Figure out better way to bring in TDCF values to the collection of measurements for the GPR
The current method for TDCF values to be applied to the measurements as input to the GPR is a kludge: each measurement has an MCMC computed, those values are saved to a separate file (copied by hand, oof!), and values are applied. We need a better way to do this that is less prone to error.
The original idea was to have calibration line generated TDCF values to be used, but these lines are turned off during measurements. The lines could be off for several hours, so reliable values are not guaranteed.
We need to think about a better, less error-prone way to do this.
Activity
changed milestone to %Implement measurement tools
One of Joe B's requests (#11 (closed)) are related to this--potentially logging MCMC values.
marked this issue as related to #10
The corresponding Integration Issue and ECR Tracker (IIET) ticket for this issue is https://services1.ligo-la.caltech.edu/FRS/show_bug.cgi?id=13730.
The Description and Comment 2 in that ticket, may be helpful for reference.
@jeffrey-kissel There seems to be two issues you mention in the ECR ticket:
- error-prone issue of hard-coding MCMC MAP values into a python file for reading in. This is addressed in !56 (merged) and !56 (merged) actually addresses both this git issue and the first part of the ECR ticket. Thus I'd like to close this git issue.
- differences in the way that different analyses are compensating differently (comment 2), which this git issue does not address. I'd like to understand better why there are differences in analysis of the high frequency roaming lines, where that code is going to slot in to our new workflow and who is going to work on that. Perhaps @sudarshan.karki is going to need to get involved to add in some new code for pyDARM to process the high frequency roaming lines
If @jeffrey-kissel and @sudarshan.karki are ok with it, I propose to close this issue and open a new issue that discusses point 2 above and link it to the ECR.
-
@evan-goetz Whatever librarianship method you chose works for me, as long as the folks eventually working on it can follow that rabbit holes to the comment / information history of the issues (be they the IIET ticket, or the comment threads on this ticket).
mentioned in merge request !56 (merged)
We've done what this issue set out to address in !56 (merged), so this particular issue can be closed. Comments left by Jeff above regarding ECR ticket comments have produced two new issues #55 and #56.
