silas/imdd_silas
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Refactor preprocessSignal usage across DSP workflows

Browse Source
This commit is contained in:
Silas Oettinghaus
2026-03-25 14:36:46 +01:00
parent 4cef384d15
commit 8c55b25112
5 changed files with 12 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

195
Tests/TEST_TODO.md Normal file
View File

@@ -0,0 +1,195 @@
# Test TODO
This file is the current backlog for building out the MATLAB test suite.
It should reflect the repository state now, not the original plan.
## Current Status
Fast profile status:
- `run_all_tests('profile','fast')`
- latest verified result: `Passed: 82 | Failed: 0 | Incomplete: 0`
- explicit placeholder tests remaining: `62`
Mandatory workflow for every future test-writing session:
1. Read the target class implementation in `Classes/...` before writing or changing tests.
2. Read the current placeholder or existing test in `Tests/...`.
3. Replace or extend the existing test. Do not create duplicate test files.
4. Use the existing `matlab.unittest` framework and extend `IMDDTestCase`.
5. Tag new tests with `TestTags`.
6. Run the relevant test entrypoint after implementation.
7. Do not consider the task complete until the test run result is checked.
## Completed Base Tests
These are already implemented and verified in the fast profile:
- `Classes/00_signals/Signal.m` -> `Tests/00_signals/Signal_test.m`
- `Classes/00_signals/Informationsignal.m` -> `Tests/00_signals/Informationsignal_test.m`
- `Classes/00_signals/Electricalsignal.m` -> `Tests/00_signals/Electricalsignal_test.m`
- `Classes/00_signals/Opticalsignal.m` -> `Tests/00_signals/Opticalsignal_test.m`
- `Classes/01_transmit/PAMmapper.m` -> `Tests/01_transmit/PAMmapper_test.m`
- `Classes/01_transmit/PAMsource.m` -> `Tests/01_transmit/PAMsource_test.m`
- `Classes/01_transmit/Signalgenerator.m` -> `Tests/01_transmit/Signalgenerator_test.m`
- `Classes/01_transmit/Pulseformer.m` -> `Tests/01_transmit/Pulseformer_test.m`
- `Classes/02_etc/Amplifier.m` -> `Tests/02_etc/Amplifier_test.m`
- `Classes/02_etc/Filter.m` -> `Tests/02_etc/Filter_test.m`
- `Classes/02_optical/Fiber.m` -> `Tests/02_optical/Fiber_test.m`
- `Classes/02_optical/EML.m` -> `Tests/02_optical/EML_test.m`
- `Classes/02_optical/Optical_Multiplex.m` -> `Tests/02_optical/Optical_Multiplex_test.m`
- `Classes/02_optical/Optical_Demultiplex.m` -> `Tests/02_optical/Optical_Demultiplex_test.m`
- `Classes/02_optical/DP_Fiber.m` -> `Tests/02_optical/DP_Fiber_test.m`
- `Classes/03_receive/Photodiode.m` -> `Tests/03_receive/Photodiode_test.m`
- `Classes/03_receive/Scope.m` -> `Tests/03_receive/Scope_test.m`
- `Classes/04_DSP/TransmissionPerformance.m` -> `Tests/04_DSP/TransmissionPerformance_test.m`
- `Classes/04_DSP/Equalizer/FFE.m` -> `Tests/04_DSP/Equalizer/FFE_test.m`
- `Classes/04_DSP/Sequence Detection/MLSE.m` -> `Tests/04_DSP/Sequence Detection/MLSE_test.m`
- `Classes/04_DSP/Timing Recovery/Timing_Recovery.m` -> `Tests/04_DSP/Timing Recovery/T_04_DSP_Timing_Recovery_Timing_Recovery_test.m`
Notes:
- There are multiple `Timing_Recovery` class names in the repo, so the implemented base-class test currently uses the disambiguated file name:
`Tests/04_DSP/Timing Recovery/T_04_DSP_Timing_Recovery_Timing_Recovery_test.m`
## Integration Tests
These are the current end-to-end regression guards:
- `Tests/integration/IMDD_base_system_minimal_integration_test.m`
- `Tests/integration/IMDD_base_system_no_impairment_integration_test.m`
- `Tests/integration/IMDD_base_system_impairment_monotonicity_integration_test.m`
- `Tests/integration/WDM_mux_demux_optical_chain_integration_test.m`
- `Tests/integration/WDM_end_to_end_two_channel_receiver_integration_test.m`
Current integration focus:
- refactor the duplicated reduced IM/DD workflow setup into shared integration helpers
- tighten the IM/DD baseline naming and thresholds so the physical intent is honest
- strengthen the WDM optical-chain and end-to-end receiver fixtures into harder regression guards
