Plots Needed for Simulation

Plots needed (suggestions…edit at will)

[we need to discus how to deal with the wall….if and when to stop a simulation…

1. a) mean no of agents at end of simulation vs ID (no one uses word “turtles”) for random initial conditions (seed is number)

b) no of agents at end vs ID for a fixed random initial condition (seed is zero)

c) no of agents at end vs ID for another fixed random initial condition.

d) no of agents at end vs ID for a bunch of simulations with different initial random arrangements, a plot you did already

Maybe we can put the four plots in one big figure.

2. plot of forces for two agents of same mass to calibrate the Minimum Force Cutoff

3. 3D Plot: number of groups vs elapsed time vs ID for a standard arrangement except for varying ID. Maybe on the same figure plot also the number of individual agents.

4. suggested (edit at will) standard arrangement is : 200 agents, MFC 10^-5, ID = 100, g 7 E-5, DistExp 2, Random Init Spat dist, mmean 0, SD 0.4; Mass Dist Random, mean 30, sd 5, dynamic dt on.

5. 3D plots:

1. total mass (agent plus groups) as a function of elapsed time and distance from the origin

2. total mass (agent plus groups) as a function of elapsed time and angle from the x axis (0 to 360 or 2Pi)

6. 3D Plot: time at which number of groups is a maximum number of groups vs actual number of groups at maximum plotted as a function of ID

7. Plot “Number of agents at end”/”Initial number of agent” versus ID, do for 100 agents and for 200 agents and 50 agents.

Experiments to do:

1. put all agents in one quadrant, only attractive forces. End up with one final group. Homogeneous culture.

2. do a standard simulation (100 agents, ID = 75, MFC = 5 x 10^-5, CoalescenceRadius=0.2, (g =7 x 10^-5 same for all runs), normal distribution spatially at (0,0), SD of 0.5 or 50, random mass dist with mean 50 and SD of of 20, dynamic dt. ) (The standard means representative, so we can infer the overall effect of a modification from a few representative runs) with a fixed seed to compare the results for R with an exponent of 3, 2 and 1, or to the .5 power, to see if there is a fundamental difference or just one of degree. Save final configuration plots. Also could plot the number of agents at end/initial number of agents versus ID for each exponent.

3. Do expt with injected agents with different mass, put one big mass in one quadrant and see what happens.

4. Do expt with one big mass in one quadrant and another big mass in another quadrant

5. Whatever looks interesting