You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When the number of elements in a genetic map gets very large,
we take a larger performance hit.
For example, the same uniform genetic map with one region vs,
say, 2.5e5 regions, results in orders of magnitude different
performance.
The reason for this is actually simple: the latter queries
each segment, bogging the simulation down by calling the Rng
machinery. (See the implementaion operator() for GeneralizedGeneticMap.)
To reproduce, we can modify cpp_neutral_benchmark/cpp_neutral_benchmark.cc
to subdived the map into a specific number of chunks.
If that number is large, we can reproduce the performance hit.
The text was updated successfully, but these errors were encountered:
When the number of elements in a genetic map gets very large,
we take a larger performance hit.
For example, the same uniform genetic map with one region vs,
say, 2.5e5 regions, results in orders of magnitude different
performance.
The reason for this is actually simple: the latter queries
each segment, bogging the simulation down by calling the Rng
machinery. (See the implementaion operator() for GeneralizedGeneticMap.)
To reproduce, we can modify cpp_neutral_benchmark/cpp_neutral_benchmark.cc
to subdived the map into a specific number of chunks.
If that number is large, we can reproduce the performance hit.
The text was updated successfully, but these errors were encountered: