ALBIUS: Achievements in reporting period: 2010 Q1

Last update Apr 12 14:00, reports are in from UCAM, MPI, ASTRON, NRAO, ESO, BORD, JIVE and UOXF

List of working documents

* Reports about CASA tests (ESO::DP)

Main Activities during 2010 Q1 and Work Plan for Q2 and Q3

  • 1.1. JIVE: work on interoperability
    • Produced Obit/ParselTongue distribution built against Casa. Allows easy installation, and use of both Casa & ParselTongue from the same Python interpreter.
    • Partially implementation of table conversion from AIPS to MS calibration tables.
    • Work on finding, solving problems related to use of EVN MS datasets in Casa.
  • 1.1.2 ESO Dirk Petry, is the developer for ALBiUS.
    • DP developed within CASA routines to convert MS CASA format into ALMA ASDM format
    • DP collected information on the existing stump of the FITS IDI interface in CASA. Scanned the documentation for FITS IDI including the recent update from March 22.
    • Performed detailed CASA testing on version 3.0.0.
    • Read publications of wide field imaging.
  • 1.2 JIVE work on parallel ParselTongue
    • Made ParselTongue work with newer Obit versions. Newer Obit versions are needed to allow adding “AIPS disks” on the fly, which in turn is needed to integrate AIPSLite in a user-friendly fashion.
    • Started investigating IPython as a vehicle to make parallelization available in a more user-friendly way.
  • 2.2.1. NRAO Some bug fixes for CASA mosaicing.
  • 2.2.4. ASTRON The work on task 6.2.5 (Distributed imaging) is continuing. An initial version for facet based correction was implemented.
  • 2.3 BORD Work for this sub-task will start within the next two weeks. Little has been done so far since I am waiting for the new hired person to start. Detailed work plan to be established soon. Laurent Chemin has been hired and he will start on 15 April 2010 (two-year postdoc with 1 year funded by FP7-RadioNet and 1 year funded by the Regional Council of Aquitaine).
  • 3.2. UOXF Evaluation of data quality characterization algorithms and excision
    • Investigated the AIPS flagging limitation. Developed an algorithm to bypass the flagging limitation within AIPS. Bottleneck: the procedure is very IO heavy.
    • Based on current GMRT datasets a baseline based bandpass calibration scheme and diagnostic plots has been developed and further quality assessment tools are planned in order to investigate bandpass problems (synergies with fringe finding [e.g. delay /delay rate plots]).
    • Synergies with the PrepSKA program PELICAN in Oxford have been identified which will result in the matching effort.
  • 3.3.1 MPI Hiring complete, Wagner starts April 12.
  • 3.3. ASTRON The work on task 6.3.3 (Source parameterization) has started

Meetings attended by people involved

  • van Langevelde attended RadioNet exec board.
  • ESO staff attended internal CASA meetings: February 26th, March 26th.
  • Laing visited Jodrell Bank (to discuss JBCA work on direction-dependent calibration; March 31) and Oxford (new calibration algorithms and CASA HPC; April 1st).
  • Jan Wagner (MPI) attended RFI mitigation workshop, Groningen March 29-31.

Problems / Issues

  • Actions were taken to get the South African partner on board for the fringe fitting task. A contract change was requested and is in progress.
  • Program was clearly underspending in 2009, progress with hires at UCAM, BORD, MPG indicate this will improve
  • We see the need for improved communications between ESO and JIVE at working level and a more active management (e.g. circulation of meeting minutes, responses to identified action items).
  • 2.3 BORD Work will start with a significant delay compared to the initial plan but should be completed by the end of FP7-RadioNet (12 + 9 months due by Bordeaux)
  • Bordeaux has issues with time-sheets that they are unclear about.
  • The Oxford contribution to ALBiUS has been significantly delayed due to issues regarding the availability of matching effort resulting from funding issues in the UK. Therefore the Oxford ALBIUS contribution started October 2009.
  • The work on task 6.3.2 (ASTRON: Data inspection) has not yet started.
Red flags - major problems


Forward Look

  • 1.1.1 JIVE work on interoperability: :
  • 1.1.2 ESO ESO deliverables for interoperability and direction-dependent calibration tasks. The primary contribution to the interoperability task will be tasks to translate from FITS-IDI to and from the CASA measurement set format. This will be the mechanism for bulk data conversion for the use cases (EVN, VLBA, e-MERLIN) identified so far.
  • 1.1.2 JIVE on paralellization:
    • About to release AIPSlite into ParselTongue; this will greatly ease deployment of AIPS and ParselTongue on HPC clusters.
  • ASTRON continues work in selected workpackages
  • 2.2 ESO The deliverable for mosaic imaging is likely to be incorporation of CASA wide-field algorithms (specifically A-projection) in mosaic software. This is something which NRAO has identified as a need, and which will be required for ALMA, EVLA and e-MERLIN.
  • 2.2.3 UCAM Taking T0 = 1 April for these purposes, the proposed work programme for the next year is as follows:
    • Review of existing approaches and implementations for wide-field polarization calibration
      • report T0 + 1 month
    • Theoretical / practical analysis of the wide-field problem T0 + 3 months
      • Proposed implementation developed
      • present proposed implementation
      • agree proposed implementation T0 + 3.5 months
    • Example implementation in CASA (to be confirmed)
      • deliver code: T0+9 months
    • Test example implementation
      • consolidated report: T0 + 12 months
  • 3.1 MPI Main Activities and Work Plan for 2010. Complete all deliverables: Implement, test, document multi-rate filtering in DiFX Implement, test, document RFI Mitigation with Focal Plane Arrays Median filtering: pieFlag port to ParselTongue) Next 3 months: Implement, test, document multi-rate filtering in DiFX
Expected Milestones
  • Report on mosaic software (month 12)

Number of Persons working in JRA

  • JIVE JIVE continues to spend funded manpower on the expected rate, although matching effort is understaffed.
  • UCAM Joern Geisbuesch has been appointed on the RadioNET ALBiUS grant (11 months starting 1 April 2010) Matching effort: Paul Alexander, Gabriel Grigorescu; effort as required: Rosie Bolton, Dominic Ford.
  • UOXF Hans-Rainer Kloeckner has been appointed for 100% (October to December 2009) and 50% (from January 2010 onwards, Matching effort (at later stages of ALBIUS) will come from: Ben More (OeRC), Fred Dulwich (OeRC), Stef Salvini (OeRC)
  • MPI serious work to start in April 2010
  • ASTRON 1.36 man-months in 2010 Q1
  • NRAO focus has been on non-ALBiUS work
  • BORD 0.5 day, but hire made
  • ESO 230 person hours, efforts by Dirk Petry (DP) Robert Laing (RAL), Paola Andreani (PA), Hans Rykaczewski (HR)

Personal Tools