Open in Colab: https://colab.research.google.com/github/casangi/casadocs/blob/v4.7-v6.1/docs/notebooks/introduction.ipynb
Release Information
These are the release notes for CASA 4.7 - CASA 6.1. CASA Docs documentation for these old CASA versions has been deprecated. However, available on this page are downloadable tar-files with the version-specific XML information regarding tasks and parameters. This XML information is also available inside the version-specific CASA packages, and can be viewed on the terminal by typing the command help(taskname). The version-specific CASA packages can be downloaded on the CASA website, but require old versions of the Linux and Mac Operating System to run.
Release Notes 5.7/6.1
CASA 5.7 (Python 2) and CASA 6.0 (Python 3) contain the same functionality.
CASA 5.7.2/6.1.2
Release Notes 5.7.2/6.1.2
CASA 6.1.2 includes a pipeline that has been validated for the VLA. CASA 6.1.2 and included pipeline are functionally equivalent to CASA 6.1.1, and have the following added features:
A script for porting the tec_maps for ionospheric calibration of VLA data has been included in version 6.1.2. To follow new server conventions at NASA’s Crustal Dynamics Data Information System (CDDIS), the TEC date retrieval mechanism within the tec_maps script has been updated. See the Ionospheric Corrections section of CASA Docs for details.
In the task polcal: (a) poltypes ‘Df’ and ‘Df+QU’ have been restored to operability (they were accidentally disabled by the introduction of correlation-dependent flagging in gain-like calibration in 5.7.0/6.1.0); (b) In poltype=’Xfparang+QU’, the solution process is now less vulnerable to fully-flagged scans defeating the solution.
WARNING: The simalma bug-fix implemented in CASA 6.1.1 is not available in CASA 6.1.2 or 5.7.2. Please use CASA 6.1.1 for running task simalma.
A CASA 5.7.2 version that includes all the 6.1.2 and 6.1.1 updates is also available, but does not include any pipeline.
Task parameters 5.7.2/6.1.2
CASA 6.1.1
Release Notes 6.1.1
CASA 6.1.1 includes a pipeline that has been validated for ALMA. CASA 6.1.1 and its included pipeline is functionally equivalent to CASA 6.1.0, and has the following added features:
The model for Mars from January 1, 2021 to December 31, 2030 was corrected. This update is not included in the CASA 6.1.2 patch! Users of CASA 6.1.2 and CASA 6.1.0 (and earlier versions) should update their local data repository.
The CASA 6.1.1 tarballs will have an up-to-date data repository. I will thus not show warnings that were present in CASA 6.1.0 to update the repository manually with python3 -m casadata –update
The task simalma was updated to produce the expected output. A bug was fixed where the task simalma failed to return properly in due to import dependency problems. In addition, a problem calculating the image size for the default imsize=’[]’ was fixed, and an extraneous primary beam scaling before combining interferometric and single dish data has also been removed. Please use CASA 6.1.1 (not 6.1.0/5.7.0 or 6.1.2/5.7.2) for running simalma!
In the task flagcmd, the plot style has been modified to prevent apparently missing flags when the resolution of the plot resolution (in pixels) is insufficient. Note that the plots will now tend to represent a higher density of flags than there really is.
Casa and Casa Viewer launcher scripts now pass font file paths to PGPLOT on macOS. This fixed a bug that axes labels did not show up in the CASA Viewer for Mac OSX.
A performance improvement has been implemented in the plotms task.
There will be no corresponding CASA 5.7.1 release.
Task parameters 6.1.1
CASA 5.7.0/6.1.0
Highlights
CASA 6.1 is available as both a full installation (the tarball, as has historically been provided) and as modular Python packages, and has been updated from Python 2.7 to Python 3.6. CASA 5.7 is scientifically equivalent and remains Python 2.7 compliant.
A new task sdintimaging is available for joint deconvolution of Single Dish and Interferometer data.- A new task sdtimeaverage is available for averaging single-dish spectral data over specified time.
A new single-dish spectral imaging mode, ‘cubesource’, is available in the task tsdimaging.
A new parameter corrdepflags has been added to the gaincal, bandpass, and fringefit tasks to permit more control of flagging subsets of correlations.
The fringefit task now includes support for dispersive delays.
The CASA simulator now uses tclean instead of clean.
The ability to correct for heterogeneous antenna pointing offsets stored in the POINTING sub-table of the MS has been added to tclean (gridder=’awproject’).
statwt now includes weighting each visibility by its exposure time, and also improved in the way the timebin parameter is interpreted when its value is an integer.
the imsmooth task has been made consistency with the rest of CASA in terms of masking
simobserve now reads and populates antenna names rather than assigning numbers, which makes comparing simulated and real data easier.
The fldmap parameter within the cal library will now support multiple fields
CARTA v.1.4 is available for users to visualize CASA data outside the CASA Viewer. CARTA contains a growing amount of functionality available in the Viewer, and includes several features that are superior to the CASA Viewer.
VLA P-band models have been made available in CASA for several standard calibrators.
Ephemeris tables for Solar System objects have been extended in time
An “execfile” python shortcut has been added to the (Python 3 based) CASA 6 environment for backwards compatibility with ALMA scriptForPI.py restore scripts.
For continuum imaging (specmode=’mfs’), the .workdirectory, where all temporary images produced by the parallel runs were placed, has been removed.
A large number of bugs have been fixed.
Release Notes 5.7.0/6.1.0
Data Import/Export
In importasdm, the ‘remove_ref_undef’ parameter has been removed. This parameter was added after CASA 4.3 so that MeasurementSets created by importasdm after CASA 4.3 could be made compatible with CASA 4.3 (and older versions of CASA). This was intended to be used for a limited time to allow for backwards compatibility of MeasurementSets with version of CASA older than CASA 4.3. That temporary parameter has now been removed.
The SUB_TYPE column in the PROCESSOR table of a MeasurementSet is now filled by importasdm using the correlatorName appropriate for that PROCESSOR as found in the ASDM CorrelatorMode table when that PROCESSOR is a correlator. If the processor is not a correlator, or an appropriate row cannot be found in the CorrelatorMode table, the processorSubType is used to fill the SUB_TYPE column (SUB_TYPE is unchanged for non-correlator processors).
Calibration
The statwt application, which consists of both the task statwt and tool method ms.statwt(), has undergone substantial changes. The algorithm used to compute weights now includes weighting each visibility by its exposure time. There are also significant changes to how the application interprets the timebin parameter if it is an integer. For details, please see the complete statwt documentation
A new parameter corrdepflags has been added to the gaincal, bandpass, and fringefit tasks to permit control of behavior when a subset of correlations is flagged. Previously (and now only by the default: corrdepflags=False), such data was ignored, as if all correlations were flagged. Now, if corrdepflags=True, solutions will be obtained for the unflagged polarization in such cases. This enables recovery of data for antennas which may have only one available polarization (common in VLBI).
VLA P-band models for the frequency range between 230 - 470 MHz have been made available in CASA for several standard calibrators, following Perley & Butler (2017). These models can be used in the setjy task.
The fldmap parameter within the cal library will now support multiple fields specified as a selection string, much as the gainfield works in the conventional calibration apply parameters (previously, only selection of a single field was supported).
The runtime of flagdata is approximately 24-34% shorter than in earlier CASA versions, because the summary mode has been optimized.
Imaging
A new task sdintimaging for the joint reconstruction of single dish and interferometer data is being released in CASA 6.1/5.7. The sdintimaging task implements the algorithm published in Rau et al, 2019. It is being released as ‘experimental’ and is currently intended for the exploration of the algorithm’s capabilities and to gather feedback for the further refinement of the interface. It supports joint deconvolution for spectral cube as well as multi-term wideband imaging, operates on single pointings as well as joint mosaics, includes corrections for frequency dependent primary beams, and optionally allows the deconvolution of only SD images in both cube and mtmfs modes. Documentation about the algorithm, usage modes that have been verified and validated prior to release, and some example images, can be found here and in the sdintimaging task pages. For a list of known issues with this initial implementation as well as planned future development/tests, please refer to the sdintimaging developer page.
The new CARTA (Cube Analysis and Rendering Tool for Astronomy) visualization software version 1.3, with limited features, is now available for users to try.
CARTA v1.3, with limited features, is now available for users to try. CARTA is an alternative data visualization option for some functionality currently available using the CASA Viewer. CARTA is a new image visualization and analysis tool designed for ALMA, VLA, and SKA pathfinders, developed by ASIAA, IDIA, NRAO and Univ. Alberta. CARTA is eventually expected to replace the CASA Viewer, but the current version 1.3 offers only limited functionality and does not include all the features that the Viewer has. Nevertheless, CARTA v.1.3 may be the preferred tool for users who are worried about proper display of header information or overall performance of the CASA Viewer, or who wish to try remote display options, Stokes wedges, custimized layouts, or visualizing HDF5 image formats with CARTA.
The ability to correct for heterogeneous antenna pointing offsets has been added to tclean (gridder=’awproject’) and augments the existing ability to apply time-dependent pointing corrections. With a choice of usepointing=True, pointing corrections are applied via vectors that are fetched from the POINTING sub-table in the MS. A new parameter, pointingoffsetsigdev, is used to define bins for grouping antennas and deciding the degree of time variability to consider when computing the phase gradients (for usepointing=True). Demonstrations of this feature on simulated as well as VLASS data are documented in our VLASS Pointing Correction Report. CASA acknowledges the ARDG’s role in providing the C++ implementation as an addition to the existing AWProject gridder.
The convolution function used by the wproject gridder in tclean has been adjusted to ensure that it is calculated on a grid which size is a composite number. Dramatic slowdowns incurred by setting specific imsize settings (leading to grids of prime numbers) are now avoided.
For deconvolver=‘MTMFS’, the runtime of the minor cycle has been improved, particularly for large imsize, niter, and number of scales for multi-scale deconvolution.
For continuum imaging with specmode=’mfs’, the creation of a directory called .workdirectory, where all temporary images produced by the parallel runs were placed, has been removed. Only the final gathered/concatenated continuum images now appear in the main directory.
Analysis
visstat now handles more gracefully commands where all data are flagged within one of the groups (sub-selections) across the reporting axes.
A change in the imsmooth task has been made for better consistency with the rest of CASA, such that combining an existing pixel mask with an OTF mask results in an output mask that is only True for pixels where both input masks are True. A practical application of this change is an apparent “bug fix” for the immoments task when used on cubes with per plane beams; before the fix, it was possible for emission outside of the input mask to end up in the output moment maps.
