Prepare O4 analysis ready frames

In O3, GWOSC prepared "ANALYSIS_READY" frames for strain data. These were similar to the h(t) frames, but they only included times when the detectors were in ANALYSIS_READY mode. Documentation for the O3 frames are here:

https://www.gw-openscience.org/O3/O3alt/

For O4, we hope to prepare similar frames, but with low-latency. One way to do this is for the RDS tool that aggregates 1 s calibrated strain frames into 4096 s frames to spit out two versions of the frames - one for all times, and one only for times that satisfy ANALYSIS_READY. Some notes:

• We want to do this for all calibrated strain in O4, but initially, we should plan on doing this for the C00 calibrated frames
• The strain channel name for h(t) should be exactly the same as in the 1 s frames, except should append _AR to the end of the channel name (see O3 docs for examples)
• The ANALYSIS_READY information should be stored in the frames, as a bit in the CALIB_STATE_VECTOR. Alternatively, maybe the same information is in the segment database, but the segment database might be unreliable in low-latency.
• We will want to be selective about which channels from the 1 s frames will be include in the AR frames. The exact channel list is under discussion.
• An example script that was used to do this for O3 is here: https://git.ligo.org/gwosc/timeseries/-/blob/master/multi_chan/O3b/makedata.py
• In the script, you can see at line 109 - 121, strain data is picked up, the channel name is appended with _AR, and then the data are written out to a new file.

I am happy to talk about this, and help as best I am able.