAC2 FITS files from the Spitzer data base for the 168 AGN we are studying.
  2. Audrey ran a program called SourceExtractor to create a "catalog" file, for each AGN, of all the "sources" present in both IRAC1 and IRAC2 channel FITS files. Each catalog file was an ASCII list of positions, in decimal degrees.
  3. We ran APT v1.1.1 on one or two of the IRAC FITS files to determine our "aperture" of interest. We determined we would use, for the IRAC1 and IRAC2 images, an aperture radius of 6 pixels. We also chose an annulus with inner and outer radii of 6 and 8 pixels, respectively, to represent our "background sky" photometry. The decision was made to utilize the same aperture and annulus for all objects in each IRAC1 and IRAC2 frame.
  4. We told APT to report the photometry to us in microJanskys. The way we did this was to convert the DN reading in APT to flux density by multiplying the DN values in the FITS files by 8.4616 MegaJanskys/steradian. I believe this required us to assume that the pixel resolution was identical for the IRAC1 and IRAC2 detectors.
  5. We ran a script to "bulk process" all 336 FITS files using APT. This resulted in 168 pairs of IRAC1 and IRAC2 data in APT output "table" files. The flux and uncertainties were presented in microJanskys; positions were in decimal degrees.
  6. We learned that we were unable to match up all the positions in the IRAC1 and IRAC2 table files. This was because our original list of positions that we sent to APT in fact included a few positions that did not have valid data in both IRAC1 and IRAC2 channels.
  7. We devised a script to read the IRAC1 and IRAC2 files, compare the positions in each, and write out a resultant file comprised of only those positions truly common to both channels.
  8. We added some additional logic to that script:
    1. we filtered the results to include only those positions that demonstrated SNR > 3 in both IRAC1 and IRAC2 channels; SNR was (flux/uncertainty)
    2. we converted the photometry from microJanskys to magnitudes using the conversion factors provided by Spitzer
    3. output format was converted from TBL to CSV to simplify importing into Excel
  9. We ran this script on all 168 pairs of IRAC1 and IRAC2 tables from APT, and this resulted in 168 CSV files. Each CSV file contained position, IRAC1 magnitude, IRAC1 mag uncertainty, IRAC2 magnitude, IRAC2 mag uncertainty.
  10. We devised an Excel workbook template to allow us to:
    1. read in a CSV file of positions, IRAC1 magnitude and uncertainty, IRAC2 magnitude and uncertainty, for a given AGN
    2. compute IRAC1-IRAC2 colors
    3. specify (by hand) the name and position of the AGN
    4. plot the AGN and all points in the frame with an IRAC1-IRAC2 color in the range of (-.6, -.1)
    5. plot a radius of one arc minute around the AGN to help us visualize potential cluster members
  11. We generated workbooks and plots for each of the 168 AGN frames.
  12. We viewed the plots to gauge which ones might depict galaxy clusters. (You can add criteria here....)
  13. We selected a group of the 10 most likely galaxy cluster candidates.
  14. We re-plotted the 10 candidates to set their X-axes to reflect true RA (East is left).
  15. We downloaded SDSS images of these same regions to perform a more accurate comparison.
  16. We made final comparisons between our plots and SDSS images, and made final suggestions about which may in fact represent galaxy clusters.


Abajian 01:20, 19 September 2011 (PDT) -- Added TRS Data and Tools to supply resources for the analysis effort. Let me know if you have any questions or need any assistance. On that page, you will find the CSV files (magnitudes), as well as a "sample" Excel workbook. The Excel workbook has a macro enabled in it, to refresh the data with a new CSV file. I can describe it better during our telecon, or give me a call.


  • Vg 16:57, 15 September 2011 (PDT) Here is the file with all the names, redshifts and RA and Dec for all our target AGN. Note that the RA and Dec are both in Sexegesimal format (h:m:s) as well as degree format which is what you should be using to plot the data. Make sure you add these AGN into your plots so that you can have a good sense of where the image is centered.


  • Spuck 15:46, 18 August 2011 (PDT) Check this out (5C6.24) what do you think? Is this a possible filament? The AGN is roughly centered on the graph?
    • File:Possible Filament.pdf
    • vg That may very well be!!! We have to check to see that the colors of the sources in the filament match those we would expect from galaxies that would be at the redshift of the AGN.
    • Mehta 16:20, 18 August 2011 (PDT) File:3C6.24.pdf This file has the main graph, plus two more, one with magnitude differences >0.1, and one for <0.1.



  • Spuck 17:31, 15 August 2011 (PDT) A couple things to remember about APT setup for this project.
    • Don't forget to go to "Preferences" and set "Celestial Coordinates Units" to "Decimal".
    • Also click on the "More Settings" button and select "Model B = Sky-median subtraction" for your Sky Algorithm. Source Algorithm should be "Model 0 = No aperture interpolation.
    • Also under "More Settings" check the box next to "Perform new image-data conversion" and use the following values.
    • New image-data conversion factor - 8.4616
    • Image-data units after conversion - µJy
    • Image-data title after new conversion - Flux Density
Retrieved from "https://vmcoolwiki.ipac.caltech.edu/index.php?title=TRS_Data_Analysis&oldid=7966"