Single Dish
Averaging SD spectral data over specified time is now available with the new sdtimeaverage task. Dedicated tasks for the baseline-subtraction and pol-averaging are already supported. This sdtimeaverage task is a further dedicated time-averaging task for more convenience.
A new spectral imaging mode, ‘cubesource’, is available in the task tsdimaging. The ‘cubesource’ mode is effective only for moving sources and tracks frequency shift during the observation. With ‘cubesource’ mode, frequency reference frame of the output image will be REST.
In tsdimaging, generation of psf image is disabled until correct implementation for single dish psf image is implemented. This is because the psf image currently generated by tsdimaging is useless in terms of single dish data reduction. This makes performance of tsdimaging faster than previous CASA 5 versions. Having said that, tsdimaging is still slower than sdimaging so that it might be better to use sdimaging for non-ephemeris imaging if performance matters.
VLBI
The fringefit task now includes support for dispersive delays (inversely proportional to the square of frequency), which parameters solved for is controlled by the new paramactive keyword. The paramactive parameter takes a Python list of Boolean arguments for the delay, rate, and dispersive components. The default settings preserve the previous behavior of the task, which is also expected to be the most common future use case.
Simulations
simobserve now reads and populates antenna names rather than assigning numbers, which makes it easier to compare plots of simulated data with those of real data. The population of antennas names is based on the fifth column of antenna configuration files passed via the antennalist parameter. Partial line comments in the header and body of such files are now also supported via the octothorpe (#) character.
Other
The task accum will be deprecated in the next CASA release, version 5.8/6.2.
An “execfile” python shortcut has been added to the (Python 3 based) CASA 6 environment for backwards compatibility with ALMA scriptForPI.py restore scripts. The “execfile” command has been tested and found to work in the same way as in (Python 2 based) CASA 5 with the exception that the treatment of global variables has changed in Python 3. For “execfile” calls within a script which itself is run via “execfile”, it is necessary to add globals() as the second argument to those “execfile” calls in order for the nested script to know about the global variables of the calling script. For example, within a script ‘mainscript.py’, calls to another script ‘myscript.py’ should be written as follows: execfile(‘myscript.py’, globals()).
Solar System objects’ positional ephemeris tables have been extended in time (Sun, Moon, Mercury, Venus, Mars, Jupiter, Saturn, Uranus, Neptune, Pluto, Io, Europa, Ganymede, Callisto, Titan and various asteroids). In addition, Mars and four asteroids (Ceres, Pallas, Vesta, and Lutetia) have explicit light curves based upon detailed thermo-physical modeling; these light curves have also been extended forward in time. The underlying brightness temperature models of these objects, and their relation to the Bulter-JPL-Horizons 2012 flux scale, have not been altered.
Bug-fixes
For plotms, a large number of bugs were addressed to improve display options and functionality, and to transition the plotms task from CASA 5 to CASA 6.
A bug was fixed that prevented uvtaper from working with Briggs style weighting (briggs, briggsabs, uniform) and perchanweightdensity=True.
A bug in tclean which incorrectly summed the imaging weights when using uniform weighting in parallel runs has been fixed.
A bug in tclean has been fixed that lead to a seg fault when imaging multiple MSes with inconsistent WEIGHT_SPECTRUM columns (i.e., the column exists in all MSes but in some has no data).
A bug in tclean has been fixed, which triggered a sementation fault when chanchunking with savemodel=’virtual’ and tclean ended with only 1 minor cycle.
A bug was fixed for imaging of ephemeris objects in tclean, which affected targets for which the field IDs observed are not in time order. This caused a noticeable offset in direction of the source in cases when the source was fast moving.
Several bugs that relate to the use of serial and parallel imaging runs in tclean have been fixed. They include a ‘latpole’ coordinate mismatch when attempting to use an output image from a parallel continuum run in a subsequent serial step, missing miscinfo image header information information in mtmfs parallel continuum runs, and the inability to use a model image from a parallel run in a subsequent serial predict-model step.
A bug in concat and importfitsidi that affected polarization visibility data in concatenated MSs, such that the cross-hands may previously have been spuriously mis-ordered on some baselines in the concatenated MS, has been fixed. The bug could be triggered when the second MS (in time-order) provided to concat had a different antenna table than the first MS. For more information, please see this CASA Knowledgebase Article. Specifically, ALMA Stokes V and VLA Stokes U, when subject to the condition above, would be affected by this bug.
A bug was fixed in the spwmap parameter, which could previously cause an exception when a spw pointed to by other spws did not point to itself. E.g., if only spw 1 has calibration available in the caltable and is to be used to calibrate spws 0,2,3, spwmap=[1,2,1,1] would fail (even if not selecting spw 1 for processing) since spw 1 points to spw 2 (for which there is no available calibration). Most users would not see this problem, since spws for which calibration is available will generally be calibrated by themselves. This is now fixed, and any permutation of spw specifications should work correctly in spwmap.
A bug has been fixed in gencal, specifically in the generation of the antenna position tables for VLA data (this bug did not affect ALMA data). Previously, if the highest-numbered antenna involved in a given observation required a position correction, it was not being registered correctly.
A segfault in imstat when using the ( - ) operator in region file has been fixed.
A bug was fixed where in certain specific circumstances (calibration absent for antenna id=0), Single Dish calibration tables were previously not being applied.
A bug in fixvis was fixed, whereby fixvis was misbehaving when processing Multi-MSs in parallel. Updates to uvcontsub and cvel were also made to ensure accurate internal copying of MMS directory structures.
A bug in flagdata, which led to occasional reports of progress values over 100%, has been fixed.
A bug was fixed in ImageBeamSet::setBeam() in which, while the beam was being replaced in the _beams Matrix object, _maxBeam was not being recalculated. This led to a problem when, in addition, the beam that was being replaced had been the current maximum beam.
In msmd.scansforfield(), the parameter name changed from intent to field, fixing an existing bug in the naming convention.
In sdimaging and tsdimaging, a bug was fixed that could cause an inaccurate direction conversion when more than one MS is given and these data contain the same antenna with the same antenna ID but with different location, i.e., station or pad where the antenna is placed. See this Knowledgebase Article for details.
In the position switch calibration mode of sdcal, a bug resulting in incorrect calibration tables has been fixed. The previous implementation, based on time gap detection, failed in some cases and OFF_SOURCE sequences were incorrectly combined. OFF_SOURCE data is now averaged over rows grouped by STATE_ID which corresponds to SUBSCAN number in the ASDM.
The wrong brightness unit of the weight image created by sdimaging has been fixed. The unit is now empty (unitless).
The simanalyze task previously corrected output images for the primary beam even when pbcor=False. This has been fixed in CASA 5.7/6.1.
Infrastructure
CASA 6.1 is available as both a full installation (the tarball, as has historically been provided) and as modular Python packages. The CASA 5 series is Python 2.7 compliant, and the CASA 6 series has been updated to Python 3.6. The configuration and startup options for CASA 5.7/6.1 are described largely in the following CASAdocs pages: Obtaining and Installing, Configuration, and Running User Scripts.
Startup options removed from CASA 6.1 which are still available in CASA 5.7: –telemetry (removed in favor of –notelemetry in CASA 6), –trace, and –maclogger. Startup options which are new in CASA 6.1: –norc, –notelemetry, –datapath, –user-site. See the Configuration page for more details.
Task parameters 5.7.0/6.1.0
Release Notes 6.0
CASA 6.0
The initial release of CASA 6 is available for general use. CASA 6 has been reorganized to offer a modular approach, where users have the flexibility to build CASA tools and tasks in their Python environment.
This first 6.0 release is targeted at developers and power users who build CASA into their own software applications and pipelines. This release includes the casatools, casatasks, and casampi modules, allowing for core data processing capabilities in parallel. The CASA 6.0 release is scientifically equivalent to the tools, tasks, and mpi capabilities available in CASA 5.6.1; the modules passed the same verification and unit testing required from CASA 5.6, and no new features are available in this first release of CASA 6.0 relative to CASA 5.6.1.
The initial CASA 6.0 Release is not intended for general users doing manual data reduction. It lacks much of the custom interactive shortcuts present in CASA 5 and many of the graphical user interfaces. Development continues on the data visualization packages, including casaplotms, casaviewer, and CARTA, as these involve external applications executing outside the python environment.
An upcoming 6.1 release scheduled for the spring of 2020 will contain a complete replacement for all CASA 5 components, including a single tar file distribution for those wishing to continue to install CASA in that manner. This monolithic package will be built from the same modular pip wheels, but adds a CASA-shell component to replicate the appearance of previous CASA versions.
Task parameters 6.0
[No XLM available]
Release Notes 5.6
This is the documentation for CASA 5.6. Changes compared to the CASA 5.5 release are listed below.
The CASA 5.6 series includes three official CASA releases to date, CASA 5.6.0, 5.6.1, and 5.6.2. All three 5.6 versions are suited for manual data processing. CASA 5.6.1 has also been scientifically validated for ALMA and includes the ALMA Cycle 7 pipeline, while CASA 5.6.2 includes the latest version of the pipeline that has been scientifically validated for the VLA.
CASA 5.6.2
Release Notes 5.6.2
CASA 5.6.2 and its included pipeline is functionally equivalent to CASA 5.6.1, containing only a pipeline bug fix for the VLA.
WARNING: all VLA users (11 Feb 2020): A bug was found in the version of the pipeline that was packaged with CASA 5.6.2-2 (the initial CASA 5.6.2 version for VLA). This bug impacts the restoration of calibrated MSs, and affects all VLA projects that restored pipeline calibration after October 21, 2019. If an existing calibration was restored using the new interface to the NRAO archive or following the instructions on the pipeline web-pages, you will obtain incorrect results. The output from a full calibration run of the pipeline is not affected by this bug, nor is data restored prior to October 21 2019, or data restored with a different version of CASA. As per Feb 19, a new CASA 5.6.2-3 version is available in which the problem has been fixed. The pipeline team will update affected users.
Task parameters 5.6.2
CASA 5.6.1
Release Notes 5.6.1
CASA 5.6.1 includes all updates of CASA 5.6.0. In addition, the following bugs that were present in CASA 5.6.0 have been fixed in CASA 5.6.1.
