1294 add lct secir model with two disease strains #1340

an-jung · 2025-08-07T11:50:59Z

Changes and Information

Please briefly list the changes (main added features, changed items, or corrected bugs) made:

added LCT SECIR Model for 2 DISEASES on the basis of the LCT SECIR model

If need be, add additional information and what the reviewer should look out for in particular:

Merge Request - Guideline Checklist

Please check our git workflow. Use the draft feature if the Pull Request is not yet ready to review.

Checks by code author

Every addressed issue is linked (use the "Closes #ISSUE" keyword below)
New code adheres to coding guidelines
No large data files have been added (files should in sum not exceed 100 KB, avoid PDFs, Word docs, etc.)
Tests are added for new functionality and a local test run was successful (with and without OpenMP)
Appropriate documentation within the code (Doxygen) for new functionality has been added in the code
Appropriate external documentation (ReadTheDocs) for new functionality has been added to the online documentation
Proper attention to licenses, especially no new third-party software with conflicting license has been added
(For ABM development) Checked benchmark results and ran and posted a local test above from before and after development to ensure performance is monitored.

Checks by code reviewer(s)

Corresponding issue(s) is/are linked and addressed
Code is clean of development artifacts (no deactivated or commented code lines, no debugging printouts, etc.)
Appropriate unit tests have been added, CI passes, code coverage and performance is acceptable (did not decrease)
No large data files added in the whole history of commits(files should in sum not exceed 100 KB, avoid PDFs, Word docs, etc.)
On merge, add 2-5 lines with the changes (main added features, changed items, or corrected bugs) to the merge-commit-message. This can be taken from the briefly-list-the-changes above (best case) or the separate commit messages (worst case).

Closes #1294

started changing Infection States, Parameters etc. to reflect 2 diseases

changed Infection States, Parameters etc. to reflect 2 diseases

…rains

codecov · 2025-08-07T12:42:56Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.36%. Comparing base (0b61820) to head (ab75b17).
⚠️ Report is 10 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1340      +/-   ##
==========================================
+ Coverage   97.29%   97.36%   +0.06%     
==========================================
  Files         174      176       +2     
  Lines       15105    15675     +570     
==========================================
+ Hits        14697    15262     +565     
- Misses        408      413       +5

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

jubicker · 2025-08-28T12:14:50Z

cpp/examples/lct_secir.cpp


 int main()
 {
    // Simple example to demonstrate how to run a simulation using an LCT-SECIR model.


If the changes in this file have don't have a specific reason, I would revert them such that the example corresponds to the main.

i have done this

jubicker · 2025-08-28T12:17:59Z

cpp/examples/lct_secir_2_diseases.cpp

+                     NumInfectedSevere_2b = 2, NumInfectedCritical_2b = 1;
+    using InfState = mio::lsecir2d::InfectionState;
+    using LctState = mio::LctInfectionState<
+        InfState, 1, NumExposed_1a, NumInfectedNoSymptoms_1a, NumInfectedSymptoms_1a, NumInfectedSevere_1a,


Can you add a comment on what the numbers in the templates are or define them as variables with corresponding names before?

and i have done this

jubicker · 2025-08-28T12:28:02Z

cpp/examples/lct_secir_2_diseases.cpp

+    ScalarType tmax = 10;
+
+    // Set Parameters.
+    model.parameters.get<mio::lsecir2d::TimeExposed_a>()[0]            = 3.;


I'm not sure whether the naming of the parameters and also of infection states is the best choice. Maybe we should use longer but more self-speaking names, like TimeExposedDisease1,... and for infection states something like: InfectedNoSymptomsDisease1Naive and InfectedNoSymptomsDisease1RecoveredDisease2? Or is that too long? @annawendler what do you think?

jubicker · 2025-09-10T06:25:52Z

cpp/tests/test_lct_secir_2_diseases.cpp

+    EXPECT_NEAR(result.get_last_time(), tmax, 1e-10);
+    ScalarType sum_pop = init.sum();
+    for (Eigen::Index i = 0; i < result.get_num_time_points(); i++) {
+        EXPECT_NEAR(sum_pop, result[i].sum(), 1e-5); // check that total pop. is constant


Suggested change

EXPECT_NEAR(sum_pop, result[i].sum(), 1e-5); // check that total pop. is constant

EXPECT_NEAR(sum_pop, result[i].sum(), 1e-5); // check that total pop is constant

jubicker · 2025-09-10T06:47:54Z

cpp/tests/test_lct_secir_2_diseases.cpp

+
+    // Initialization vector for lct2d model.
+    Eigen::VectorX<ScalarType> init_lct2d = Eigen::VectorX<ScalarType>::Constant((Eigen::Index)InfState2d::Count, 0);
+    init_lct2d[0]                         = 200; // lct and lct2d use different infection states


