GitLab

Add retries to LALInferenceMCMC checkpointing

See merge request lscsoft/lalsuite!1165

Remove trailing whitespace from lalapps string code

See merge request lscsoft/lalsuite!1183

Remove trailing whitespace from lalapps_power_veto

See merge request lscsoft/lalsuite!1184

Remove trailing whitespace from lalapps inspiral code

See merge request lscsoft/lalsuite!1182

Switch from manylinux2010 to manylinux2014

See merge request lscsoft/lalsuite!1164

It's important to retain manylinux1 because it is supported by
fairly old versions of pip (since 8.1.0). However, manylinux2010
support in pip is fairly recent (since 19.0), so it should be fine
to remove manylinux2010 while adding manylinux2014.

Remove trailing whitespace from lalapps pulsar code

See merge request lscsoft/lalsuite!1181

Reverse order of found_order and found_approximant in XLALSimInspiralDecomposeWaveformString

Closes #251

See merge request lscsoft/lalsuite!1178

Add test of SEOBNRv4PHM in the aligned-spin limit

See merge request lscsoft/lalsuite!1179

This checks that the ringdown (2,2) mode frequency of
SEOBNRv4PHM is consistent with SEOBNRv4HM in the aligned spin
limit, in particular for cases where the final spin is negative
with respect to the orbital angular momentum

MakeSFTs: remove example files / install CopySFTs

See merge request lscsoft/lalsuite!1171

add example python unit test for lalsimulation

See merge request lscsoft/lalsuite!1175

Remove trailing whitespace in lalapps python code

See merge request lscsoft/lalsuite!1167

- Not modified since ~2005 so almost certainly out of date

The waveform string "SEOBNRv4pseudoFourPN" was decomposed as waveform = "SEOBNRv4p"
phase_order = "seudoFourPN".  Instead parse the order first then the approximant.

Fixes #251.

Merge release branch

See merge request lscsoft/lalsuite!1176

(cherry picked from commit cf134e003dfb8ce4c54f9a7740ea07b40f4228fa)

When m1 < m2, the surrogate code previously did the following:
1. Swap the labels of the two BH, i.e. m1 <-> m2, chi1 <-> chi2.
2. Multiply the odd-m modes by -1 to undo this swap.

The assumption here is that the first step is only a rigid rotation, and
the second does undoes this rigid rotation.

However, the first step is only a rigid rotation if after the swap the
in-plane spins are also rotate by pi. This is now fixed.

(cherry picked from commit 9ed713a0857fe716fc768d6e89fe66a01bf9334c)

(cherry picked from commit f1ba9db504fa3179cf00479a30dfb165dd7e6876)

(cherry picked from commit f28e118d083085fdc8c0c420f1cb38d394e8a529)

This fixes a bug when the final total spin is negative with respect
to the orbital angular momentum. It involves 2 changes:

if \hat{J}\cdot\hat{L}<0
i) flipping the sign of the final spin
ii) flipping the sign of the precession frequency in the post-
merger extension

(cherry picked from commit a8b4cbb3f17765404bf82aaa312ebc0221850abf)

since we build for python3 as well now

(cherry picked from commit 8023cd0b47eac52a1320de33eb75220d32d9b1aa)

Fstatistic/Makefile.am: remove disabling of tests if CONDOR_ENABLED

See merge request lscsoft/lalsuite!1169

MakeSFTs: add test script for lalapps_MakeSFTs

Closes CW/software/lalsuite#63

See merge request lscsoft/lalsuite!1170

Merge release branch

See merge request lscsoft/lalsuite!1168