WARNING: CASA versions prior to 5.6.1-8 cannot be used to import raw ALMA Cycle 7 data due to a change in the data format. CASA 5.6.1-8 and the bundled Cycle 7 ALMA pipeline has been validated to calibrate and image raw ALMA data from Cycles 3-7. Restoration of previously calibrated ALMA data is recommended to be done in the CASA version originally used for the calibration, however, we believe that CASA 5.6.1-8 will restore pipeline calibrated data from CASA 5.4.0 or more recent.
Bug-fixes
In tclean, a bug was fixed that prevented uvtaper from working with Briggs style weighting (briggs, briggsabs, uniform) and perchanweightdensity=True.
In tclean, a bug was fixed in which the convolution function support size may change for different chunks. This bug appeared when chanchunk > 1 and the fractional bandwidth is large enough for the primary beams to be significantly different between chunks.
A bug in tclean has been fixed which was triggered when a chanchunk is created with only one channel, causing a change in interpolation and discontinuity in sensitivity.
In tclean, a bug that caused errors in the image center for high proper motion ephemeris targets has been fixed.
When running the polfromgain task with the caltable parameter, a bug prevented the removal of the newly created caltable using rmtables(). The new caltable is now cleared from the table cache to avoid this problem.
In gaincal, a user-specified spwmap list that is too long (cf the number of spws in the MS) is now trapped and will result in an error message. Also, delay tables in the multi-band case which are applied using the cal library will now use the correct pivot frequency per spectral window, matching the non-cal library case.
In plotms, symbols in the plots were updated to improve legibility and maintain consistency between RH6 and RH7.
Task parameters 5.6.1
CASA 5.6.0
Highlights
Task bandpass now supports relative frequency-dependent interpolation when applying bandpass tables.
In tclean, a new version of the parameter ‘smallscalebias’ is implemented for deconvolver = ‘mtmfs’ and ‘multiscale’ to more efficiently clean signal on different spatial scales.
A new suite of verification tests were written to evaluate various imaging modes related to mosaicing and wideband/widefield imaging in tclean.
The ‘mosaic’ and ‘awproject’ gridders in tclean can now use the ‘usepointing’ parameter to use the POINTING subtable of the MS during gridding convolution.
In tclean, the auto-multithresh now functions with polarization data.
A bug was fixed in tclean where explicitly setting the restoringbeam parameters produced a wrong flux scaling of the residuals.
For fringefit, a new keyword niter determines a maximum number of iterations for the global least squares solver.
plotms can show atmospheric and sky temperature curves for the image sideband, enhanced pointing plots and more accurate axis labels.
A new task nrobeamaverage is available Nobeyama data, to average on-source data over different beams.
tsdimaging has new modes to track ephemeris objects.
In sdcal, the ‘otf’ calibration mode has improved.
simobserve can create multi-channel MeasurementSets from a component list. ALMA receiver temperatures were updated.
The statwt task log has been updated to provide explicitly the channels used in the statwt calculations.
The ATM library of atmospheric models has been updated in CASA.
A large number of memory leaks have been addressed in the code.
Release Notes 5.6.0
Data Import/Export
In importasdm, CalAntennaSolutions tables with the new spectralWindowId field required by the VLA can be filled to a table using the asis option. Versions of this table produced by ALMA without that field are still supported.
MS Viewing, Editing, Manipulation
A new feature was added to plotms to show the atmospheric curve for the image sideband (in black) along with the atmospheric transmission curve or the sky temperature curve. This feature is enabled with a new showimage parameter (True/False, default False), or a checkbox on the Axes tab in the GUI. The image sideband curves can be shown only if (1) showatm or showtsky is also selected, (2) the MS (or the MS associated with a calibration table) is available and has not been split resulting in reindexed spectral window IDs, and (3) this MS has an ASDM_RECEIVER table.
Pointing plots in plotms have the following enhancements:
B1950 and GALACTIC reference frames are now supported as xframe and yframe parameters.
The value ‘cubic’ in the xinterp and yinterp parameters is no longer allowed and has been replaced with ‘cubic spline’.
Axes labels and values are formatted and sorted in accordance with the reference frame.
The ia.modify() tool now handles ellipses/disks correctly. Further, pixels with centers inside the disk will have the same values, even if a pixel straddles the edge of the disk. Pixels which straddle the edge of the disk which have centers outside the disk are given values of zero. Thus, one should not expect the flux density of the disk to be exactly the provided value to the component list; for a given size disk, the computed flux density will be closer to the expected value for images with smaller pixels.
Calibration
In task bandpass, relative frequency-dependent interpolation is now supported when applying bandpass tables, e.g., to permit use of offset bandpass calibration (e.g., for HI and similar cases where line features may contaminate an ordinary bandpass calibration scan).
In the experimental fringefit task, a keyword niter has been added to set a maximum number of iterations for the global least squares solver. The default is 100; in cases with high signal-to-noise this limit is not reached.
In the fringefit task, the calibration tables were extended to anticipate the integration of code for correcting dispersive atmospheric delays in an upcoming CASA version. This means that calibration tables created with fringefit in CASA 5.5 cannot be used in CASA 5.6 (see Known Issues for details).
The fluxscale task has been updated so that the output fluxtable is no longer able to overwrite the input caltable. The user can still specifically set append=True to append the current flux-scaled results to the existing table.
The fluxscale task now populates the fitFluxd and fitRefFreq keys in the returned dictionary if an SDM contains only a single spectral window. No fitting is performed, but fitFluxd and fitRefFreq are filled with the values from fluxd and freq, respectively.
Imaging
The capabilities of clean are now in tclean. The clean task will be deprecated in the near future! Please use tclean instead. All major functionality from clean is present in tclean via a modified interface along with additional algorithmic options.
In tclean, a suite of 30+ functional verification tests have been written to evaluate various imaging modes relating to joint mosaicing and wideband imaging. Data were simulated for the VLA L-band, with one 1 Jy point source with spectral index of -0.5, located in-between two overlapping pointings of a mosaic, and with three channels spanning 1-2 GHz. Tests include single-field as well as two-field (joint mosaic) imaging modes, spectral cube and multi-term wideband imaging, gridders ‘mosaic’, ‘awproject’ and ‘standard’ along with appropriate post-deconvolution pb-correction steps in order to achieve flat-sky fluxes and spectral indices, with all tests operating in serial as well as in parallel modes. Several Known Issues are still in place for widefield and wideband imaging in tclean, so users should continue to use these algorithms at their own discretion.
In tclean, the gridders ‘mosaic’ and ‘awproject’ can now make use of the ‘usepointing’ parameter to use entries supplied in the POINTING subtable of the MS rather than the FIELD table for applying a phase gradient during gridding convolution. This ensures that the image-domain primary beam is formed from antenna voltage patterns centered at the true pointing direction per antenna. Please read the chapter on widefield imaging for details on how both gridders deal with antennas pointed in different directions.
In tclean, the parameter ‘smallscalebias’ has been added to the option deconvolver = ‘mtmfs’ to more efficiently clean signal that shows structure at different spatial scales. See CASA Memo 9 for details.
In tclean, the Multiscale CLEAN algorithm (deconvolver = ‘multiscale’) has been improved. For a flat spectral model (nterms=1), deconvolver = ‘multiscale’ and deconvolver = ‘mtmfs’ now produce deconvolved images that are similar. See CASA Memo 9 for details.
In tclean, the default setting for perchanweightdensity, which enables the calculation of the weight density for each channel in a cube independently, has been changed to True.
Auto-multithresh now functions with polarization data. It applies the same algorithms to the Stokes QUV images as used for the Stokes I image. This means that the full masking process is applied to the positive emission (including the prune and grow steps), but that the masking of the negative emission only includes the initial threshold mask (no prune or grow).
Analysis
The statwt task log has been updated to provide explicitly the channels used in the statwt calculations.
The ATM library of atmospheric models has been updated in CASA to the latest version, which includes an improved ozone profile, line-specific pressure broadening factors, line specific temperature dependency coefficients, and other fixes.
Single Dish
A new task nrobeamaverage is available. The task is for ‘ON-ON’ observation datasets from Nobeyama 45m telescope, to average on-source data over different beams.
tsdimaging has new modes of phasecenter to track moving objects. One is phasecenter=’TRACKFIELD’ to use ephemeris table attached to MS, and the other is to set ephemeris table name explicitly to phasecenter.
In sdcal, the ‘otf’ calibration mode has improved, as higher order interpolation has been implemented to get pointing direction for each spectral data.
Simulations
The simobserve task is now able to create a multi-channel MeasurementSet from a component list using a new subparameter of the componentlist parameter called comp_nchan.
The ALMA receiver temperatures assumed by the simobserve task have been updated to the values assumed in the Cycle 7 Technical Handbook. The simalma task now also uses these values. Please see the simobserve task page for specific Trx values for each band.
Other
ia.toascii()/fromascii() have been deprecated and will be removed in an upcoming release.
Bug-fixes (selected)
A large number of memory leaks have been addressed in the code.
A bug has been fixed where explicitly setting the restoringbeam parameters (including ‘common’ beam) in tclean produced a wrong flux scaling of the residuals. See this memo for details.
In plotms, several bugs have been resolved that deal with correct scaling and labeling of the axes.
Running tclean or plotms on a dataset with a mixed polarization setup (some windows exclude cross-hand terms) where only the LL and RR correlations have been split out, is now working correctly.
A bug in specsmooth for the boxcar smoothing option has been fixed, such that the frequency labeling in the header is now correct.
In gaincal, multiband delay solving (gaintype=’K’, with combine=’spw’) can now properly handle heterogeneous spws (e.g., spws with different bandwidths, channelizations, etc.).
In importvla, a number of bugs that affected the import of archival VLA data have been addressed.
In tclean, the interpolation (when set to ‘linear’ and ‘cubic’) of data prior to gridding was changed to fix a problem that was found when chanchunk >1 or in parallel. The problem was that at the edge of chunks (or parallel cube partition) the psf will exhibit jumps in size. The interpolation alignment with the data channel was changed to fix this issue.
tclean runs now without errors in parallel cube mode when the number of channels is lower than the number of parallel servers.
imregrid can now regrid an image with multiple channels to another cube which overlaps by 1 channel.
A bug in applycal which resulted in single dish caltables being left open when using the cal library has been fixed.
An error caused by leaving table lock on the internal mask when tclean was restarted with savemodel=‘modelcolumn’ has been fixed.
