Is there a way to simulate a simple SIR model to Iris

[ripatti]

I wonder if there is a way to simulate a simple SIR model with Iris?

The model code would like like this

%% Definitions
!transition-variables 
  S, I, R

!parameters 
  pi_r = 1/14; % per-period recovery rate
  pi = 2.5/14; % =R0*pi_r


%% Model block
!transition_equations
  S =# S{-1} - pi*I{-1}*S{-1};
  I =# I{-1} + pi*I{-1}*S{-1} - pi_r*I{-1} ;
  R =# R{-1} + pi_r*I{-1};

And the code to run

m = Model.fromFile('SIR.model','linear', false,'comment','SIR model');
len = 150;
idb = struct();
idb.I = 0.001;
idb.S = 1-idb.I;
idb.R = 1 - idb.S - idb.I;
R0 = m.pi/m.pi_r;
db = simulate(m, idb,1:len,'method','period');

But the last option does not work because we have not solved the model. There is no steady state either.

[jaromir-benes]

Hi - the first order solution is used by default to start iterations when method=stacked or method=period. Use startIterations="data" and supply some initial values (time series on the simul range) for all variables in the input databank. Best J

…

On Mon, Aug 8, 2022, 1:56 PM Antti Ripatti ***@***.***> wrote: I wonder if there is a way to simulate a simple SIR model with Iris? The model code would like like this %% Definitions !transition-variables S, I, R !parameters pi_r = 1/14; % per-period recovery rate pi = 2.5/14; % =R0*pi_r %% Model block !transition_equations S =# S{-1} - pi*I{-1}*S{-1}; I =# I{-1} + pi*I{-1}*S{-1} - pi_r*I{-1} ; R =# R{-1} + pi_r*I{-1}; And the code to run m = Model.fromFile('SIR.model','linear', false,'comment','SIR model'); len = 150; idb = struct(); idb.I = 0.001; idb.S = 1-idb.I; idb.R = 1 - idb.S - idb.I; R0 = m.pi/m.pi_r; db = simulate(m, idb,1:len,'method','period'); But the last option does not work because we have not solved the model. There is no steady state either. — Reply to this email directly, view it on GitHub <#332>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGCVKKX2RT4MTLYE2TWNWUDVYDYYLANCNFSM5543IQCQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[ripatti]

Thanks a lot!

Although these SIR models are easy to solve and can be programmed easily, Iris provides the wide set of other tools (consider, eg, different parameterizations) to experiment with the model. Therefore, I try to use Iris in performing any model simulation or development.

This modification works:

idb.I = Series(1:len,0.001.*ones(1,len));
.
.
.
db = simulate(m, idb,2:len,'method','period', 'startIterationsfrom','data');

the range has to start from 2 to get the initial state information.

[ripatti]

It seems that assigning multiple parameter values does not work with this model object?
eg

p.pi = [0.1 0.2];
assign(m,p); 

does not work. (latest version of Iris)

[jaromir-benes]

m = alter(m, 2); m.pi = [ ...];

…

On Tue, Aug 9, 2022, 3:26 PM Antti Ripatti ***@***.***> wrote: It seems that assigning multiple parameter values does not work with this model object? eg p.pi = [0.1 0.2]; assign(m,p); does not work. (latest version of Iris) — Reply to this email directly, view it on GitHub <#332 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGCVKKVWJF3HGECFDNZSCVDVYJL7VANCNFSM5543IQCQ> . You are receiving this because you commented.Message ID: <IRIS-Solutions-Team/IRIS-Toolbox/repo-discussions/332/comments/3357997@ github.com>