- keep all integration fixtures deterministic and small enough to remain reviewable
## Remaining Core Tests
Highest-value unfinished unit tests to implement next:
1. `Classes/02_optical/Polarization_Controller.m`
`Tests/02_optical/Polarization_Controller_test.m`
Focus:
- deterministic transform behavior
- shape preservation
2. `Classes/01_transmit/AWG.m`
`Tests/01_transmit/AWG_test.m`
Focus:
- DAC-like resampling and quantization behavior
- length and sampling-rate invariants
- deterministic small fixtures
Follow-up around the WDM integration fixtures:
- `Classes/01_transmit/AWG.m`
- `Classes/02_optical/EML.m`
- `Classes/02_optical/Polarization_Controller.m`
- `Classes/02_optical/DP_Fiber.m`
- `Classes/02_optical/Optical_Multiplex.m`
- `Classes/02_optical/Optical_Demultiplex.m`
- `Classes/03_receive/Photodiode.m`
These should be revisited together when the WDM integration tests are tightened.
Immediate integration refactor backlog:
1. Extract shared reduced IM/DD integration helpers
Targets:
- `Tests/integration/IMDD_base_system_minimal_integration_test.m`
- `Tests/integration/IMDD_base_system_no_impairment_integration_test.m`
- `Tests/integration/IMDD_base_system_impairment_monotonicity_integration_test.m`
Goal:
- remove duplicated reduced-workflow builders
- keep assertions local to each test file
2. Revisit the "no impairment" IM/DD fixture
Target:
- `Tests/integration/IMDD_base_system_no_impairment_integration_test.m`
Goal:
- ensure the name, comments, and thresholds reflect the actual physical regime
3. Harden the WDM optical-chain regression guards
Targets:
- `Tests/integration/WDM_mux_demux_optical_chain_integration_test.m`
- `Tests/integration/WDM_end_to_end_two_channel_receiver_integration_test.m`
Goal:
- improve channel-separation assertions
- add propagated-chain checks where stable
- avoid brittle waveform snapshots
## Next DSP Layer
After the items above, continue with these:
- `Classes/04_DSP/Coding/Duobinary.m`
`Tests/04_DSP/Coding/Duobinary_test.m`
- `Classes/04_DSP/Coding/MRDS_coding.m`
`Tests/04_DSP/Coding/MRDS_coding_test.m`
- `Classes/04_DSP/Sequence Detection/MLSE_viterbi.m`
`Tests/04_DSP/Sequence Detection/MLSE_viterbi_test.m`
- `Classes/04_DSP/Sequence Detection/MLSE_new.m`
`Tests/04_DSP/Sequence Detection/MLSE_new_test.m`
- `Classes/04_DSP/Timing Recovery/Time_Shifter.m`
`Tests/04_DSP/Timing Recovery/Time_Shifter_test.m`
- `Classes/04_DSP/Timing Recovery/Godard_Timing_Recovery.m`
`Tests/04_DSP/Timing Recovery/Godard_Timing_Recovery_test.m`
- `Classes/04_DSP/Timing Recovery/MaxVar_Timing_Recovery.m`
`Tests/04_DSP/Timing Recovery/MaxVar_Timing_Recovery_test.m`
Then move on to the remaining equalizer variants:
- `EQ.m`
- `EQ_copy.m`
- `EQ_silas*.m`
- `FFE_*`
- `ML_MLSE*.m`
- `Postfilter.m`
- `VNLE.m`
## Lower Priority Areas
Leave these for later unless a current task depends on them:
- `Classes/Warehouse_class/*`
- `Classes/DataBaseHandler/*`
- template/example classes like `class_template`
## Excluded From Automated Testing
Do not create or run automated tests for lab device classes in this repository.
Excluded area:
- `Classes/05_Lab/*`
Reason:
- the device IPs are reachable in this environment
- executing these classes can change instrument set points
- this can interfere with external or ongoing experiments
Rule for future sessions:
- do not implement `Tests/05_Lab/*`
- do not run files that communicate with lab devices
- do not add these classes back into the active TODO backlog unless explicitly requested by the user for a controlled setup
## Session Prompt Checklist
Every future agent prompt should include these requirements:
- read the class implementation first
- read the current test file second
- implement the test in the existing file
- prefer deterministic synthetic fixtures
- avoid plotting and hardware unless required
- never run or implement `05_Lab` tests in normal agent sessions
- run `run_all_tests('profile','fast')` after implementation
- report the changed files and the verification result