Task parameters 5.6.0
Release Notes 5.5
CASA 5.5.0
This is the documentation for CASA 5.5.0. CASA 5.5.0 is intended for general use. Changes compared to the CASA 5.4 release are listed below.
Highlights
Several new tasks have been introduced: apparentsens, polfromgain, and sdpolaverage.
The tasks fringefit, gaincal, polcal, immath, and plotms have a number new capabilities, as described below.
In tclean, new weighting options are added, and improved noise calculation was implemented for “auto-multithresh”.
rerefant now supports rereferencing solutions produced by the fringefit task.
importfitsidi now imports flagging information and correctly concatenates WEATHER tables.
statwt has the excludechangs parameter re-introduced, and can now also use data selection by scan.
The simulator has been updated to include the spectral variation of each component in a componentlist.
The CASA Docs documentation has a new look and layout
Release Notes 5.5.0
Documentation
CASA Docs now has a more modern look and layout, with easier navigation
CASA Docs has a section added that collects CASA memos and knowledgebase
Inline help and parameter description on the CASA Docs task pages are being improved, to provide a more complete documentation on task-parameters. This is work in progress that will continue for CASA 5.6
Data Import/Export
importfitsidi now imports flagging information.
importfitsidi now correctly concatenates WEATHER tables if multiple input files are given.
MS Viewing, Editing, Manipulation
In plotms, a new interactive flagging mode has been added, “Flag all/Unflag all”. Contrary to the existing mode that applies flagging operation to selected regions within plot panels, the new mode offers per-panel flagging operation to the user.
plotms now has the capability to plot antennas pointing directions. Refer to Ant-Ra / Ant-Dec axes, and to the Antenna Pointing Direction Parameters section in plotms documentation. Averaging is currently not supported for these axes. This feature has only been implemented for ALMA, ASTE, and NRO data.
In plotms, a feature has been added to allow connections between plotted visibilities in calibration tables (MeasurementSets are not yet supported). Connected points will have the same metadata (including flag) with only the x-axis value changing. Points will be colorized based on their connection metadata. No connectors are drawn between flagged and unflagged points, even when flagged points are not displayed.
Calibration
The task fringefit has several new capabilities:
fringefit now allows specification of multiple reference antennas and automatically chooses an alternative reference antenna if no data for the initial reference antenna is available.
fringfit now allows specification of search windows for delay and rate.
fringefit now has the option to skip the globalization step.
fringefit uses a new convergence criterion that matches classic AIPS more closely. This speeds up the task in most cases as it requires fewer iterations in solving.
rerefant now supports rereferencing solutions produced by the fringefit task.
The task polcal has two new parameters, Xfparang+QU and PosAng. These are considered experimental in CASA 5.5.
Xfparang+QU uses parallactic angle coverage on a significantly linearly polarized calibrator to estimate the cross-hand phase spectrum prior to instrumental polarization calibration (for either basis). It solves for the apparent cross-hand phase spectrum (channelized) and integrated (per-spw) fractional linear polarization, Q and U. Xfparang+QU is part of “standard” polarization calibration for ALMA (linear feed basis), and is under study as an option for the VLA (circular basis).
PosAng solves directly for the absolute position angle offset spectrum, for either the linear or circular basis. PosAng is redundant with traditional cross-hand phase spectrum calibration (poltype=’Xf’) for the circular feed basis (but permits a semantic distinction in heuristics); for the linear basis, an external calibration reference more accurate than the presumed feed registration is required (e.g., a polarized beacon such as ALMA is constructing).
A new task polfromgain is added, which can estimate calibrator linear polarization from gain solutions (which assumed zero linear polarization in the model) in the linear basis. This is part of the “standard” ALMA polarization calibration. Task polfromgain is considered experimental in CASA 5.5.
A new task apparentsens is added, which will estimate nominal imaging sensitivity from the visibility weights as a function of data selection and imaging weight parameters. Also reports relative sensitivity compared to pure natural weighting. Supports only continuum (specmode=’mfs’) imaging at this time.
A new single-dish task sdpolaverage is added to export data averaged over different polarizations.
In task gaincal, the new parameters solmode and rmsthresh have been added to support robust solving modes: L1 and/or on-the-fly outlier rejection while solving. This feature is experimental in CASA 5.5.
statwt can now use data selection by scan (under selectdata).
In statwt, the excludechans parameter has been re-implemented in statwt. It now takes a Boolean setting such that, when True, the fitspw selection is inverted and the specified channels are excluded from the computation of the weights.
Imaging
The capabilities of clean are now in tclean. The clean task will be deprecated in the near future! Please use tclean instead. All major functionality from clean is present in tclean via a modified interface along with additional algorithmic options.
A new visibility weighting option, weighting=’briggsabs’, is now available in tclean. The advantage of this weighting option is that it is much more aggressive with respect to changes in npixels, the uv-cell size. The Briggsabs option is experimental in CASA 5.5 - please use at own discretion!
A new parameter perchanweightdensity has been implemented in tclean, which when set to perchanweightdensity=True provides a per channel way to calculate Briggs weight density for image cubes. This results in a more uniform sensitivity per channel for cubes, but with generally larger psfs than the default setting of perchanweightdensity=False (prior behavior).
In tclean, the ability is added to do full AW-Projection corrections, using frequency and time dependent primary beams, suitable for VLA widefield and wideband imaging. The use AW-Projection is experimental in CASA 5.5 - please use at own discretion! The corrections are azimuthally asymmetric to account for feed leg structures and sidelobes of the primary beam. They also include beam squint, which is corrected during gridding as a function of time, band, and frequency within a given band. The frequency dependence of the primary beam within the data being imaged is included in the calculations and can optionally also be corrected for during gridding. In addition, w-corrections to account for sky curvature and non-coplanar baselines are now possible. The AW-Projection algorithm is currently undergoing commissioning for the VLA Sky Survey pipeline and is being tested for a variety of other use cases. We aim to officially endorse it for general use in CASA 5.6.
In tclean, a new improved noise calculation was implemented for “auto-multithresh” automask and n-sigma threshold stopping. It allows more robust noise estimate in the presence of extended source structures and/or absorption. Set a newly introduced parameter, fastnoise, to False to active the new noise calculation. When the default, fastnoise=True, is chosen a simple median absolute deviation (MAD) based noise calculation will be used as in the previous releases. In addition, the sidelobe threshold is now being correctly chosen over the noise threshold for the cases where the peak residual is negative and both automask and n-sigma threshold stopping are used together.
The ia.decimate tool now ensures that the beam information is flushed to disk even if the image tool is not closed properly.
Analysis
immath now employs two new modes, lpoli and tpoli, for linear and total polarization images. The poli mode, which was previously available (and for backwards compatibility has not been changed), is flexible and will fall back to tpoli when Stokes V is provided and lpoli when Stokes V is not available.
A new parameter prec has been added to the task immath and tool ia.imagecalc(), which will be used to explicitly indicate what the precision of the output image should be, float or double.
Single Dish
a new SD task sdpolaverage is added to export data averaged over different polarizations.
Other
The simulator has been updated to include the spectral variation of each component in a componentlist. For a component with a flat spectrum, the predicted visibilities now show the frequency variation of the beam.
CASA 5.5 introduces a great improvement in memory use for large mosaics when using mosweight=True in tclean. As downside of the memory improvement, mosaics with large number of fields and MSs may suffer from an “too manby open files” error if the limit for open files is set too low (see Known Issues for details).
Bug-fixes (selected)
A bug in tclean when using deconvolver=’multiscale’, such that the model column was being written incorrectly when tclean finished by threshold (the behavior was fine when finishing by niter), has been fixed.
A bug in tclean was fixed where parallel processing with parallel=True failed when mosweight=True
Task parameters 5.5.0
Release Notes 5.4
CASA 5.4.1/5.4.2
Release Notes 5.4.1/5.4.2
CASA 5.4.1 and 5.4.2 are intended for general use, and CASA 5.4.2 is also the official VLA pipeline release. For outside users, CASA versions 5.4.1 and 5.4.2 are identical. The official ALMA pipeline release is CASA 5.4.0. The CASA 5.4 releases build on CASA 5.3. Changes compared to the CASA 5.3 release are listed below. The differences between 5.4.1/5.4.2 and 5.4.0 are labeled “5.4.1/5.4.2 only” on this page.
IMPORTANT: NEW VERSION CASA 5.4.1-32
A patch to CASA 5.4.1 with the version number CASA 5.4.1-32 is officially released by CASA and approved by the VLA, and can be downloaded from the CASA download page. The 5.4.1-32 version solved a critical bug for large mosaic cubes, where the signal-to-noise in the first pointing of a field could be noticeably lower than in the rest of the cube. This solution is also implemented in subsequent CASA versions (e.g. CASA 5.4.2 and up).
Problem: The problem occurred when the chanchunk parameter in tclean is put to a value greater than the default value of 1. Note that for chanchunks = -1 (which is often used as default for ALMA), chanchunks > 1 is triggered whenever the estimated memory requirement is > 80% of the system memory. The value of chanchunks used to create a cube is printed in the casa .log file, shortly after tclean starts. The chanchunk feature is useful for large images cubes, where the gridders can run into memory limits as they loop over all available image planes for each row of data accessed. To prevent this, the chanchunk parameter can be used to grid subsets of channels in sequence, so that at any given time only part of the image cube needs to be loaded into memory. The bug found for values of chanchunk >1 affects primarily mosaic images, where each “source” is made up of more than one “field”, but technically also affects single field observations, where each “source” is made up of one “field”. The specifics of the bug in tclean are:
gridder=’mosaic’: the first integration of the first field of each source included in the mosaic is dropped.
gridder=’standard’: the first integration of each source (i.e. each field in this case) included in the image is dropped (though typically the ‘standard’ gridder is used when there is only one field).
The loss in signal-to-noise in the affected pointing scales with sqrt(#integrations -1)/ #integrations, so gets less problematic when a pointing is visited with a large number of integrations. For example, if a field has > 5 integrations throughout the data selected for the image, the loss in rms noise will be < 10%. So only images with few integrations per field are adversely affected - typically these will be large mosaics and very short snapshot observations.
Solution: We recommend VLA users to use the official 5.4.2-5 version, but at least the CASA 5.4.1-32 version or up, in particular for large image cube mosaics.
ALMA users are recommended to use CASA 5.4.0-70, which also has this bug-fix for chanchunk included. Compared to CASA 5.4.1-32, the version CASA 5.4.0-70 excludes several features that were implemented for the VLA and not yet testing by ALMA (these features are not relevant for ALMA Cycle-6 production).
