allow repeating the results of MultipleRuns for models that use random numbers

[pedro-andrade-inpe]

Think about how to allow repeating the results of MultipleRuns for models that use random numbers. It is interesting for published papers, in order to guarantee that the results of the paper could be reproduced exactly as published. To handle this, the random seed should be passed from the result of one simulation to the input of another simulation. Note that in this case the model must have seed as argument.

[agojunior]

(all strategies have a "repeats variable" implemented as in the last release 0.1.2) Since all strategies can repeat the simulation, which will generate repeated results, can I close this issue?

[agojunior]

Oh wait, sorry I had misunderstood the issue. If the TerraME random seed and the model random seed are equal in both models, they will have the same results as far as I know. I will create tests to be sure.