1003 implement ODE SEIR metapopulation model by charlie0614 · Pull Request #1273 · SciCompMod/memilio

charlie0614 · 2025-05-12T12:24:47Z

Changes and Information

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

implement a different mobility model without the graph approach and only using ODEs

Replacement for #1128

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 for new functionality has been added (Doxygen in the code and Markdown files if necessary)
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).

…s-integrated-into-the-odes

…function

pycode/memilio-epidata/memilio/epidata/getNRWCounties.py

HenrZu

Great work :) Only some comments, and then we can merge :)

pycode/memilio-epidata/memilio/epidata/getNRWCounties.py

cpp/models/ode_seir_metapop/regions.h

cpp/models/ode_seir_metapop/parameters.h

cpp/models/ode_seir_metapop/model.h

cpp/models/ode_seir_metapop/CMakeLists.txt

HenrZu

Sorry for the delay. Let's get this done soon. Overall, it looks very good and is an interesting addition. However, we still need to work on the presentation and documentation :)

HenrZu · 2026-03-06T12:31:05Z

docs/source/cpp/ode_metapop.rst

+As the model is given as a large system of ODEs, the simulation can be performed by solving the system using a standard ODE solver. The model is build on the same setup as the simpler ODE-models, so we refer to :doc:`ode` for more details.
+
+
+How to: Set up and run the ODE metapopulation model


Suggested change

How to: Set up and run the ODE metapopulation model

How to: Set up and run the ODE-based metapopulation model

HenrZu · 2026-03-06T12:41:17Z

cpp/examples/ode_seir_metapop.cpp

+
+    auto result = simulate(t0, tmax, dt, model);
+
+    result.print_table();


What is the insight of this output? Maaybe you can output smth more interesting, or at least label the cols.

HenrZu · 2026-03-06T12:42:17Z

cpp/models/ode_seir_metapop/model.h

+                    const size_t Rjn = population.get_flat_index({Region(region_n), AgeGroup(age_j), Recovered});
+                    const size_t Sjn = population.get_flat_index({Region(region_n), AgeGroup(age_j), Susceptible});
+
+                    const double Nj_inv = 1.0 / (pop[Sjn] + pop[Ejn] + pop[Ijn] + pop[Rjn]);


similar to the model.h in the locla seir model, we should check if the population is > Limits::zero_tolerance())

HenrZu · 2026-03-06T12:46:32Z

cpp/models/ode_seir_metapop/model.h

+                    const size_t Sjn = population.get_flat_index({Region(region_n), AgeGroup(age_j), Susceptible});
+
+                    const double Nj_inv = 1.0 / (pop[Sjn] + pop[Ejn] + pop[Ijn] + pop[Rjn]);
+                    double coeffStoE    = 0.5 *


This might potentially be a bigger problem. The coeffStoE is independent of the region. Is it possible to choose a region specific damping factor? At the moment, we just have one global contact matrix right?
I think we can solve this using the ContactMatrixGroup and the indices are the regions then.

HenrZu · 2026-03-06T12:49:34Z

cpp/models/ode_seir_metapop/model.h

+{
+
+/********************
+* define the model *


there is no readme file in this dir.
I think it is important to note there, that is just an implementation of a implicit mobility scheme and that we also have a generic mobility scheme with explicit mobility (the graph-ode) for each of the available ODE models.

HenrZu · 2026-03-06T13:03:41Z

docs/source/cpp/deterministic_metapop.rst

+Simulation
+----------
+
+The simulation runs in discrete time steps using standard numerical integration schemes. Several schemes including adaptive step size methods are available, see here.


HenrZu · 2026-03-06T13:04:00Z

docs/source/cpp/deterministic_metapop.rst

+How to: Set up and run a simulation of the deterministic metapopulation model
+----------------------------------------------------------------------------
+
+To set up a simulation with the deterministic SEIR metapopulation model you need to initialize the model with the desired number of regions and age groups, e.g., 3 regions and 1 age group:


not sure about the deterministic here.

HenrZu · 2026-03-06T13:05:19Z

docs/source/cpp/deterministic_metapop.rst

+    model.parameters.template get<mio::oseirmetapop::ContactPatterns<>>()
+    .get_cont_freq_mat()[0]
+    .get_baseline()
+    .setConstant(2.7);
+
+    model.parameters.set<mio::oseirmetapop::TimeExposed<>>(3.335);
+    model.parameters.set<mio::oseirmetapop::TimeInfected<>>(8.097612257);
+    model.parameters.set<mio::oseirmetapop::TransmissionProbabilityOnContact<>>(0.07333);


these parameters are now set global. It would be more interesting to allow them to be set locally.

