Gnssrefl

Added refraction model number (refr_model) to gnssir_input and thus station analysis json.
on the way to implementing more than one refraction model. Default will remain refr_model = 1.
0 is no model. All information about the refraction models are in the code for gnssir_input

Double-checked that vwc code is creating a txt output file.

2023 September 26

Remove zero points from SNR traces in window_new.py in gnssir_v2.py

2023 September 25
added nmea2snr changes from naoyakadota. Invalid lines no longer crash the code.

multiple frequencies allowed

https://github.com/kristinemlarson/gnssrefl/pull/194

2023 September 20

Added plot changes to subdaily (e.g. allow fontsize choice from command line)

Fixed a weirdness where GNSS files with no SNR data would create a file with
all zeros data. Which is technically correct, but not terribly friendly. The
GPS orbit option exited with a message. Now GNSS files also exit with a message.

changed nyquist to max_resolve_RH so it makes more sense to learners. i.e. this is the max
resolvable RH you can expect for a given station location.

2023 September 19

Allow Hatanaka and Hatanaka + unix compressed files (.Z) in the local directory
for the nolook option. Hatanaka files also allowed in $REFL_CODE/YYYY/rinex/ station directory

Write out error message when people inadvertently throw out all data because
they did not take into account that their receiver had an elevation mask and
used e1 that is too small.

2023 September 17

rinex2snr SNR file creation fails are written to the log file. And users are told this on the screen.
This does not seem to be helping people. This version writes the log file's explicit name to the screen

Various docs were hopefully improved.

gzipping the SNR files is now the default when running gnssir. Should save space
and I don't think it will be a huge time sink. If it is, users can set gzip to False
on the command line.