Add tree seq format as output to make the simulation msprime compatible.
We could either use the public API, or directly plug the simulation output using the c library tskit that is very much C++ compatible (it's very orienty ^^).
The primary goal is to report the output in this format, but eventually it could be fun to try to turn the whole simulator natively to treeseq. The smoothest way to get this done I think will be adding a new method to the world class, somewhere among the other saving methods. I have hard times to think it through though, as I still don't have a good grasp of tree sequence format, perhaps it got be integrated in the generation process because tree sequence records ancestry as well. Hmm, Is the backtrace output is functionally equvalent to the tree sequence format? Because if yes, then the easiest would be to follow exactly the same logic. Anyway.... just random thoughts I have now.
Add tree seq format as output to make the simulation msprime compatible.
We could either use the public API, or directly plug the simulation output using the c library tskit that is very much C++ compatible (it's very orienty ^^).
The primary goal is to report the output in this format, but eventually it could be fun to try to turn the whole simulator natively to treeseq. The smoothest way to get this done I think will be adding a new method to the world class, somewhere among the other saving methods. I have hard times to think it through though, as I still don't have a good grasp of tree sequence format, perhaps it got be integrated in the generation process because tree sequence records ancestry as well. Hmm, Is the backtrace output is functionally equvalent to the tree sequence format? Because if yes, then the easiest would be to follow exactly the same logic. Anyway.... just random thoughts I have now.