paths containing spaces
psrchive configure gfortran
Edit: I just modified the feedstock and created a PR for it. See here: https://github.com/conda-forge/psrchive-feedstock/pull/68
Add aarch64 on conda-forge
Merge branch 'master' of ssh://git.code.sf.net/p/psrchive/code
minor formatting/spelling fixes
Information about handling PSRFITS SEARCH-mode data added
By default, print a warning message when loading PSRFITS SEARCH-mode data.
pam -R 0 is broken
commit e6bfb590a651d8ac108a92b40896d583f72363b4 Author: Willem van Straten vanstraten.willem@gmail.com Date: Tue Nov 25 06:36:57 2025 +1300 bugs/508 record the correction of relative DM/RM, even if it is zero
Merge branch 'master' of ssh://git.code.sf.net/p/psrchive/code
bugs/508 record the correction of relative DM/RM, even if it is zero
pam -R 0 is broken
Merge branch 'master' into Draft2025
optionally specify the FWHM of a von Mises component
Thanks for being so responsive. Not having to create workarounds makes things very easy. I tested the above on a clean container build, and it all works
Merge /u/vhaasteren/psrchive/ branch healpix into master
:-) No worries! (Sorry I didn't test more thoroughly before accepting !)
Include submodule of epsic
Include submodule of epsic
Merge /u/vhaasteren/psrchive/ branch healpix into master
Update healpix.m4 macro
Thanks, Rutger! Thanks, Colin!
Oh, and also here, thanks to Colin Clark for the help with finding this issue!
Update healpix.m4 macro
fix check for autoconf version
Add numpy2 support
Awesome! Thank you, Rutger!
Add numpy2 support
pam --update_DM uses DIST_DM1 parameter to update DM
I think that this bug is fixed in commit 1d3fce040d3682260a80575df0a572694f04e3e1 Author: Willem van Straten <vanstraten.willem@gmail.com> Date: Fri Oct 31 09:47:08 2025 +1300 bugs/506 new TextParameters and ParametersDM unit tests verify correct behaviour
Excellent thanks Willem!
Thanks, Emma. I'll fix this bug while implementing the change to the POLY/TAYLOR default.
bugs/506 new TextParameters and ParametersDM unit tests verify correct behaviour
bugs/506 new ParametersDM class manages DM variations in ephemeris files
pam --update_DM uses DIST_DM1 parameter to update DM
FITSArchive::unload/load DynamicResponse more resiliant to faults
udpate pac help
test_ReceptionModel can simulate fixed dipole observations
Merge branch 'pcm_use_baseline_as_constraint' of ssh://git.code.sf.net/p/psrchive/code into pcm_use_baseline_as_constraint
optionally print a report on the covariances between each pair of free model parameters
remove baseline before automatically selecting phase bins as constraints
AT3-1081 do not attempt to access profile data when in search mode
TemplateDifference::difference -> TemplateDifference::reference
psrdiff -W prints the inverse off-pulse variance weighted reduced chisq
optionally include the off-pulse baseline as a source of zero Stokes V
PolnProfileStats::get_total_squared_invariant and StandardData::get_stokes
connect pat -n and -c to any shift estimator that inherits HasMaxHarmonic
epsic update and minor documentation addition
update to latest epsic
Archive::unload does not create path returned by dirname if empty
successfully link against latest versino of epsic
use latest version of epsic, which compiles with clang version 18.1.8
pathname -> dirname; duplicate basename removed
ambiguous operator overload
Fix merged with additional compilation fix commit 4c3d1df97f9cfc3edbb9f9e73b9f3d5cbf6a181a Author: Willem van Straten <vanstraten.willem@gmail.com> Date: Thu Aug 28 06:53:45 2025 +1200 show clang that ConfigurableProjection::Transformation inherits VariableTransformationManager::Transformation before overriding get_transformation.
show clang that ConfigurableProjection::Transformation inherits
bugs/505 additional overloaded operators (== and !=) in Reference::To
This fix is implemented in commit 826be506b4e77b68bef3baa73884685f894fdb1c Author: Willem van Straten <vanstraten.willem@gmail.com> Date: Thu Aug 28 06:39:03 2025 +1200 bugs/505 additional overloaded operators (== and !=) in Reference::To take priority over built-in candidate operators and eliminate ambiguity ... still a couple more clang compiler errors to fix before this will be merged onto the main branch
Additional overloaded operators (== and !=) in Reference::To<Type> take a non-const Type* argument and take priority over built-in candidate operators, thereby eliminating the ambiguity. To handle Reference::To<const Type> the new non-const overloads are conditionally enabled using std::enable_if
ambiguous operator overload
Merge branch 'master' of ssh://git.code.sf.net/p/psrchive/code
disable 'make dist' if autoconf is out of date
Merge branch 'master' of ssh://git.code.sf.net/p/psrchive/code
disable if autoconf is out of date
python: update to build with numpy 2.
ac_python_devel: remove distutils (deprecated but not needed).
psrdnu searches for optimal frequency offset like pdmp
release instructions updated and simplified
add instructions to tag the release
bump release date to 2025-08-21
indentation; no change
python: update ignore statement (avoids Jones issues)
needless FITSArchive::load_Pointing correcting RA_SUB
pcmdiff now works when pcm.fits does not contain CalibratorStokes extension
relax check for correction flag consistency when int:aux:dm|rm equals zero
pcm once again works with purely polar decomposition
indentation policy set in .vscode/settings.json
verbosity clean up
'zap cal robust' can remove outliers from ConfigurableProjectionExtension
pcm -s erases all gain variation, including those built into model
To address this issue, the normalized Stokes parameters are now multiplied by the gain of the current best estimate of the instrumental response, as derived from absolute value of the determinant of the Jones matrix. commit 7f84d430bdabe037e67aaf529dbb40517e4b99ec Author: Willem van Straten <vanstraten.willem@gmail.com> Date: Mon Aug 4 18:35:51 2025 +1000 bugs/503 multiply the normalized Stokes parameters by the gain of the current best estimate of the instrumental response
bugs/503 multiply the normalized Stokes parameters by the gain
pcm -s erases all gain variation, including those built into model
pcm computes incorrect first guess for pulsar Stokes parameters
ReceptionCalibrator::integrate_pulsar_data now calls SystemCalibrator::get_transformation (const MJD& epoch, Signal::Source source, unsigned ichan), which correctly configures the model and returns a transformation object that correctly computes the current best estimate of the instrumental response. commit f8f447e8309f4ddd043883c4bf17ef2d6eb9cc4f Author: Willem van Straten <vanstraten.willem@gmail.com> Date: Sun Aug 3 21:03:48 2025 +1000 pac -i enables emulation/simulation of calibrated effects...
pcm computes incorrect first guess for pulsar Stokes parameters
pac -i enables emulation/simulation of calibrated effects
known projection of configurable projection can be set
psrsh TF result differs from FT result when aux:rm/dm used
commit 0747da532b08b33ce2fe416c7bc6d483701cdb30 Author: Willem van Straten <vanstraten.willem@gmail.com> Date: Mon Jul 28 20:24:18 2025 +1000 bump release date to 2025-07-28
bump release date to 2025-07-28
bugs/501 compute delta only if there is a history
bugs/501 more fully separate/clarify relative and absolution corrections
psrsh TF result differs from FT result when aux:rm/dm used
further clean up and documentation of ManualPolnCalibrator