Add tests: `pet -> raw` simulation chain

[fhagemann]

I slightly modified the pet -> raw simulation chain I was given by @sagitta42 to act as a test script.

Disclaimer: the tests will fail now because LegendTestData.jl still links to an old commit of legend-test-data where invcoax-metadata.json is still in the old metadata format. Once LegendTestData.jl is updated, these tests should pass. I would ask @oschulz to please take care of this.

Running the tests, the function propdict threw a deprecation warning. I updated it to use the PropDicts readprops function. We could think about getting rid of propdict in LegendGeSim.jl and replace all propdict by readprops. Any thoughts about this?

[fhagemann]

Don't merge until LegendTestData.jl is updated.

[sagitta42]

@fhagemann I updated this json to be of the correct format at some point when it stabilized, I think before the collab meeting

https://github.com/legend-exp/legend-testdata/blob/main/data/ldsim/invcoax-metadata.json

[fhagemann]

Yes, but the package LegendTestData.jl is not yet linking to the commit where this is updated. So, when using LegendTestData.jl, it's still pulling a config file in the old format from before the collaboration meeting.

[fhagemann]

You can see the commit hash it's linking to the commit 7df474059cb6cf1d7c7ff836ddfb7e8d3373abb0 (see line 6 in Artifacts.toml)

which is this one (still in the old format): https://github.com/legend-exp/legend-testdata/blob/7df474059cb6cf1d7c7ff836ddfb7e8d3373abb0/data/ldsim/invcoax-metadata.json

[fhagemann]

With the new releases of LegendTestData.jl and RadiationDetectorSignals.jl, the current tests should pass. I also added tests for waveforms with noise applied to them.

[codecov[bot]]

Codecov Report

Patch coverage: 100.00% and project coverage change: +36.24% :tada:

Comparison is base (eaae6ee) 13.60% compared to head (bd00eca) 49.84%.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #39 +/- ## =========================================== + Coverage 13.60% 49.84% +36.24% =========================================== Files 22 22 Lines 669 664 -5 =========================================== + Hits 91 331 +240 + Misses 578 333 -245 ``` | [Files Changed](https://app.codecov.io/gh/legend-exp/LegendGeSim.jl/pull/39?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=legend-exp) | Coverage Δ | | |---|---|---| | [src/temp\_utils.jl](https://app.codecov.io/gh/legend-exp/LegendGeSim.jl/pull/39?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=legend-exp#diff-c3JjL3RlbXBfdXRpbHMuamw=) | `55.55% <100.00%> (+38.88%)` | :arrow_up: | ... and [14 files with indirect coverage changes](https://app.codecov.io/gh/legend-exp/LegendGeSim.jl/pull/39/indirect-changes?src=pr&el=tree-more&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=legend-exp)

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[sagitta42]

I started adding tests after @fhagemann told me "add tests" :D But he also added tests and submitted the PR before me.

I think I guess we just work with this PR and I scratch what I added. Although, I would recommend to not use json files in the tests but explicitly written out dictionaries so that it's clear what's happening there.

For example, this is what I ended up having:

@testset "Package LegendGeSim" begin
    environment_settings = Dict(
        "crystal_temperature_in_K" => 77,
        "medium" => "vacuum"
    )

    simulation_settings = Dict(
        "method" => "ssd"
    )

    for (filename, capacitance) in test_dict
        detector_metadata_filename = joinpath(testdata_path, filename)

        @testset "Field Simulation of $filename (SSD)" begin
            sim = LegendGeSim.simulate_fields(detector_metadata_filename, environment_settings, simulation_settings)
            C = LegendGeSim.capacitance_matrix(sim)

            @testset "Capacitances" begin   
                @test isapprox(C[1,2], capacitance, atol = 0.2u"pF")
            end
        end

        @testset "Pulse Simulation" begin
            path_to_pet_file = joinpath(LegendTestData.legend_test_data_path(), "data", "ldsim", "single-invcoax-th228-geant4.csv")
            # --- ideal
            # pet -> pss
            pss_table, pss_truth = LegendGeSim.simulate_pulses(det_metadata, path_to_pet_file, environment_settings, simulation_settings)

            # --- realistic
            setup_settings = Dict(
                "preamp" => Dict(
                    "type" => "generic",
                    "t_decay" => 50,
                    "t_rise" => 15,
                    "max_e" => 10000,
                    "offset" => 2000
                ),
                "fadc" => Dict(
                    "type" => "generic",
                    "sampling_interval" => 16
                ),
                "trigger" => Dict(
                    "type" => "trapezoidal",
                    "window_lengths" => [250,250,250],
                    "threshold" => 9
                ),
                "daq" => Dict(
                    "type" => "generic",
                    "nsamples" => 3750,
                    "baseline_length" => 1875
                )
            )
            # pss -> raw
            raw_table = LegendGeSim.pss_to_raw(pss_table, pss_truth, setup_settings)
            # pet -> raw
            raw_table1 = LegendGeSim.simulate_waveforms(det_metadata, path_to_pet_file, environment_settings, simulation_settings, setup_settings)
        end

        # ToDo
        # - noise model: FromSim & FromData
    end 
end

Note: to make this work I had to add a function that can do pss_to_raw(pss_table, ...) (I only added a wrapper that takes pss_name and reads table, doesn't take table itself, the mid step was missing)

[fhagemann]

@sagitta42 With your example test file, I get the error that det_metadata is undefined. So you want det_metadata to be defined in the tests as dictionary instead of being read in from LegendTestData.jl?

And we also cannot use the CSV file with hits for an ICPC detector for a random BEGe detector if we use your loop that iterates over all detector geometries that we want to test (?)

[sagitta42]

@fhagemann

With your example test file, I get the error that det_metadata is undefined.

So you want det_metadata to be defined in the tests as dictionary instead of being read in from LegendTestData.jl?

Sorry that was just copy-paste typo. In my snippet, it should be detector_metadata_filename.

No, metadata is indeed easier to read from LegendTestData, but settings are better defined as dictionaries rather than json. This way it's more clear what's being tested, and also can loop e.g. for method in ["SSD", "siggen"] ...

And we also cannot use the CSV file with hits for an ICPC detector for a random BEGe detector if we use your loop that iterates over all detector geometries that we want to test (?)

Well, inside pet->stp there's the part that removes events outside of the detector :D So I thought this might work anyway, just not perfectly tailored to each detector, but some events would eventually be in the bulk x)

What I have in plan now is to have a function in LegendGeSim which can toy-monte-carlo a pet table for given detector with given SSE/MSE fraction. This is useful way beyond testing for many a study for PSS. I'm planning to make a call about it. Should be nice in Julia because I could basically generate positions in a cylinder and then use that same feature to remove the ones outside of the detector. And, it belongs in LegendGeSim! Generate your PET sample to test PSD or do a study of noise impact at lower energies or whatnot, then process right there with LegendGeSim. Sorry I got carried away

[fhagemann]

The tests now pass with the newest version of LegendHDF5IO.jl

[oschulz]

@sagitta42 merge?

[fhagemann]

Two possibilities:

• Merge this PR now and modify the tests based on @sagitta42 comments afterward. That way we have a baseline of tests which pass before modifying, so we can always default back.
• Already incorporate the proposed changes now. Might result in fewer PRs and fewer merge conflicts.

[oschulz]

@sagitta42 what's your preference?

[sagitta42]

@fhagemann @oschulz Let's merge now, then deal with other PRs, then add tests and other potential mods.