Add sonar beamformers

[PhilClemson]

This PR adds two new detectors for application to sonar tracking problems. Both take hydrophone data and convert these signals to a set of detections using a windowing method.

capon is a standard Capon beamformer. rjmcmc is a novel beamforming algorithm based on reversible-jump Markov chain Monte Carlo. The latter algorithm is designed for multiple target tracking, but the Metropolis-Hastings algorithm currently used does not perform well in these situations. I've therefore disabled the multiple target detection for now by forcing the update type in proposal_func(). The current algorithm has since been replaced with a more effective no-u-turn sampler (NUTS) but this is still undergoing testing (I'll include this in another PR at a later date).

The usage is demonstrated in the corresponding test file test_beamformers.py, where I've included a small simulated dataset fixed_target_example.csv.

[codecov[bot]]

Codecov Report

Merging #490 (4a6b60e) into main (1ec7db2) will increase coverage by 1.80%. The diff coverage is 72.53%.

:exclamation: Current head 4a6b60e differs from pull request most recent head 6535754. Consider uploading reports for the commit 6535754 to get more accurate results

@@            Coverage Diff             @@
##             main     #490      +/-   ##
==========================================
+ Coverage   91.42%   93.23%   +1.80%     
==========================================
  Files         144      145       +1     
  Lines        7009     7344     +335     
  Branches     1034     1110      +76     
==========================================
+ Hits         6408     6847     +439     
+ Misses        484      389      -95     
+ Partials      117      108       -9     

Flag	Coverage Δ
integration	65.66% <14.32%> (?)
unittests	90.56% <72.53%> (-0.87%)	:arrow_down:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
stonesoup/detector/beamformers.py	72.53% <72.53%> (ø)
stonesoup/metricgenerator/ospametric.py	86.39% <0.00%> (+0.59%)	:arrow_up:
stonesoup/simulator/simple.py	97.76% <0.00%> (+1.49%)	:arrow_up:
stonesoup/base.py	94.65% <0.00%> (+1.52%)	:arrow_up:
stonesoup/initiator/simple.py	97.32% <0.00%> (+1.78%)	:arrow_up:
stonesoup/sensor/radar/radar.py	97.23% <0.00%> (+2.20%)	:arrow_up:
stonesoup/types/particle.py	96.22% <0.00%> (+3.77%)	:arrow_up:
stonesoup/feeder/geo.py	93.61% <0.00%> (+4.25%)	:arrow_up:
stonesoup/metricgenerator/manager.py	100.00% <0.00%> (+4.87%)	:arrow_up:
stonesoup/metricgenerator/plotter.py	95.00% <0.00%> (+5.00%)	:arrow_up:
... and 11 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 1ec7db2...6535754. Read the comment docs.

[PhilClemson]

I have now added some descriptions / references above the beamformer classes. I also added a user-set random seed so that the RJMCMC algorithm will give reproducible output (for testing).

[PhilClemson]

Thanks for the comments @sdhiscocks. I've commited some changes but it's not ready for review yet.

[PhilClemson]

Okay, I think I've made all the requested changes but let me know if I missed anything!

[PhilClemson]

Thanks again for looking at this @sdhiscocks and apologies for taking so long to get back to you with the changes (unfortunately Christmas got in the way and then I was busy with other work in January).

I think I've fixed all the issues that were mentioned but it might be worth checking the new reader implementation with itertools is usable for streaming data.