count result is not what is expected

[mhrlittle]

Many thanks for creating this package. I've noticed the count produced by run_model to not be what would be expected. In the introduction example the initial count puts 1000 in state A with a transition probability of 0.1 to state B. Therefore, I would expect a count of 900 in the first cycle but the bottom the example shows a count of 950. I'm not sure if this is intended or not. Could you let me know how the count of 950 is arrived at.

[jrdnmdhl]

This is a result of the half-cycle correction method, which defaults to life-table (i.e. averaging the counts from adjacent times).

[mhrlittle]

Thanks