Simulate wideband TOAs using zima

[abhisrkckl]

I have added wideband TOA simulation capability to zima. The underlying functionality is of course implemented in pint.simulation.

This is invoked using two new command line options --wideband and --dmerror).
Here is an example command:
$ zima --startMJD=56600 --addnoise --wideband --dmerror=1e-4 J0100+8023.par J0100+8023.sim

This feature computes the expected DM time series using the model (from the input par file), adds white noise based on the --dmerror option, and sets appropriate -pp_dm and -pp_dme flags.

The injected error takes care of the DMEFACs and DMEQUADs as follows: -pp_dme are set equal to the --dmerror, and the injected noise is computed with the help of model.scaled_dm_uncertainty(toas).

[abhisrkckl]

My use case for this is that I need to simulate WB TOAs to test PR #1426.

[dlakaplan]

I haven't used zima extensively. Have you checked this using the straight python interface in pint.simulation? When I tested those for NB data they did respect EFAC/EQUAD in the model, but I don't know if I did WB data.

[abhisrkckl]

The earlier behaviour in pint.simulation was to put arbitrary dm values in the toa files instead of computing them based on the model. I have changed that.

The wierd thing is that the parameters are estimated correctly for narrowband data. The under/overestimate happens only for wideband. I had checked this before doing the previous PR and re-checked now.

I am planning to do more simulations of the same par file and see if this is systematic or happens randomly.

[codecov]

Codecov Report

Base: 66.16% // Head: 66.19% // Increases project coverage by +0.02% 🎉

Coverage data is based on head (ee61945) compared to base (7c840a7).
Patch coverage: 86.11% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1429      +/-   ##
==========================================
+ Coverage   66.16%   66.19%   +0.02%     
==========================================
  Files          91       91              
  Lines       21353    21367      +14     
  Branches     3701     3702       +1     
==========================================
+ Hits        14128    14143      +15     
+ Misses       6371     6370       -1     
  Partials      854      854              

Impacted Files	Coverage Δ
src/pint/residuals.py	73.49% <ø> (ø)
src/pint/simulation.py	83.44% <76.19%> (+1.75%)	⬆️
src/pint/scripts/zima.py	100.00% <100.00%> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[dlakaplan]

Can you add a test in test_fake_toas to do the wb toas without zima?

[abhisrkckl]

I have added the test in test_fake_toas.

[dlakaplan]

This looks good. Do you want me to merge or wait until the meeting to discuss?

[abhisrkckl]

We had discussed this a few times in the past meetings. I think this is OK to merge.

[dlakaplan]

OK, I think you might just need a CHANGELOG entry