All new features and bug fixes throughout the various 5.4 release versions will be included and tested for both ALMA and VLA in the next CASA 5.5 release. To the best of our knowledge, this chanchunk bug was introduced in CASA 5.0 and does not affect single dish imaging tasks (tsdimaging)
General
During testing for the CASA 5.4 releases, several issues involving mosaicing in clean and/or tclean for CASA versions 4.7 - 5.1 were uncovered, as explained in more detail below. These issues have been fixed in CASA 5.4.
A number of tasks have been deprecated in CASA 5.4: autoclean, boxit, clearplot, csvclean, importevla (use importasdm instead), mosaic, oldhanningsmooth, oldpartition, oldplotants, plotuv (capabilities transferred to plotms), plotxy (capabilities transferred to plotms), ssoflux, widefield, visstatold and visstat2.
The deprecated task importevla is now replaced by importasdm, but several additional step in parameters specification and flagging of zero and shadowed data are required to replicate the functionality of importevla, as described in the CASA Docs task pages of importasdm.
The task statwt2 is now statwt (the previous startwt is now oldstatwt)
5.4.1/5.4.2 only: statwt has changed its syntax for choosing which channels to take into account regarding the weights calculation. While in CASA 5.4.0, the parameter excludechannels was introduced to exclude channels from the weights calculation, CASA 5.4.1 uses fitspw to specify channels to be included in the weights calculation.
Starting CASA 5.4.1, telemetry data is by default generated for each CASA run and sent anonymously to NRAO to measure statistics on CASA task useage. This feature can be disabled by setting the CASA_ENABLE_TELEMETRY environment variable to “false”.
Documentation
We continue to update CASA Docs, but are still in a transitioning phase of making CASA Docs the main source of information on CASA. All task descriptions are now in CASA Docs, and we will gradually phase out the inline help descriptions. We are slowly rolling out additional documentation on remaining areas of CASA as we are able to test and verify the functionality matches the description.
Data Import/Export
Ephemeris data for the Moon, based on JPL-Horizons with a cadence of 1hour, has been added as a standard DE200 ephemeris table.
importnro now imports Nobeyama data with data stored in UTC, rather than JST (UTC+9).
MS Viewing, Editing, Manipulation
flagmanager now returns a dictionary for mode=’list’, returning the flag version names and comments.
Calibration
5.4.1/5.4.2 only: gaincal now has the option to use the median values for normalizing amplitude solutions
plotms inherited the functionality of plotcal and now supports all calibration tables. plotms now also offers plotting scalar residuals. In addition, plotms supports simple text export, which just exports X and Y values, in addition to the default verbose export.
5.4.1/5.4.2 only: plotms now also the capability to connect points for calibration tables (support for MeasurementSets will be added later).
Imaging
During testing for the CASA 5.4.0 release, several unknown issues were uncovered that affected mosaic imaging in clean and/or tclean for CASA versions 4.7 - 5.1. The problems have been investigated and quantified for ALMA by the North American ALMA Science Center (NAASC) Software Support team, and can be found in this memo. In short, these are the problems uncovered:
ALMA 7m-array mosaics imaged with CASA versions up to and including 4.7.2 (Cycle 0-4): The ALMA 7m-array mosaic imaging used a non-optimal definition for the antenna primary beam in both clean and tclean up to CASA 4.7.2. For Nyquist-sampled mosaics, after primary beam correction, within the 50% mosaic primary beam response, the result is flux densities that are too high by ~10%. At the 20% mosaic primary beam response level, the flux densities are too large by ~30%.
All mosaics imaged with gridder=’mosaic’ in tclean in CASA 5.0 and 5.1 (this includes all ALMA mosaics made with Cycle-5 pipeline version 5.1.1): In CASA 5.0 and 5.1, a truncated version of the PB model was erroneously used for gridding the data in tclean. This truncation causes aliasing (at some level) in all CASA mosaic images made with the mosaic gridder. The manifestation of the problem in the image plane depends sensitively on a range of factors including: mosaic shape, mosaic spatial sampling, and uniformity of mosaic sensitivity per pointing. In extreme cases errors as large as 15% have been observed inside the 50% mosaic primary beam response, but the median error for most mosaics is a few percent inside the 50% mosaic primary beam response. In particular for rectangular mosaics, the situation rapidly deteriorates beyond the 50% mosaic primary beam response, with flux density errors as large as 30% seen at the 20-50% primary beam response for mosaics with size-ratio in the actual sky-coverage pattern of 1.5 or larger.
The latter issue affects images made with the mosaic gridder for all telescopes (including the VLA) at some level, although the effect of the errors is expected to be smaller for larger antennas. VLA observers who created image mosaics using the mosaic gridder in previous CASA 5 versions, especially in the case of non-square image sizes, are encouraged to re-image using the latest CASA version.
The reported problems only affected data imaged with the gridder=’mosaic’ mode. These known issues have been fixed in CASA 5.4.0.
Automasking has been further improved with refined performance.
tclean now fully supports mosaics of ephemeris objects spanning one or more execution blocks through the new parameter phasecenter. For ALMA data, phasecenter=’TRACKFIELD’ uses the ephemeric tables attached to each MS. For the latest ephemeric information, or for non-ALMA data, one may set phasecenter to the name of the ephemeris file, or a body for which a DE200 ephemeris table is distributed with CASA (‘MERCURY’, ‘VENUS’, ‘MARS’, ‘JUPITER’, ‘SATURN’, ‘URANUS’, ‘NEPTUNE’, ‘PLUTO’, ‘SUN’, ‘MOON’)
For mosaic imaging with Briggs/uniform weighting in tclean, the mosweight sub-parameter of gridder=’mosaic’ has been added to tclean with default setting mosweight = True (compared to the old mosweight = False default in clean). There is the current Known Issue that mosweight=True is now silently invoked also for other multi-field imaging. The default mosweight = True optimizes noise characteristics for ALMA mosaics. However, it may potentially cause memory issues for large VLA mosaics, and may cause the major and minor axis of the synthesized beam to be up to ~10% larger than with mosweight=False. Please change to mosweight=False to get around these issues.
Analysis
To calculate robust statistics in imstat, an algorith = ‘biweight’ option has been added to determine the center and width of a distribution while downweighting points that are far off, e.g. due to the presence of significant, wide spread signal.
Most image analysis tasks in CASA can now handle complex-double and double precision images.
Bug-fixes (selected)
In imregrid, the conversion between galactic and J2000 coordinates that previously caused a wrong tilt to high declination targets has been fixed.
In simobserve and the underlying tool ia.modify, point-sources from a component-list were assigned the wrong WCS coordinates in CASA 5.3, because the x and y pixel positions were swapped. This issue has been fixed for CASA 5.4, and only affected CASA version 5.3.
A bug was fixed where multi-EB ephemeris data sets produced a wrong image with tclean parameter settings deconolver=’mtmfs’ and nterms=2
In rflag, the parameters ‘timedevscale’ and ‘freqdevscale’ now operate in the ‘apply’ mode, whereas previously the values were written to the flag file during the ‘calculate’ mode only, and any settings during the ‘apply’ stage were silently ignored.
In split, specifying a spw selection with multiple channel ranges within a single spw (e.g., spw=’1:26;1012’) now creates the correct DATA_DESCRIPTION subtables
In sdimaging, a bug was fixed that caused wrong computation of pointing directions when multiple fast-scan MS files are given as input.
In importnro, several bugs that causes data inconsistency in output MS have been fixed
5.4.1/5.4.2 only: In tclean, the default of conjbeams is set to False for all applicable gridders, including mosaic and awproject (was mistakenly set to conjbeams=True in CASA 5.3 and 5.4.0).
5.4.1/5.4.2 only: In statwt, the default of the parameter timebin has temporarily been set to 0.001s to work around a bug detailed in the CASA 5.4.1 Known Issues.
Task parameters 5.4.1/5.4.2
CASA 5.4.0
Release Notes 5.4.0
The CASA 5.4 release builds on CASA 5.3. Changes compared to the CASA 5.3 release are listed below. The CASA 5.4.0 release includes usage of the ALMA pipeline (the VLA pipeline will be accommodated in a subsequent CASA 5.4.1 release).
IMPORTANT: NEW VERSION CASA 5.4.0-70
A patch to CASA 5.4.0 with the version number CASA 5.4.0-70 is officially released by CASA and approved by ALMA, and can be downloaded from the CASA download page. The 5.4.0-70 version solved a critical bug for large ALMA mosaic cubes, where the signal-to-noise in the first pointing (typically a corner) could be noticeably lower than in the rest of the cube.
Problem: The problem occurred when the chanchunk parameter in tclean is put to a value greater than the default value of 1. Note that the ALMA pipeline (and often manual imaging) uses chanchunks = -1, which means that chanchunks > 1 is triggered for any cube whenever the estimated memory requirement is > 80% of the system memory, even if this is not explicitly set by the user. The value of chanchunks used to create a cube is printed in the casa .log file, shortly after tclean starts. The chanchunk feature is useful for large images cubes, where the gridders can run into memory limits as they loop over all available image planes for each row of data accessed. To prevent this, the chanchunk parameter can be used to grid subsets of channels in sequence, so that at any given time only part of the image cube needs to be loaded into memory. The bug found for values of chanchunk >1 affects primarily mosaic images, where each “source” is made up of more than one “field”, but technically also affects single field observations, where each “source” is made up of one “field”. The specifics of the bug in tclean are:
gridder=’mosaic’: the first integration of the first field of each source included in the mosaic is dropped.
gridder=’standard’: the first integration of each source (i.e. each field in this case) included in the image is dropped (though typically the ‘standard’ gridder is used when there is only one field).
The loss in signal-to-noise in the affected pointing scales with sqrt(#integrations -1)/ #integrations, so gets less problematic when a pointing is visited with a large number of integrations. For example, if a field has > 5 integrations throughout the data selected for the image, the loss in rms noise will be < 10%. So only images with few integrations per field are adversely affected - typically these will be large mosaics and very short snapshot observations.
Solution: We recommend ALMA users to use the CASA 5.4.0-70 version, in particular for large image cube mosaics. ALMA users should be aware that the CASA 5.4.1 versions have not been validated for ALMA users.
