IPP Progress Report for the week YYYY.MM.DD - YYYY.MM.DD

(Up to IPP Progress Reports)

Eugene Magnier

* continuing work on the ippToPsps modifications needed to (a) work with the parallel dvo and (b) speed up the general ingest process. I've converted some of the sql in the original ippToPsps code from a pair of insert/select + updates with select into / load from commands. this speeds things up a lot -- a factor of 2 gain in the smf ingest (P2) and something quite large for the stack ingest.

Serge Chastel

Note: This is the report for both my IPP and MOPS activities.

  • Off on Monday (Holiday)
  • IPP
    • ipp001 clean up
  • MOPS
    • Installed the IPP from https://svn.pan-starrs.ifa.hawaii.edu/repo/ipp/tags/ipp-20121218 / Rev.: 35023 (IPP "tag" = ipp-20121218) on the NEO cluster (ipp user account). Two execution contexts: one for megacam, one for detection efficiency. MOPS users who want to use it should give me a public key (send e-mail to schastel@...)
    • Comets classification: talked with Larry
    • Helped BB with MegaCam
    • Setup SOAP interface on NEO cluster
    • Continued with re-implementation of synthetic trails. While exploring the IPP code to implement the ra,dec to pixel coordinates conversion, I discovered that there was a tool (ppCoord) that was doing the job, that is, giving the pixels coordinates and the chip for an exposure when ra and dec and its astrometry (smf) file are provided. I started re-implementing the synthetic trails generation using this tool this weekend. Will be finished and tested on Monday.

Heather Flewelling

Mark Huber

  • md.gr0 -- still mired in cleaning up bugs to config psastro for skycal MD and finish the rerun of MD stack photometry

Bill Sweeney

Chris Waters

  • Added code to psphot to optionally enable extended source fitting for all objects. This will allow MOPS to have trail fits applied to all objects. This still needs a test before being called complete.
  • PSPS work/science. Discovered there is a threshold rate beyond which requests fail. Noted odd counts for Ndetection/Nstack/object.nX, where the object table claims more detections than can be found in the detection and stack tables.
  • Split LAP queues into hour slices for the winter hole.
  • Determined that the OTAs that tend to have higher numbers of diff detections by area are those where we have seen problems before. 2-sigma OTA60, OTA71. 1-sigma OTA01 OTA06 OTA10 OTA16 OTA17 OTA74.
  • Sorted out my issues with ppImage and file IO, and produced first continuity corrected background model. However, this correction is wrong, as the layout of chips and cells are not the same. Fixing this is first goal for next week.

Chris Waters