lalapps_cbc_sbank_sim_pipe.py: unexpected hacked gpsend time
On line 311, the GPS end-time is calculated using "50*(njobs*injperjob)" for a reason related to 'pruning'. However, it is not clear that anyone is actually pruning or that it should be hard-coded. I propose the following fix:
217,218d216
< time-step = 10
< time-interval = 5
308d305
< timestep = int(cp.get("inspinj", "time-step"))
314c311
< gpsend = gpsstart + (njobs * timestep)
---
> gpsend = gpsstart + 50 * (njobs * injperjob)
317c314
< #cp.set("inspinj", "time-step", str(1))
---
> cp.set("inspinj", "time-step", str(1))
So, here we see several changes:
- First, I am reading in a time-step and time-interval variable rather than ignoring them. This happens in the "documentation" part of the code.
- Next, I actually read in the timestep.
- Consecutively, I remove the "50 *" part as it is not required
- Lastly, I multiply the number of jobs by the timestep rather than injperjob as 'njobs * injperjob' gives a unitless number of injections rather than a time.
Alternatively, we could leave injperjob and continue to reset the timestep to 1 in the code but then I think we should add a note that timestep will be ignored. However, I see little reason to leave the 50*.
Marnix
<\pre>