IPP Progress Report for the week 2099.11.16 - 2209.11.20

(Up to IPP Progress Reports)

Eugene Magnier

Heather Flewelling

• chased down lots of processing problems
  • thousand of diffs failed, a full ipp016 was the cause, and some incorrect queries also caused problems. Bill and Paul helped out to debug this mess.
  • a few failures from each stage, mostly caused by the 0 byte neb bug. These have been dealt with, and Chris put in a fix that prevents them. The 0 byte bug often occurs when there is a high load on a machine.
• manned summitcopy and registration again, since there was stuff to download. investigated registration failures.
• figured out which raw files exist only on cabinet 2.
• started copying megacam files into nebulous, about halfway done. (do 'neb-ls megacam' to find them)
• started md09 processing. working on diffs and greater. things are distributing. yay.
• started md10 processing. chips are in, should go through to warp this weekend.
• created a nebulous volume log on wiki. ​http://svn.pan-starrs.ifa.hawaii.edu/trac/ipp/wiki/NebulousVolumes
• mask fractions for random chips and warps in throughput.20091104:
  • ​http://svn.pan-starrs.ifa.hawaii.edu/trac/ipp/wiki/WarpMasksStatistics
  • ​http://svn.pan-starrs.ifa.hawaii.edu/trac/ipp/wiki/ChipMasksStatistics

Bill Giebink

• Got DIMM operational by repairing relays; serviced DIMM
• Located and prepared requisitions for replacement and spare relays.
• Attended scaffold training.
• Spent a small amount of time on realtest.
• Worked with Tom and Steph to setup a laptop to serve accelerometer data.

Paul Price

• Stack photometry
  • Warp compared with warp: 0.016 mag (aperture) to 0.024 mag (PSF)

scatter

• Warp compared with stack: 0.076 mag (PSF) scatter
• These STS warps have bad PSFs! Generating own stack with "good" inputs failed because bad PSFs despite iq_m[234] being decent: double images and/or triangular PSFs.
• Kepler is sparse in terms of PS1 coverage; ditto SA92.
• Using Gene's flattest exposures. Photometry comparison residuals:
  • warp-warp PSF 0.010
  • warp-warp Aper 0.010
  • warp-stack PSF 0.016
  • warp-stack Aper 0.014
• Very pleased to see Nigel's values photometry comparison plots: good aperture photometry suggests stack pixels are fine.
• Using simtest to verify that stack pixels are fine.
• Discovered RA=0 boundary bug in DVO/getstar
• 1 simtest stack failed; no time to investigate
• Simtest run in /data/ipp022.0/price/stack/simtest

• Fixed headers (WCS, concept values, software versions) for unconvolved stacks.
• Threading for pmReadoutFake, trunk@26202. If there are problems with this while I'm away, "svn merge -c -26202" should remove this change.
• Fixed division by zero in magic_tree.pl that was preventing STS processing.
• Preparing for UK trip.

Bill Sweeney

• Worked with people at MPIA to get the current version of the IPP built and installed at their site.
• Installed a data store at MPIA for them to use to submit postage stamp requests
• Modified the ipp data store tools to work through a proxy server.
• Helped monitor and debug issues in the destreaking and distribution portions of the pipeline.
• Figured out why diffs were taking so long to bundle. There was a bug in the pending query that caused runs to get bundled multiple times. Fixed this and deleted the duplicates from the data store.
• Finished postage stamp server "get_image" functionality. This will allow users to randomly select and download gpc1 data through the postage stamp server.
• Added the ability to select psf, cmf, and background model image to be included in postage stamp data products.
• At this point the processing portions of the postage stamp server are functionally complete. (The documentation needs to be updated though).

Chris Waters

• Crosstalk: Finished and tested crosstalk masking in psastro. On test exposure, correctly identified and masked all crosstalk artifacts known to follow the rules defined by John Tonry.

• Data quality statistics: Wrote new task/script/tool group to create fits table bundles of exposure statistics for transfer to OTIS. All code is written, but no testing has been performed to confirm functionality.