VLA users are recommended to use CASA 5.4.1-32, which also has this bug-fix for chanchunks included. In addition, the original CASA 5.4.1 and patch 5.4.1-32 include several other features that are explicitly described in the CASA 5.4.1 Release Notes (those new 5.4.1 features are not yet tested by ALMA and not relevant for ALMA Cycle-6 production).
All new features and bug fixes throughout the various 5.4 release versions will be included and tested for both ALMA and VLA in the next CASA 5.5 release. To the best of our knowledge, this chanchunk bug was introduced in CASA 5.0 and does not affect single dish imaging tasks (tsdimaging)
General
During testing for the CASA 5.4 releases, several issues involving mosaicing in clean and/or tclean for CASA versions 4.7 - 5.1 were uncovered, as explained in more detail below. These issues have been fixed in CASA 5.4.
A number of tasks have been deprecated in CASA 5.4: autoclean, boxit, clearplot, csvclean, importevla (use importasdm instead), mosaic, oldhanningsmooth, oldpartition, oldplotants, plotuv (capabilities transferred to plotms), plotxy (capabilities transferred to plotms), ssoflux, widefield, visstatold and visstat2.
The deprecated task importevla is now replaced by importasdm, but several additional step in parameters specification and flagging of zero and shadowed data are required to replicate the functionality of importevla, as described in the CASA Docs task pages of importasdm.
The task statwt2 is now statwt (the previous startwt is now oldstatwt)
statwt has changed its syntax for choosing which channels to take into account regarding the weights calculation. It now uses the parameter excludechannels to exclude channels from the weights calculation.
Starting CASA 5.4.0, telemetry data can be generated for each CASA run and sent anonymously to NRAO to measure statistics on CASA task useage. This feature can be enabled by setting the CASA_ENABLE_TELEMETRY environment variable to “True”. Starting CASA 5.4.1, telemetry will turned on by default, but for CASA 5.4.0 it need to be turned on manually.
Documentation
We continue to update CASA Docs, but are still in a transitioning phase of making CASA Docs the main source of information on CASA. All task descriptions are now in CASA Docs, and we will gradually phase out the inline help descriptions. We are slowly rolling out additional documentation on remaining areas of CASA as we are able to test and verify the functionality matches the description.
Data Import/Export
Ephemeris data for the Moon, based on JPL-Horizons with a cadence of 1hour, has been added as a standard DE200 ephemeris table.
importnro now imports Nobeyama data with data stored in UTC, rather than JST (UTC+9).
MS Viewing, Editing, Manipulation
flagmanager now returns a dictionary for mode=’list’, returning the flag version names and comments.
Calibration
plotms inherited the functionality of plotcal and now supports all calibration tables. plotms now also offers plotting scalar residuals. In addition, plotms supports simple text export, which just exports X and Y values, in addition to the default verbose export.
Imaging
During testing for the CASA 5.4.0 release, several unknown issues were uncovered that affected mosaic imaging in clean and/or tclean for CASA versions 4.7 - 5.1. The problems have been investigated and quantified for ALMA by the North American ALMA Science Center (NAASC) Software Support team, and can be found in this memo. In short, these are the problems uncovered:
ALMA 7m-array mosaics imaged with CASA versions up to and including 4.7.2 (Cycle 0-4): The ALMA 7m-array mosaic imaging used a non-optimal definition for the antenna primary beam in both clean and tclean up to CASA 4.7.2. For Nyquist-sampled mosaics, after primary beam correction, within the 50% mosaic primary beam response, the result is flux densities that are too high by ~10%. At the 20% mosaic primary beam response level, the flux densities are too large by ~30%.
All mosaics imaged with gridder=’mosaic’ in tclean in CASA 5.0 and 5.1 (this includes all ALMA mosaics made with Cycle-5 pipeline version 5.1.1): In CASA 5.0 and 5.1, a truncated version of the PB model was erroneously used for gridding the data in tclean. This truncation causes aliasing (at some level) in all CASA mosaic images made with the mosaic gridder. The manifestation of the problem in the image plane depends sensitively on a range of factors including: mosaic shape, mosaic spatial sampling, and uniformity of mosaic sensitivity per pointing. In extreme cases errors as large as 15% have been observed inside the 50% mosaic primary beam response, but the median error for most mosaics is a few percent inside the 50% mosaic primary beam response. In particular for rectangular mosaics, the situation rapidly deteriorates beyond the 50% mosaic primary beam response, with flux density errors as large as 30% seen at the 20-50% primary beam response for mosaics with size-ratio in the actual sky-coverage pattern of 1.5 or larger.
The latter issue affects images made with the mosaic gridder for all telescopes (including the VLA) at some level, although the effect of the errors is expected to be smaller for larger antennas. VLA observers who created image mosaics using the mosaic gridder in previous CASA 5 versions, especially in the case of non-square image sizes, are encouraged to re-image using the latest CASA version.
The reported problems only affected data imaged with the gridder=’mosaic’ mode. These known issues have been fixed in CASA 5.4.0.
Automasking has been finalized, with refined performance and improved noise calculations.
tclean now fully supports mosaics of ephemeris objects spanning one or more execution blocks through the new parameter phasecenter. For ALMA data, phasecenter=’TRACKFIELD’ uses the ephemeric tables attached to each MS. For the latest ephemeric information, or for non-ALMA data, one may set phasecenter to the name of the ephemeris file, or a body for which a DE200 ephemeris table is distributed with CASA (‘MERCURY’, ‘VENUS’, ‘MARS’, ‘JUPITER’, ‘SATURN’, ‘URANUS’, ‘NEPTUNE’, ‘PLUTO’, ‘SUN’, ‘MOON’)
For mosaic imaging with Briggs/uniform weighting in tclean, the mosweight sub-parameter of gridder=’mosaic’ has been added to tclean with default setting mosweight = True (compared to the old mosweight = False default in clean). There is the current Known Issue that mosweight=True is now silently invoked also for other multi-field imaging. The default mosweight = True optimizes noise characteristics for ALMA mosaics. However, it may potentially cause memory issues for large VLA mosaics, and may cause the major and minor axis of the synthesized beam to be up to ~10% larger than with mosweight=False. Please change to mosweight=False to get around these issues.
Analysis
To calculate robust statistics in imstat, an algorith = ‘biweight’ option has been added to determine the center and width of a distribution while downweighting points that are far off, e.g. due to the presence of significant, wide spread signal.
CASA can now handle Double and Complex-Double images.
Bug-fixes (selected)
In imregrid, the conversion between galactic and J2000 coordinates that previously caused a wrong tilt to high declination targets has been fixed
In simobserve and the underlying tool ia.modify, point-sources from a component-list were assigned the wrong WCS coordinates in CASA 5.3, because the x and y pixel positions were swapped. This issue has been fixed for CASA 5.4, and only affected CASA version 5.3.
A bug was fixed where multi-EB ephemeris data sets produced a wrong image with tclean parameter settings deconolver=’mtmfs’ and nterms=2
In rflag, the parameters ‘timedevscale’ and ‘freqdevscale’ now operate in the ‘apply’ mode, whereas previously the values were written to the flag file during the ‘calculate’ mode only, and any settings during the ‘apply’ stage were silently ignored.
In split, specifying a spw selection with multiple channel ranges within a single spw (e.g., spw=’1:26;1012’) now creates the correct DATA_DESCRIPTION subtables
In sdimaging, a bug was fixed that caused wrong computation of pointing directions when multiple fast-scan MS files are given as input.
In importnro, several bugs that causes data inconsistency in output MS have been fixed
Task parameters 5.4.0
Release Notes 5.3
CASA 5.3.0
Release Notes 5.3.0
The CASA 5.3 release builds on CASA 5.1. Changes compared to the CASA 5.1 release are listed below.
Documentation
We continue to update CASA Docs but are still in a transitioning phase, where in particular the description of the tools is not yet complete. Please use the old CASA Toolkit Manual if the tool description is not yet available on CASA Docs. All task descriptions are now in CASA Docs, and we will gradually phase out the inline help descriptions.
Data Import/Export
The tasks importfits now have a feature enabled that writes a history to the image it creates
The task importmiriad has a built in limit of 64 antennas; when more than 64 antennas are present, an exception is now thrown instead of crashing. There has also been a fix to the indexing so that importmiriad can now deal with data that contains ‘wide’ channels (spectral averages) for each spectrum.
importasdm now checks and eliminates duplicate data rows.
importfitsidi now (1) has the added parameter “specframe”, with default value “geo”, to indicate the spectral reference frame; (2) works with specifying full paths; (3) has the scanreindexgap_s parameter defined on a per-antenna basis, effectively enabling subarraying in the decision regarding when to start a new scan.
MS Viewing, Editing, Manipulation
The task plotants has been rewritten to remove unwanted dependencies and add requested features for plotting antenna positions. New parameters allow the user to:
display antenna ID in addition to name;
plot the logarithmic positions;
exclude antennas from the plot or plot only those in the MeasurementSet main table (e.g. after a split);
set the title for the plot.
The old plotants task has been renamed oldplotants and will be deprecated.
The task plotms has been given the following features:
ability to compute and overlay an atmospheric transmission curve or a sky temperature curve, as available in plotbandpass. The parameters showatm (default ‘False’) and showtsky (default ‘False’) have been added to plotms for this feature. It can also be enabled in the plotms GUI by selecting the desired Overlay on the Plot > Axes tab (default ‘None’).
ability to provide information about calibration tables by clicking on the Summary menu in the top menu bar. If All is chosen from the pull-down menu next to Type, listobs-style output about scans, correlator configurations, and antennae will be written to the console or log. Other options for a subset of the MeasurementSet include Where, What, How, Main, Tables, Antenna, Feed, Field, Observation, History, Polarization, Source, Spectral Window, Spectral Window and Polarization, SysCal, and Weather. For calibration tables, options in the Summary menu include All, Where, What, How, Main, Tables, Antenna, Field, Observation, History, and Spectral Window. See chapter page about the use of plotms.
ability to display a header above the plots. The contents of the header can be edited interactively from the GUI, or specified from the CLI using the new “headeritems” parameter.
all remaining functionality of plotcal has been implemented in plotms, pending the deprecation of plotcal. This includes the option to provide values in the parameter plotrange with numpy types, e.g., plotrange=[0,0,0,np.max([0.0,10.0])]. Only things not implemented in plotms compared to plotcal are support of baseline-based tables and connecting the dots in the plot.
