I am going to assume you are the author of MASON and this was agents.jl?
They make a lot of claims on how much faster they are than MASON, Netlogo, and Mesa. But in practice I am not finding that to be the case. Also they arent counting the Julia compilation step which takes an absurdly long time, and by the time that gets done similar simulations are already done, then they start the clock on their own benchmark.
Agents.jl and Mesa have the selling point of having better languages / libraries for numerical computation. But thats really a subset of msor ABM I think.
They make a lot of claims on how much faster they are than MASON, Netlogo, and Mesa. But in practice I am not finding that to be the case. Also they arent counting the Julia compilation step which takes an absurdly long time, and by the time that gets done similar simulations are already done, then they start the clock on their own benchmark.
Agents.jl and Mesa have the selling point of having better languages / libraries for numerical computation. But thats really a subset of msor ABM I think.