This comment is a bit confusing because below you initialize the exactly same values for the normal LCT model

jubicker · 2025-09-10T06:58:25Z

cpp/tests/test_lct_secir_2_diseases.cpp

+
+// Run the model with more than one subcompartment for states E,I,C,H,U
+// and calculate the TimeSeries with no subcompartments from the result
+TEST(TestLCTSecir2d, testSubcompartments)


The only thing this test does is checking whether the simulate function advances until tmax which is tested already in simulateDefault. What was the intention of this test? If it was to check whether the calculate_compartment method works, then test whether the Exposed value of this timeseries at t=0 corresponds to the sum of the Exposed compartments of the other timeseries at t=0.

jubicker · 2025-09-10T07:39:23Z

docs/source/cpp/models/lsecir2d.rst

+| The LCT-SECIR-2-DISEASES model is an extension of the :doc:`model with one disease <lsecir>`.
+| The model is ODE-based and uses the Linear Chain Trick to allow for more general Erlang distributed stay times in each compartment instead of just exponentially distributed stay times induced by basic ODE-based models.
+| With the SECIR structure the model is particularly suited for pathogens with pre- or asymptomatic infection states and when severe or critical states are possible. 
+| For the two diseases :math:`a` and :math:`b` the model assumes no co-infection, a certain independence in the sense that prior infection with one disease does not affect the infection with the other disease (e.g. probability to get infected, time spend in each state, chances of recovery etc.), and perfect immunity after recovery for both diseases.


Suggested change

| For the two diseases :math:`a` and :math:`b` the model assumes no co-infection, a certain independence in the sense that prior infection with one disease does not affect the infection with the other disease (e.g. probability to get infected, time spend in each state, chances of recovery etc.), and perfect immunity after recovery for both diseases.

| For the two diseases or variants of one disease :math:`a` and :math:`b` the model assumes no co-infection, a certain independence in the sense that prior infection with one disease does not affect the infection with the other disease (e.g. probability to get infected, time spend in each state, chances of recovery etc.), and perfect immunity after recovery for both diseases.

jubicker · 2025-09-10T07:40:44Z

docs/source/cpp/models/lsecir2d.rst

+Each infection is simulated using a LCT-SECIR model. The states are labeled according to infection (first or second) and disease (:math:`a` or :math:`b`),
+so the full model is given by the combination of infections :math:`1a`, :math:`2a`, :math:`1b`, and :math:`2b`.
+
+Below is a visualization of the infection states split into LCT-states and transitions without a stratification according to socisdemographic groups.


Suggested change

Below is a visualization of the infection states split into LCT-states and transitions without a stratification according to socisdemographic groups.

Below is a visualization of the infection states split into LCT-states and transitions without a stratification according to sociodemographic groups.

annawendler

First part of review

annawendler · 2025-09-10T15:59:02Z

cpp/models/lct_secir_2_diseases/README.md

+
+## Examples
+
+A simple example can be found at [LCT2D minimal example](../../examples/lct_secir_2_diseases.cpp).


In a recent commit to the main we updated the docu of memilio. In this commit we changed the structure of the README files of the models. Now, we only refer to the docu and to an example, see https://github.com/SciCompMod/memilio/blob/main/cpp/models/lct_secir/README.md. Can you adapt this? Can you also capitalize Linear Chain Trick in the README file of the LCT SECIR model?

annawendler · 2025-09-10T16:00:30Z

cpp/models/lct_secir_2_diseases/infection_state.h

+    InfectedCritical_2bToDead_b                = 38,
+    InfectedCritical_2bToRecovered_ab          = 39,
+    Count                                      = 40
+};


Do you use the transitions somewhere? I think in the LCT SECIR model they are only used in the initializer_flows file for the initialization. If the transitions are not currently used, I would remove them to avoid confusion.

annawendler · 2025-09-10T16:00:42Z

cpp/models/lct_secir_2_diseases/parameters.h

+**********************************************************/
+
+/**
+ * @brief Average time spent in the Exposed compartment for disease a in day unit.


Suggested change

* @brief Average time spent in the Exposed compartment for disease a in day unit.

* @brief Average time spent in the Exposed compartment for disease a in day unit for each group.

Same for the other parameters.

annawendler · 2025-09-10T16:00:52Z

cpp/models/lct_secir_2_diseases/parameters.h

+
+/**
+ * @brief Average time spent in the TimeInfectedNoSymptoms before developing 
+ *  symptoms or recover for disease b in day unit.


Suggested change

* symptoms or recover for disease b in day unit.

* symptoms or recovering for disease b in day unit.

Same for other parameters.

annawendler · 2025-09-10T16:01:14Z

cpp/models/lct_secir_2_diseases/parameters.h

+};
+
+/**
+ * @brief The start day in the LCT-SECIR-2-DISEASES  model.


Suggested change

* @brief The start day in the LCT-SECIR-2-DISEASES model.

