I have updated the GPR HDF5 files for use in computing response function uncertainty.

Scripts used to produce the HDF5 files:
trunk/Runs/O3/H1/Scripts/Uncertainty/process_allmeas_writeGPRHDF5_model20190416-A.py   r8224
trunk/Runs/O3/H1/Scripts/Uncertainty/process_allmeas_writeGPRHDF5_model20190416-C.py   r8228

GPR HDF5 files for use by RRNom.py:
trunk/Runs/O3/H1/Results/Uncertainty/O3_H1_A_GPR_20190416_multi.hdf5   r8186
trunk/Runs/O3/H1/Results/Uncertainty/O3_H1_C_GPR_20190416_multi_nofsQcorr_fmin30Hz_lengthscaleZp5.hdf5   r8230

Note that there is a new file to use for the GPR of the optical plant. The reason for the change is that below 30 Hz, the MCMC does a poor job fitting the data. This impacts the fit for optical gain and coupled cavity pole frequency (see G1901479, slides 10, 11, and 19), which need to be correctly fit in the frequency range of importance. Also, currently C00 and C01 do not correct for any spring frequency or quality factor Q, so we should not be correcting our measurements by these values--our systematic error at low frequency will naturally expand because of this.

Actuation and sensing MCMC values for the reference model remain the same, as per our procedure.