Page 2 of 2

Re: Recent Miriad changes

Posted: Tue Mar 18, 2014 1:28 pm
by Mark.Wieringa
An update with the latest Miriad changes (since Sep 2013):
  • Invert's theoretical noise estimates could be way too high for large datasets, due to a recently discovered bug where precision was lost in accumulating weights. This has now been fixed. The estimates are still not as good as they used to be before CABB, due to more limited Tsys info.
  • gpcopy is now more relaxed about copying tables, avoiding creation of many small intervals in 'merge' or 'apply' mode.
  • gplist can now list the contents of binned leakage and gains tables.
  • linmos now has a cutoff parameter to specify how much of the primary beam to use and an 'alpha' option to mosaic mfs images with an I*alpha plane which can then be used to create spectral index mosaics. Mfspin has had a corresponding change to deal with these mosaics.
  • uvfmeas, the flux measuring task has had a number of improvements, including fitting multiple frequency ranges, more output options and handling larger datasets
  • a number of smaller bug fixes and improvements

---
Here is the full changelog

atlod.for
fix rpfits version reporting
apply vjm fixes to string overflows

atmos.for
Add drive settling time
update slew rates

cgdisp.for
Fix circle overlay for non lat/long axes

gpcal.for
Check nchan>=nfbin

gpcopy.for
use relax option for all modes
fix solution interval
change the way the relax option works

gplist.for
List details of gainsf/leakagef table

imload.for
fix docs - 256 char limit

invert.for
Fix bug in theoretical rms for large datasets
add ncp option

linmos.for
fix cutoff option
fix docs
add cutoff keyword
add alpha option

mfspin.for
cope with restored images

mospsf.for
fix hdelete call

pbplot.for
add options alpha and frequency

pgflag.for
Add option to output screen to a third plot device with D.

selfcal.for
Initialize nchan in point source case

uvdflag.for
Initial revision

uvfmeas.for
replace len_trim with len1
Show warning about too many uv points when buffer is exceeded while
uvhist option is given, but continue on while not accumulating and don't
Fix bug with the way the uvdistance bins were calculated so the bins cover
the full range of measured distances.
Make it possible to write out the binned uv-dist vs residual amplitude to the log file.
Made it possible to do spectral fits for large datasets by not calculating
uv-distance quantities when options=uvhist is not specified.
Fix a bug in the calculation of average user fit error.
Added option to output machine-readable fit coefficients, an option
to make an mfcal compatible flux string for further calibration based
Clean up unused variables and subroutines, and unused code blocks.
Improve documentation of the task.
Made it possible to plot multiple frequency ranges on the same plot and
do a simultaneous fit.

uvgen.for
do bandpass even if antenna gains are constant
Fix bug in bandpass code

uvplt.for
add amplitude unit (Jy)

uvsfit.for
Don't read/accumulate data if not fitting

uvspec.for
fix bug in adding units
fix axis label
add amplitude unit (Jy)
Add noise and mnoise option

Re: Recent Miriad changes

Posted: Fri Jan 23, 2015 2:12 pm
by Mark.Wieringa
Miriad changes since the last summary (Mar 2014):

arrayplt: a new task to plot the array layout from the antenna table

atlod.for: updates to the rfiflag files, better handling of antenna tables

fits.for: some changes to handle ASKAP data, topo option for uv export to CASA, cope with HPX (HealPix) map rotation

gpcal.for: Allow higher order spectral models and check for flagged bins

linmos.for: improve the clipping of large mosaics

mfboot.for: allow higher order spectral models, fix to avoid creating inconsistent gain tables

mfcal.for: allow higher order spectral models, fix so it can still work with options=delay when the model flux is specified.

pgflag.for: make less verbose in nodisp mode and fix some bugs

restor.for: improve beam fitting

selfcal.for: fix interpolation of binned solutions

uvedit.for: Correctly handle w coordinate.

uvfmeas.for: Numerous fixes and improvements including new goodness of fit measures

uvplanet.for: Added pparams option

uvrms.for: cope with >2^30 visibilities

Re: Recent Miriad changes

Posted: Tue May 31, 2016 12:33 pm
by Mark.Wieringa
Miriad changes since Jan 2015:
  • A new mm flux scale for 1934-638 was introduced starting May 2016. The tasks gpcal, mfcal were updated and the 'oldflux' option can be used to switch back to the previous mm flux scale.

    The task gpedit can do bandpass smoothing and a new 'extend' option can fill in missing channels in the bandpass.

    The tasks gpscal, selfcal and uvmodel take spectral parameters for the point source flux.

    Various tasks print more significant digits to allow capture of the output for subsequent processing.

    The task imstat has an option 'rrms' to plot 'robust' statistics (median, absolute deviation)

    The invert task takes a position angle for the taper to allow greater control over the shape of the synthesized beam.

    Task uvlin has a 'ratio' mode to produce a normalised residual spectrum.

    Various tasks were updated to accept longer filenames to cope with systems with long path names.

    Various bug fixed were made to: atlod, cgcurs, gpcal, imstat, linmos, uvsplit.