HenrZu · 2026-03-06T13:05:42Z

docs/source/cpp/deterministic_metapop.rst

+    model.parameters.set<mio::oseirmetapop::TimeInfected<>>(8.097612257);
+    model.parameters.set<mio::oseirmetapop::TransmissionProbabilityOnContact<>>(0.07333);
+
+Construct an ``Eigen::MatrixXd`` of size :math:`n_{regions} \times n_{regions}` which describes the fraction of individuals commuting from one region to another. The matrix should satify the sum of each row equal to 1.0, e.g.:


Theres not real "commuting from one region to another" in this model

HenrZu · 2026-03-06T13:06:52Z

docs/source/cpp/ode_metapop.rst

+ODE-based metapopulation model
+================================
+
+Introduction


just reading the file name, i would assume that this also includes the graph-ode. And whats the difference between the determinnistic_metapop and ode_metapop files?

charlie0614 added 30 commits April 11, 2024 16:29

add model with Populations with Regions

2d4aec1

working SIR model without infections during commuting

9416256

add infections during commuting and age groups

12c9c09

read in mobility data for new model

2b883b6

only add edges if weight is big enough

d49a048

correct age resolution

56c4dd4

add tests

0efa8db

add seir version of mobility model

05e3717

small adjustments sir version

f475e3c

Merge branch 'main' into 1003-implement-model-in-which-the-mobility-i…

68632c5

…s-integrated-into-the-odes

fixes after merge

f19bcf6

Renaming

6ace8cf

Renaming

232f705

Set up examples for comparison

6045e1f

changes for plots and corrections

dfdac2d

add improved model

16a07da

changes for comparing simulations

23196bc

Merge branch 'main' into 1003-implement-model-in-which-the-mobility-i…

4193cd6

…s-integrated-into-the-odes

adjust plot file

bfc2562

reformat py file

c6b56e9

py file

5e51d69

py file

0eff4d1

change commuting strengths to contact matrix and implement indicator …

3d3b007

…function

return to factor 0.5

9f0fba7

read in data and time measurement 400 counties

8908d01

small corrections model

06950d9

corrections again

aad78a9

change integrator in graph model

2bdb791

fix bug with age groups and discard transmissions during commuting

3c6f630

adapt structure of old model

aa6b393

change error handling to include

172c674

charlie0614 requested a review from HenrZu May 14, 2025 11:38

HenrZu reviewed Jun 4, 2025

View reviewed changes

pycode/memilio-epidata/memilio/epidata/getNRWCounties.py Outdated Show resolved Hide resolved

charlie0614 added 8 commits June 11, 2025 07:48

correct docu for contact matrix

8b85683

use existing region struct

4619fac

adjust to new data structure

e7c4def

Merge branch 'main' into 1003-implement-ode-seir-metapopulation-model

f788f7f

simplify metapop

6c75e60

adjust test

371f12a

fix bug

6649d49

go back to previous version

58856cd

mknaranja changed the title ~~1003 implement ode seir metapopulation model~~ 1003 implement ODE SEIR metapopulation model Nov 13, 2025

charlie0614 added 11 commits December 1, 2025 16:16

readjust test, write some documentation

f43dba2

Merge branch 'main' into 1003-implement-ode-seir-metapopulation-model

043de91

specify templates

b5f1df6

fix bug

3aaecd9

forgot SimulationTime

8185db7

forgot FP

0b812e8

still forgot SimulationTime

3e57b61

fix bugs in tests

7861885

Still forgot ScalarType

e8d8e30

readd the tests for parameter checks

fd60e1b

add license header

fdbf3a7

charlie0614 requested a review from HenrZu December 1, 2025 18:40

Merge branch 'main' into 1003-implement-ode-seir-metapopulation-model

7edc608

HenrZu requested changes Jan 5, 2026

View reviewed changes

charlie0614 added 2 commits January 7, 2026 15:43

apply suggestions from review

8e122aa

add serialize and deserialize

d1296e3

charlie0614 requested a review from HenrZu January 8, 2026 08:22

HenrZu requested changes Mar 6, 2026

View reviewed changes

		As the model is given as a large system of ODEs, the simulation can be performed by solving the system using a standard ODE solver. The model is build on the same setup as the simpler ODE-models, so we refer to :doc:`ode` for more details.


		How to: Set up and run the ODE metapopulation model

	How to: Set up and run the ODE metapopulation model
	How to: Set up and run the ODE-based metapopulation model


		auto result = simulate(t0, tmax, dt, model);

		result.print_table();

Conversation

charlie0614 commented May 12, 2025 • edited by HenrZu 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!

Uh oh!

HenrZu left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

HenrZu 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!

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!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

charlie0614 commented May 12, 2025 •

edited by HenrZu

Loading