Relationship between transprob() and multinomial()?

[choisy]

I am understanding correctly if I say that the transprob() function is a particular case of the multinomial() function? In the same sense that exponential() is a particular case of the gamma() distribution?

[thinhong]

I made the multinomial() to solve this problem:

The a and b are defined with transprob() and the 0.9 and 0.1 are defined with multinomial(). But now we can define it as you pointed out in another issue:

I also implemented an algorithm so that if we don't define the 0.9 and 0.1 with multinomial()

I will calculate this case by:

• S -> I: a*S move to I
• S -> V: the rest after moving to I which is S - a*S, so (S - a*S)*b move to V

But now after knowing that we can make what I made so complicated before by:

I will think carefully whether we still need the multinomial() or we can do the entire thing with transprob()

[choisy]

Yes, please do. I am not necessarily saying that you should get rid off anything here, just trying to understand the logic, which is still not too clear to me at the moment. As for the diagramme, I think you should be careful about the notation as well, see in particular the issue #22.