Beta Planning ------------- Schedule -------- April 15: Tester patch available for distribution May 21: ALMA CDR5 (Documents due May 1st) June 30: Patch 1 July 15: Moderate testing by NAUG, NAASC Aug 15: Patch 2; Initial freeze Aug 15; Thorough testing by NAUG, NAASC -Sep 07: bug fixing/polishing -Sep 21: distribution builds and testing/documentation Sept 30: Tentative beta release date Resources --------- Scientists: (3+1.5): JM: 0.75, SB: 0.75, GM: 0.75, KG: 0.75; RR(Canada): 1.0, TT(Japan): 0.5 SE: (3.75+1.5): WY: 0.75, DS: 1.0, BW: 1.0, DK(grant): 1.0, SJ(Canada): 1.0, HY(ALMA): 0.5 Visualization (David, Honglin (0.5)) Plotting (Urvashi (<0.5), Shannon) Calibration (George, Sanjay) Imaging (Kumar, Sanjay) Simulation (Kumar, George) Flagging (Urvashi, Wes) Image Analysis (Raymond) Single Dish (Tak) Note: Not needed for beta but must be in distribution for commissioning availability Interface (Joe+) System (Wes, Darrell, Boyd) ***** KEY [easy=1 week or less, med=1 month or less, hard=>1 month] C = indicates needed for commissioning ***** Beta Release ------------ ------------ Data to be filled and/or processed: ------------------------------------ Fill, calibrate, image & analyze (at least minimally): - All VLA 7mm to 21cm continuum and spectral line taken from 1995 (or earlier?) to present (except for wide-field imaging experiments). - This includes single field, polarized emission, and mosaics as well as multiple configuration imaging. - Wide-field imaging is not a high priority for ALMA so the beta release should not promise to image VLA data that requires this. - ALMA science data model (ASDM) format data General Development goals: -------------------------- Focus on making the software more stable. Provide better user feedback for: - when an error occurs (e.g. clearly state the problem and, if possible, how to fix it). - in log messages during nominal operations (e.g. assurances that the software is doing the right thing). Make task parameters consistent across the package. The user should *only* need tasks to process the above datasets, no tools. - Note, task parameters should be consistent with the tools also. Improve ability to find and flag bad data and examine calibration results. Make sure CASA is easy to install (test extensively before the beta release). DEVELOPMENT: Coordinated activities --------------------------------- All med C MS Selection All hard C Error handling All med C table locking/table still in cache All med C Crashes (see below) All hard C Documentation - Cookbook (NAUG; tools in advanced sections?) - URM (CASA) - in-line help (task, parameter and tool; NAUG/CASA) - script repository author + editor/reviewer for each module/task All easy - progress meter (for calibrater, imager, viewer) All easy - unify task parameter names All hard - Logging/tool messages Sys med - asynchronous tasks Sys easy - 64-bit support Sys med - Simplify task robustness/development (developer/astronomer) Functionality Development ------------------------- Image Analysis easy - expand imhead (really enable setting all header values) - split put/list functionality (incl. axes names)? - documentation needs examples for basic manipulation of header quantities easy - expand? regridimage (make this a task) easy - imagemaths task (easy scaling, etc) med - fitting components in the image plane Imaging med - interactive clean (per channel) implementation easy - masks for cleaning easy - a few missing important parameters in task clean med - imaging big cubes (FTMachine enable by plane; > 1024x1024x256 too big) med - mosaic performance easy - more flexibility for imaging RR, LL, Q, U, V only (med - wfproj imaging) Flagging/Plotting med - flag versions support (flagxy, flagger, ms) hard C flag tool (with update flagdata task); includes unflag data/flag versions easy C extend scope of channel averaged/time averaged flags med - improvements in plotting speed, memory usage (change in backend?) Calibration easy - solution snr threshold and reporting of value easy - Combined solution for specified IF's easy - post-solve bandpass normalization med - calplot updates (e.g., examine spline solutions) med C Cal Table revision (alignment with CalDM) easy - cal table lister (uses data selection) easy - visibility lister (uses data selection) med - frequency scaling of calibrater models easy - Application of fluxscale solution into ms model column med - Fast application of point source model med - Making ms corrected, model, weight columns with input of data easy - One time application of gaincurve and opacity med - D-term calibration easy - Application of polarization angle correction Tablebrowser med C stability Display easy - support for other color maps easy - support for beam ellipses easy - multi-image tracking easy - color bars easy - black/white background selection (at least) easy - multi-tracking ? - transparency for flags med - polygon support easy - profile/stats for integrated flux region med - plot spectra from different locations easy - export as mpg med - annotations med - improve print output (postscript) ------ Notes: ------ Commissioning - mainly displays, averaging - TP vs time, elevation, hour angle - visibility vs freq, time or any time-related parameter (elevation, air mass) - visibility vs freq-averaged or time averaged parameter - need total power calibration support in SD Simulation - make additional task (single field) - stress FITS image ingestion (utilize puthead routines to aid) - add data corruption facilities -------------------- Other issues - simulation (1 TB initiative) to demonstrate/stress the system EF G192 data size (MS data size) EF some imaging benchmarks Crashes ------- - imager (infinite loop when dirty beam is poorly defined) - imager (corrupted history table) - split task - flagdata task - flagxy when no data to plot - plotxy when no MS ----------------------------------------------------------- msselect parameter becomes set of parameters: ==> field, spwid, baseline, time, correlation, freqrange, uvrange *** ==> field, spwid, antenna1, antenna2, time, correlation, channel-freq-velocity *** range, uvrange) POSSIBLE ADDITIONAL TASKS - EF listvis mscorr (corrections to visibility data): example antenna position change, source position change, phase change (to remove jumps), gain factor, ionosphere correction, atmosphere correction) pbcor convolveimage ------------ From Debra ------------ Priority 2 --------------------------------------------------------- --------------------------------------------------------- Data to be filled and/or processed: ------------------------------------ Fill, calibrate, image & analyze (at least minimally): - PdBI data in correct ASDM format, 1 & 3mm spectral line taken within the last 5 years. Maybe earlier if the correlator structure has not changed. Add no functionality beyond what is currently in the code. ==> It will be PdB's responsibility to ensure their data is converted properly to ASDM format. Image & analyze: - Feather reduced, single dish images and synthesis images, starting from CASA or FITS format (includes images from any telescope or array that can be put in image FITS format). CASA should not be too "fussy" about the FITS format. - Semi-joint deconvolution of single dish images and synthesis data in which the single dish image is used as an initial model to deconvolve the synthesis data (includes single dish images from any telescope that can be put in the correct format and any calibrated synthesis uv data that in in MS or UVFITS format). General Development goals (in addition to above): ------------------------------------------------- Make sure CASA does not require root (super-user) priviledges to install. Priority 3 --------------------------------------------------------- --------------------------------------------------------- Data to be filled and/or processed: ------------------------------------ Fill from uvfits: yes. Calibrate and image: yes if no additional functionality is needed beyond what is available (e.g., Tsys correction for SMA data, correlator-specific resampling): - CARMA or SMA data in standard uvfits format (e.g. it can be filled into AIPS). General Development goals (in addition to above): ------------------------------------------------- Convert all tool-based functionality into tasks