Changes between Version 11 and Version 12 of MSEReport20110215


Ignore:
Timestamp:
03/24/2011 04:06:50 PM (8 years ago)
Author:
Tim Axelrod
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • MSEReport20110215

    v11 v12  
    5151|| Enable rescheduling of failed job || 3/22/11 || 4/4/11 || Pietrowicz Steve[75%] || 0% ||  ||  || 
    5252|| Create Calibrated Image on-the-Fly from Raw plus Provenance || 2/23/11 || 3/8/11 || Lim KT[25%] || 0% ||  ||  || 
    53 || ''MOPS Analyses'' || 10/10/08 || 4/19/11 || ''Axelrod Tim[0%]'' || 53% ||  ||  || 
     53|| ''MOPS Analyses'' || 10/10/08 || 4/19/11 || ''Axelrod Tim[0%]'' || 53% ||''Candidate tracks found by MOPS are evaluated for plausibility prior to sending them off to the orbit fitter, which is a computationally expensive task.  The plausibility check in MOPS involved fitting a quadratic in time to the ra and dec of the observed points, and cutting on the rms of the fit.   This cut was not tied in any quantitative way to the actual probability that the candidate is a real track.   In the last month we have completely overhauled this section of MOPS to use a probability of chi_squared formalism.   Crucially, it was recognized that topocentric corrections must be applied before fitting the polynomial in ra, and this has been incorporated into the procedure.  The result is a plausibility cut which is still computationally much faster than full orbit fitting, but is at least 10x better in discriminating true from false tracks.   We have also quantified the capabilities of MOPS on different classes of solar system objects.  This has confirmed our intuition that, with the current limits on velocities and acceleration, our NEO detection efficiency falls well short of 50%.   We are investigating a hybrid approach in which MOPS detects and eliminates the large population of main belt objects, and then leaves the remaining fast movers to a more specialized code.''   ||  || 
    5454|| Complete analysis of runs of FindTracklets/LinkTracklets and OrbFit driven by simulated DIASources || 10/10/08 || 3/4/11 || Axelrod Tim[20%],Myers Jonathan[75%] || 75% ||  || 3/23/11  || 
    5555|| Estimate amount of CPU required for LSST MOPS || 11/8/10 || 3/11/11 || Axelrod Tim[20%],Myers Jonathan[75%] || 50% ||  || 3/23/11 ||