I'd like to try and do some runtime tests where I run the same scenarios over an elevation transect, and plot runtime vs the number of cohorts, and plot separate lines for a linear, exponential, and custom root distribution. I hypothesize that runtime is going to increase as you go up the elevation gradient, and have more cohorts. I also think the runtime will be lowest for linear roots because of the analytic solution instead of a spin-up for the initial conditions. It'll be medium for exponential root distribution, and highest for a custom distribution, because we use R's solver for the integrals.
I'd like to try and do some runtime tests where I run the same scenarios over an elevation transect, and plot runtime vs the number of cohorts, and plot separate lines for a linear, exponential, and custom root distribution. I hypothesize that runtime is going to increase as you go up the elevation gradient, and have more cohorts. I also think the runtime will be lowest for linear roots because of the analytic solution instead of a spin-up for the initial conditions. It'll be medium for exponential root distribution, and highest for a custom distribution, because we use R's solver for the integrals.