A bug has been solved in plotms that affected delay tables: when an antenna has been completely flagged, plotms now produces a blank plot instead of skipping the plot for that antenna.
Calibration
The calibration library has been updated and supports the CASA calibration tables that are used in standard calibration. The calibration library is a file that can be used in applycal to calibrate all fields in a single execution, or in plotms to calibrate data on-the fly without an execution of applycal.
gaincal can now perform multi-band cross-hand delays
A new flux standard ‘Perley-Butler 2017’ is now available in setjy. It is also the new default.
The task plotbandpass now correctly handles data from the Nobeyama Radio Observatory, following an update of the getWeather function in the analysisUtils that are called by plotbandpass
The experimental tasks accor and fringefit have been added to CASA for the reduction of VLBI data. The task accor determines the amplitude corrections needed due to errors in sampler thresholds using measurements of auto-correlations spectra (typically required for data correlated with a DiFX correlator, such as VLBA, and not for data from a SFXC correlator like the EVN). The task fringefit determines phase offsets, group delays and delay rates with respect to a specified reference antenna by a two-dimensional fast Fourier-transform and subsequent last-squares optimization.
Imaging
phase centers in tclean can now be moving objects such as planets or ephemeris files
tclean has a new stopping criterion nsigma that end cleaning by given multiples of the MAD noise of the residual image
A tclean mask will now be extended to all Stokes parameters if it was only created for a single Stokes plane
A new stokes parameter ‘pseudoI’ has been added to allow imaging data with visibilities that are partially flagged in one polarization hand but not the other.
For widefield imaging (wprojectft) in tclean, a bug that under some cases caused increased peak residuals and higher noise towards the edges of the image has been fixed.
For very large images and number of wprojection planes there was a big spike in memory usage when running tclean. This has been addressed, and CASA will now try to respect amount of memory available for calculating the convolution function from the start of tclean
Performance improvements in the auoboxing ‘auto-multithresh’ algorithm; the older options (auto-thresh and auto-thresh2) will be deprecated in favor of the new algorithm
Improvements in memory usage for tclean single and parallel mode (including a memory resource predictor)
Analysis
statwt2 is a new task that will replace current statwt in CASA 5.4. In addition to the statwt functionality, statwt2 displays the mean and variance of the computed weights (also in preview mode), and prints those in the logger. Also, the datacolumn selection now supports ‘residual’ and ‘residual_data’ options, which will compute weights based on DATA - MODEL_DATA and CORRECTED_DATA - MODEL_DATA, respectively. statwt2 also has options to use robust statistic methods and has better control of flagging data. field, spw, and states can be combined for computation.
imstat has significantly improved its performance for many cases where the paramter algorithm is not equal to ‘classic’ and parameter axes is specified.
imfit had pixel coordinates (pixelperarcsec) added to each component subrecord, to facilitate using these values in scripts.
The tasks imrebin, imregrid, immoments, imfit, imcollapse, imcontsub, imhead, immath, and impv now have a feature enabled that writes a history to the image it creates
A bug in the task impv has been solved, in which impv failed in mode=’length’ for pa=’90deg’ or pa=’270deg’
The task immath can create an image of the total polarized intensity using the parameter mode=’poli’, which now calls the po.totpolint() tool. For more detailed information, see the task description pages of immath.
Single Dish
A new task sdsidebandsplit was implemented to separate single-dish data observed by DSB receivers
The sdbaseline and sdsmooth tasks have new parameter, reindex, to control whether or not indices in MS subtables are re-indexed based on data selection.
The task sdcal has a new interpolation option available. It can be used for Lissajous-scanned Single Dish images to suppress large gradient pattern along elevation direction observed when setting interp=’linear’. In addition, performance for sdcal has been improved, and a bug has been fixed so that sdcal no longer fails to generate a sky caltable if the input data is Multi-MS.
A new experimental task, tsdimaging, has been defined. It is supposed to be equivalent to sdimaging but is based on new imager framework.
Performance for total power (TP) calibration has improved, especially for Solar data (double-circle fast scan).
Tools
The “limb-darkened disk model” is not working in through the cl tool and any reference to this has been removed from the documentation
The utilstool.compare_version method has been changed to return a string instead of a list of numerals.
Many ms tool functions been refactored to use newer infrastructure code for selection and iteration, but have the same arguments and functionality. The exception is that all selection is now cumulative; conflicting selections will result in a null selection (a MeasurementSet with zero rows). For example, when selecting different data description IDs with ms.selectinit, you must reset the selection (ms.reset() or ms.selectinit(reset=True)) between the calls to selectinit.
Bug-fixes
mstransform now resizes the ANTENNA subtable when selecting on antennas, thus correctly handling MeasurementSets that include different antenna designs (e.g., 7m and 12m antennas in ALMA).
Other
When going from CASA 5.3 to CASA 5.4, there are a number of tasks that will be deprecated. The list of tasks that will still be in CASA 5.3, but no longer in CASA 5.4 is the following:
autoclean
boxit
clearplot
csvclean
importevla (functionality is now in importasdm)
mosaic
oldhanningsmooth (functionality is now in hanningsmooth)
oldplotants
plotuv (functionality is now in plotms)
plotxy (already deprecated in 5.3, functionality is captured in plotms)
ssoflux (functionality is now in setjy)
statwt2 (will revert back to statwt, as described earlier in these CASA 5.3 Release Notes)
visstatold / visstat2 (functionality is now in visstat)
Task parameters 5.3.0
Release Notes 5.1
CASA 5.1.2
Release Notes 5.1.2
The CASA 5.1.2 release builds on CASA 5.1 and is mostly intended as a patch for the VLA pipeline. Changes compared to the previous CASA 5.1.1 patch include:
reordering the INTENTs for VLA data, to accommodate older data that have TARGET intents for requantizer scans;
fix to task plotweather, to close all table calls properly when no WEATHER data is available in the MS;
enabled the support of Total Electron Content (TEC) corrections in the VLA pipeline tasks;
Pipeline task hif_makeimlist: for data sets where the central channel of a spw is flagged, the synthesized beam is now derived from a non-flagged channel nearest to the central flagged channel.
Task parameters 5.1.2
CASA 5.1.1
Release Notes 5.1.1
CASA 5.1.1. is a patch of the 5.1.0 release that contains bugfixes for the ALMA and VLA pipelines. In particular, the ALMA pipeline restore script was updated. For the VLA pipeline, we changed the interpolation method in applycal to circumvent a bug that flags additional edge channels. We also fixed a potential bug that may select the incorrect calibrator when the BANDPASS intent is not defined.
Task parameters 5.1.1
CASA 5.1.0
Release Notes 5.1.0
The CASA 5.1.0 release builds on 5.0.0 and contains new, updated ALMA and VLA pipelines. In addition, we added the following new features:
Documentation
We continue to update CASA Docs but are still in a transitioning phase for task and tool descriptions. Please use the inline help if the task or tool description is not yet available on CASA Docs
Data Import/Export
The matrix syntax in FITS files has been modernised and now follows standard 3.0. The way exportfits handles PC matrix convention is updated:
gridline rotation as seen in casaViewer is now consistent and correct in both casa image format and fits image format.
pc matrix convention “PCi_j” is now the convention and the casaViewer now shows results consistent with ds9.
MS Viewing, Editing, Manipulation
The CASA 5.0 task visstat has been renamed visstatold for CASA 5.1. The CASA 5.0 task visstat2 has been renamed visstat for CASA 5.1, although it is still available with the name visstat2.
cvel2 & mstransform: regridding (regridms=True) has been corrected to follow the same procedure as in tclean. mstransform does not any longer pre-average channels when the ratio between the output and input width is >2.
Calibration
Ionospheric Total Electrion Content (TEC) corrections (low frequency cm observing):
TEC “at site” plot now stored in file.
tec_maps.create0 now generates (non-interactive) plot file
Reference antenna selection:
The gaincal task now supports “strict” refant application via a new refantmode parameter, which will flag solutions timestamps (per spw) if the specified refant is unavailable (flagged or absent). Useful for polarimetry where refant cross-hand phase stability is important.
New ‘rerefant’ task which permits revising the refant on an existing caltable.
Flux scaling: in the CASA 5.0 “Known Issues”, a bug was reported for fluxscale, where it generated incorrect results if the flux density calibrator field, as specified in the reference parameter, was also specified among the other calibrators in the transfer parameter. This bug has been fixed, in that the reference calibrator that is also listed in the transfer parameter will now be excluded from the scaling process.
Imaging
Automasking:
The growiterations subparameter has been added for multithreshold automasking in tclean (usemask=’auto-multithresh’) to control the number of iterations used by binary dilation to expand the mask into low signal-to-noise regions. Lowering this number can help to improve the processing speed of tclean particularly when it involves a large cube, but may lead to less effective masking of extended emission. The default has been selected to optimize masking of extended emission. Some experimentation may be needed to determine the most effective lower value for your purposes.
An experimental feature has been added to auto-multithresh algorithm (usemask=’auto-multithresh’) to mask absorption, which is controlled by the sub-parameter negativethreshold. This value is a signal-to-noise based parameter similar to the noisethreshold and lownoisethreshold parameters in the auto-multithresh algorithm. Currently, the masks produced using this threshold are not independent of the emission threshold-based masks, so use of this parameter may lead to masking of low-lying positive emission near absorption. This feature assumes that the data have been continuum subtracted.
Single Dish
ATNF Spectral Analysis Package (ASAP) is removed from CASA. The sd tool and the following single dish tasks are removed, sdaverageold, sdbaselineold, sdbaseline2old, sdcalold, sdcal2old, sdcoaddold, sdfitold, sdflagmanagerold, sdflagold, sdgridold, sdlistold, sdmathold, sdplotold, sdreduceold, sdsaveold, sdscaleold, and sdstatold.
The sdimprocess task is renamed sdfixscan. Also, mode=’basketweave’ and ‘press’ are renamed ‘FFT_mask’ and ‘model’, respectively.
The sdimaging task now supports several types of projections. Currently, SIN (slant orthographic; default), CAR (plate carree), TAN (gnomonic), and SFL (Sanson-Flamsteed) are available.
The sdimaging task now supports overwriting brightnessunit of the CASA image. Currently, ‘K’ (Kelvin) and ‘Jy/beam’ are available.
The parameter verbose is available in sdbaseline task when blfunc=’variable’. When verbose=False (default), logging of per spectrum fitting parameters is supressed.
