Thermal noise still needs attention

Jameson Graef Rollins authored 6 years ago and Jameson Rollins committed 6 years ago

for consistency with all other noise calculation functions.

This reverts commit 4a60fc58.

This was not an appropriate fix to whatever was needing to be addressed.

returns list of (param, value, other_value) tuples.

command line interface added as well.

This is not elegant, but it works.

--dump option was not working anyway because of changes to precompIFO (which
hopefully will be going away).  text output now does not do precomp.

matplotlib.pyplot throws a RuntimeError when no display is present.  One
way around that is to set the backend to AGG.  But we don't want AGG when
we do have a display.  So we only want to set it opportunistically.  But
it also has to be set before anything else is imported.  Kind of convoluted.

So we really just import pyplot as opportunistically as possible, and then
rely on setting AGG backend when we're sure we don't need it in main().

All noise calculation functions in pygwinc take the arguments (freq, ifo)
in that order.  It's annoying that for some reason this one is different.

The products of precompQuantum (primarily fSQL) are currently only
used in squeeze.py (which is itself not used anywhere).  Remove this
function from preocomp, and move relevant bits into squeeze.py

Also remove unused variables in precompPower that could confuse
memoization.

adds CI documentation builder

See merge request !27

With this tweak, most noises agree within 1 ppb.  Exceptions are
seismic and substrate TE (which still have open bugs in matgwinc).

Arguments added to gwinc.test for setting a tolerance level, and a
list of noises to skip in the comparison.

The CI pipeline will now fail when tests fail.

substratethermal: use adiabatic approx for carrier density & thermorefractive noises

See merge request !35

Note: these are cryo-silicon-only noises.
The adiabatic approx is quite accurate and evaluates faster than
the exact integral.
See also https://git.ligo.org/rana-adhikari/CryogenicLIGO/issues/37

coatingthermal: update for field penetration and freq-dependent loss

See merge request !34

Update quantum code from matgwinc, and vectorize

See merge request !32

Christopher Wipf authored 6 years ago and Lee McCuller committed 6 years ago

The matgwinc updates include a new shotradSignalRecycled(), and minor
corrections to the previous code, which is still available as
shotradSignalRecycledBnC().

shotradSignalRecycled() has been vectorized for good performance,
with its two TF inversions pre-computed by sympy.

Speed up quad sus tf calculation 10x, by pre-solving the system symbolically

See merge request !31