make BudgetItem.freq a property

See merge request !149

<unset> removes struct entries with inheritance again

See merge request gwinc/pygwinc!148

remove vestigial function calculating squeeze angle

See merge request gwinc/pygwinc!150

Fix bugs in load_budget

See merge request gwinc/pygwinc!144

Pytest enhancements

See merge request gwinc/pygwinc!151

update Voyager FC input coupler and detuning

See merge request gwinc/pygwinc!156

Improves total noise a little more than 5% and quantum noise a little
more than 10% around 30 Hz.

fix bug in squeezer type options

See merge request gwinc/pygwinc!147

Removed the function sqzOptimalSqueezeAngle from quantum.py. It is no
longer used and wasn't great to begin with.

* Fixes bug where the squeezer exists even if Squeezer.Type is 'None'
* Removes the squeezer if Squeezer.AmplitudedB = 0
* Adds unit tests to check squeezer options

fix bug when updating ifo struct in budget run

See merge request gwinc/pygwinc!146

Quick fix so that the Freq Independent squeezing option works again

See merge request gwinc/pygwinc!145

simple inheritance in yaml files through load_budget

See merge request gwinc/pygwinc!127

Adds residual gas damping and additional molecular species to the residual gas calculations. The gas scattering and gas damping are now tracked separately for each species in an excess gas sub-budget. The damping calculation is done in the infinite volume limit by default but adds corrections for squeezed film damping if specified in the yaml file.

H2, N2, H2O, and O2 are considered. The numbers for aLIGO are taken from the link in this commit. The numbers for Aplus and Voyager are copied from aLIGO.  For the existing LIGO facilities the O2 numbers are unknown but are suspected to be low, and are therefore set 10x below the expected noise floor.

Closes #63.

precomp_suspension was being called for no reason in SuspensionThermalHorizTM.calc()

Following Amann et al. (2020) -- ET site selection paper