Convergence of sample paths produced by Kappa

[debdaspaul]

Hello,

I am wondering 

1. does Kappa output the average over n simulations or shall I have to run Kappa 1000 times to obtain 1000 sample paths?

if not, then, 

2. how to test the convergence of the sample paths produced the Kappa using Gillespie's simulation?

Many thanks 

Debdas

[David C Sterratt]

I think the answers are:

1. No. You need to run Kappa n times and analyse the results

2. Gillespie's method is guaranteed to generate unbiased samples from the chemical master equation, i.e. the probability distribution of all possible states over time. Yes, if you run more simulations, your estimates of the summary statistics of this distribution should converge.

David

--
You received this message because you are subscribed to the Google Groups "kappa-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kappa-users...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Jérôme]

Exactly,

Jérôme