The sdgaincal task now allows to specify a list of caltables to be applied on-the-fly during computing calibration factor using the parameter applytable. You can specify spwmap and interpolation method (interp) for each caltables specified by applytable. Syntax for spwmap and interp is same as applycal task.
The sdcal and sdgaincal tasks support calibration of single polarization data.
Tools
The ms tool has new functions ending in “2” (for example, getdata2), which change the underlying implementation to a newer infrastructure but have the same arguments and functionality. The old functions will eventually be deprecated and replaced by the new ones.
There is a new msmd tool method timesforspw() to retrieve all time stamps associated with a specified spectral window
Other changes
The use of the ‘casadef’ module is deprecated. The CASA version information has been moved to the ‘Class utilities (cu)’ module. This module contains functions cu.version and cu.compare_version to retrieve the CASA version number and compare version numbers. The subversion information that is found in the ‘casadef’ module is no longer up to date. CASA now uses git for source code version control and no analogous git state is available. The ‘casadef’ module will be removed in CASA 5.2 or CASA 5.3 depending on the timing of these releases.
Task parameters 5.1.0
Release Notes 5.0
CASA 5.0.0
Release Notes 5.0.0
The development of CASA 5.0 focused on internal refactoring, a new build and workflow process, and code and performance improvements. This includes a new way to access, store and process visibilities that affects many tasks but should be invisible to the user.
New features include:
Documentation
We implemented a completely new documentation system, CASAdocs. CASAdocs replaces the Reference Manual and Cookbook and contains other pieces of documentation that were previously available in external documents. CASAdocs will also contain all current CASA inline help, but during a transition period only a limited number of task and tool descriptions are available in CASAdocs.
Data Import/Export
channel averaging in exportuvfits and ms.tofits() has been removed. We advise to use mstransform or split to bring the MS in the required shape before exporting to fits.
MS Viewing, Editing, Manipulation
In plotms it is now possible to choose multiple datacolumns for the same visibility axis (amp, phase, real, imag). This can be done for x- and y-axes, or via multiple y-axes.
The msmd tool has now additional methods to report the number of arrays (msmd.narrays(), useful for concatenated data) and spectral window names (msmd.spwsfornames())
new mode ‘antint’ in flagdata to flag integrations if all baselines to a specified antenna are flagged.
Calibration
In polcal we are deprecating poltype=’X’ (XMueller), use poltype=’Xf’ from now on
All calibration solving tasks are now using WEIGHT_SPECTRUM if available.
Imaging
tclean is maturing and replaces and improves on more and more functionality of current clean. This includes
experimental, new automasking algorithm designed to work on a wide range of emission morphologies available in tclean via usermask=’auto-multithresh’
an increasing peak residual is not a termination criterion anymore
new EVLA beams are now the default for VLA data.
a bug was fixed where model visibilities were interpolated incorrectly for data that had three conditions fulfilled: negative frequency widths, cubes that are decreasing in frequency, and interpolating over more than 2 channels.
Image Analysis
new task imdev and tool method ia.deviation are now available to create images that represent the rms noise of an input image
Simulation
added possibility to set outframe of simulated MS, and changed how simobserve treats the reference spectral pixel of a sky model image -
this is to allow more precise treatment of noise.
Heterogeneous array configurations e.g. ALMA 12m+7m are supported, as are new telescopes with control of the primary beam patterns using the vpmanager
An Airy pattern will be used for unknown telescope primary beams
Single Dish
The following single dish tasks are renamed (name in CASA 4.7 -> 5.0). Note all tasks with ‘old’ at the end of the name will be deleted in future releases.
tsdbaseline -> sdbaseline
tsdcal -> sdcal
tsdfit -> sdfit
tsdsmooth -> sdsmooth
sdaverage -> sdaverageold
sdbaseline -> sdbaselineold
sdbaseline2 -> sdbaseline2old
sdcal -> sdcalold
sdcal2 -> sdcal2old
sdcoadd -> sdcoaddold
sdfit -> sdfitold
sdflag -> sdflagold
sdflagmanager -> sdflagmanager
sdgrid -> sdgridold
sdlist -> sdlistold
sdmath -> sdmathold
sdplot -> sdplotold
sdreduce -> sdreduceold
sdsave -> sdsaveold
sdscale -> sdscaleold
sddstat -> sdstatold
time average and polarization average are available in Task, sdfit
Renaming of a parameter in sdimprocess: masklist -> maskwidth
Task importnro supports import of Nobeyama 45-m observations in the spectral window mode
The task importasdm in lazy mode now also accept FLOAT_DATA, which is used in single dish applications (ocorr_mode=”ao”).
Other changes
Global tool instances, im, tb, ms, at, etc., are no longer exported for use outside of the CASA command line interface (CLI) by default, e.g. in user created tasks. End users can still use these as they like, but they are not systematically exported for use in tasks. In general, outside of the CASA CLI new instances of the tools should be created where needed, e.g. myim = imtool.create( ). Some tools must be finalized; please consult the documentation for individual tools.
In an effort to improve CASA, a crash reporter has been introduced. This crash reporter collects information to help diagnose why CASA crashed. If you would like to prevent reporting information about a crash use the –nocrashreport CASA command line flag when starting CASA. The collected information includes:
program call stack
filesystem mount information
CASA log
memory information
operating system version
CPU information
The Boolean shortcuts T for True and F for False are not accepted anymore, please write out the full True and False Boolean values as they are defined in python.
iPython is now upgraded to version 5.1.0
The environment variable CASA_MAX_MEMORY can be defined to limit memory usage in plotms
Added a ‘–trace’ CASA startup command line option to check on imported modules
ATM library update to ATM-0_9_0 (ALMA-JAN2017)
Task parameters 5.0.0
Release Notes 4.7
CASA 4.7.2
Release Notes 4.7.2
CASA 4.7.2 is a second patch for 4.7.0 that fixes a CASA 4.7.1 bug concerning the restoration of mosaics in tclean. The primary beam correction was inadvertently applied to the model image, which causes problems during cleaning and causes the output restored image to have the primary beam corrected model added to uncorrected residuals, which is incorrect. This behavior is independent of the value of the pbcor parameter (for mosaics the pbcor=True option is not yet implemented). The issue only occurred in CASA 4.7.1.; 4.7.0 is not affected.
Task parameters 4.7.2
CASA 4.7.1
Release Notes 4.7.1
CASA 4.7.1 includes the most recent official VLA data calibration pipeline (including a correction for the online tropospheric delay bug, see below). It also contains the candidate ALMA Cycle 4 pipeline patch, although this has not yet been accepted for operations use.
Bug-fixes
Added a correction for the tropospheric delay bug that affects VLA data taken between August 9, 2016, and November 14, 2016. The fix is included in gencal when used with caltype=’antpos’.
A non-standard record of center versus reference frequencies in MTMFS images and FITS files has been resolved.
importasdm is not failing anymore in the case of mismatched and in particular missing BDFs (this bug was most visible for bdfflags).
Improved divergence checks in tclean based on peak residuals
tclean pblimit can now be set as a negative number to omit the application of a Boolean mask in images and residuals.
Task parameters 4.7.1
CASA 4.7.0
Release Notes 4.7.0
Major feature improvements over the previous version of CASA include:
Imaging
tclean is a new imaging task, eventually replacing clean. It has a new, more straightforward interface, allows more combinations of algorithms, include options for output files, has a more flexible outlier approach, and includes algorithms for autoboxing. tclean modes that have been commissioned include mfs with nterms=1&2 for both, the standard and mosaic gridders, and spectral line mode cube with the hogbom deconvolver in the LSRK frame. We also tested the common option of restoringbeam and as well as multi-scale clean. Commissioning of additional modes are currently underway, including support for parallel computing. If you find any problems with tclean please send us a note through the CASA helpdesk.
Data examination/editing/import/export
plotms for calibration tables is now capable to display tsys, spgain, tec, and SNR axes
polarization selection for calibration tables in plotms
plotms has a new option to average across fields
plotms displays the reference frame for frequency axes
plotms has now the ability to set font size for labels
height, width and dpi has been added to the plotms task export feature
plotms does now allow one to iterate over correlation
pre-averaging in flagdata
Calibration
setjy is now performing linear interpolation of model values
new asteroid flux models are now available in setjy
new standard “Stevens-Reynolds 2016” for an updated flux density polynomial of 1934-638
new cb.createcaltable tool method to create empty calibration table
Data and Image Manipulation
New task: plotprofilemap to show a “profile map” of CASA image data, as spectra binned up over user-defined boxsize, in RA/Dec.
msmd can retrieve spw ids from spw names
improved image processing histories are now recorded in CASA images
imsubimage has options to drop degenerate axes
imhead now also returns a python dictionary
immath allows specification of an image that is used to define the metadata of the output image
imsmooth scales images with units Jy/beam and Jy/beam.km/s via the inverse of their beams but leaves images with other units unscaled by default
Data analysis
new task specflux reports spectral line characteristics when averaged over a region
ia.findsources does now accept Jy/beam.km/s images
masks will be stretched individually when they are used in immath LEL expressions
improved ASCII file output format for imfit
Viewer
The precision of the spectral coordinate labeling can now be selected
contour thickness can be set in imview
Single Dish
New task: sdgaincal: Uses mode=’doublecircle’ and interpolates time-dependet gain variation for Fast-Mapped Double circle observations.
New task: importnro: Convert Nobeyama NOSTAR format to ASDM format.
New task/parameter in sdimaging: clipminmax; rejects the maximum and minimum values of the data ensemble used to compute pixel values during gridding.
Extended task: sdimaging accepts and rounds float values input to imsize and cellsize
Extended task: sdimaging selects target intents by default
New fitmode in tsdfit: fitmode=’auto’; auto-detect line emission and return gaussian fit characteristics
New function/parameter in sdbaseline: blfunc=’sinusoid’, enables removal of selected (or dominant) fourier components from spectral data.
New calmode in tsdcal: calmode=’otf’: Generation of sky calibration table assuming integrations at a map edge as off source measurements. Map edge detection is operated on the offset coordinate with respect to the source for known moving sources, e.g., Sun, Moon, etc.”
Performance
Parallelization of tclean is still underway
Other
new Jansky VLA beams are now available
Experimental tasks
tclean testing of some modes is still underway, this includes parallelization of the imaging processes
new task: visstat2 (MS-format replacement for sdstat and visstat)
mstransform has an experimental addition to perform continuum subtraction