* @brief The start day in the LCT-SECIR-2-DISEASES model.

annawendler · 2025-09-10T16:03:13Z

cpp/models/lct_secir_2_diseases/parameters.h

+ */
+struct TimeExposed_a {
+    using Type = Eigen::VectorX<UncertainValue<ScalarType>>;
+    static Type get_default(size_t size = 1)


Is the default ever used? Below when constructing a Parameters object, you have num_groups as input argument. Isn't the default always overridden then?

annawendler

In a recent commit the use of floating point types was reworked to allow for automatic differentiation. This also affected the LCT-SECIR model, e.g. we now have an additional template parameter FP for the model class. Can you adapt your model accordingly so that the models are consistent? If you have any questions just let us know :)

annawendler · 2025-09-25T12:46:10Z

cpp/models/lct_secir_2_diseases/model.h

+/* 
+* Copyright (C) 2020-2025 MEmilio
+*
+* Authors: Annika Jungkalus, Lena Ploetzke


Suggested change

* Authors: Annika Jungkalus, Lena Ploetzke

* Authors: Annika Jungklaus, Lena Ploetzke

an-jung added 5 commits June 5, 2025 13:07

change year in copyright

3ef230f

lct for 2 diseases on basis of lct_secir model,

1caf0d0

started changing Infection States, Parameters etc. to reflect 2 diseases

lct for 2 diseases on basis of lct_secir model,

a75cb66

changed Infection States, Parameters etc. to reflect 2 diseases

Working LCT2D Model (no age groups) with test

e2dcde1

LCT2D Model with age groups

62945e1

an-jung linked an issue Aug 7, 2025 that may be closed by this pull request

Add LCT SECIR model with two disease strains #1294

Open

2 tasks

an-jung added loc::backend This issue concerns the C++ backend implementation. model::lct Models using the linear chain trick labels Aug 7, 2025

fix gitignore

424020a

an-jung marked this pull request as ready for review August 7, 2025 12:00

an-jung added 2 commits August 7, 2025 14:05

Merge branch 'main' into 1294-add-lct-secir-model-with-two-disease-st…

2d2ade6

…rains

header fix

72678a5

an-jung added 3 commits August 7, 2025 15:28

new test for codecov

4165ebe

readthedocs for lct 2 diseases model

f5b26f0

lct_infectionstate fix

1da5cf5

an-jung requested review from annawendler and jubicker August 22, 2025 10:06

jubicker reviewed Aug 28, 2025

View reviewed changes

an-jung added 2 commits August 29, 2025 08:59

revert changes in lct_secir exmple, explain numbers in template

d7ce14a

CI fix in test_lct_secir_2_diseases

ab75b17

jubicker reviewed Sep 10, 2025

View reviewed changes

annawendler requested changes Sep 10, 2025

View reviewed changes

annawendler reviewed Sep 25, 2025

View reviewed changes

	EXPECT_NEAR(sum_pop, result[i].sum(), 1e-5); // check that total pop. is constant
	EXPECT_NEAR(sum_pop, result[i].sum(), 1e-5); // check that total pop is constant

	\| For the two diseases :math:`a` and :math:`b` the model assumes no co-infection, a certain independence in the sense that prior infection with one disease does not affect the infection with the other disease (e.g. probability to get infected, time spend in each state, chances of recovery etc.), and perfect immunity after recovery for both diseases.
	\| For the two diseases or variants of one disease :math:`a` and :math:`b` the model assumes no co-infection, a certain independence in the sense that prior infection with one disease does not affect the infection with the other disease (e.g. probability to get infected, time spend in each state, chances of recovery etc.), and perfect immunity after recovery for both diseases.

	Below is a visualization of the infection states split into LCT-states and transitions without a stratification according to socisdemographic groups.
	Below is a visualization of the infection states split into LCT-states and transitions without a stratification according to sociodemographic groups.


		## Examples

		A simple example can be found at [LCT2D minimal example](../../examples/lct_secir_2_diseases.cpp).

	* @brief Average time spent in the Exposed compartment for disease a in day unit.
	* @brief Average time spent in the Exposed compartment for disease a in day unit for each group.

	* symptoms or recover for disease b in day unit.
	* symptoms or recovering for disease b in day unit.

	* @brief The start day in the LCT-SECIR-2-DISEASES model.
	* @brief The start day in the LCT-SECIR-2-DISEASES model.

	* Authors: Annika Jungkalus, Lena Ploetzke
	* Authors: Annika Jungklaus, Lena Ploetzke

1294 add lct secir model with two disease strains #1340

Are you sure you want to change the base?

1294 add lct secir model with two disease strains #1340

Uh oh!

Conversation

an-jung commented Aug 7, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Changes and Information

Merge Request - Guideline Checklist

Checks by code author

Checks by code reviewer(s)

Uh oh!

codecov bot commented Aug 7, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jubicker Aug 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

annawendler left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

annawendler left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

an-jung commented Aug 7, 2025 •

edited

Loading

codecov bot commented Aug 7, 2025 •

edited

Loading

jubicker Aug 28, 2025 •

edited

Loading