Monte Carlo Simulations and Hierarchical Fail-up

Latona algorthms can generate movement through random processes. Sometimes these processes run into problems. Maybe all the tracks ended in the corner and cannot move as described in the next phrase of the music. As a result, we created a hierarchal structure for trying new things within limits and failing up to previous versions of the run.  Building on this, we also created a system for optimization where we define criteria we are looking for in the pattern. If we have a successful run, this run is ranked in terms of the criteria we are targeting. The longer we let the algorthmm run, the better is our target value. The better the target value, the more interesting the pattern is likely to be artistically.  As an example, we have a optimation for X and Y axis crosses. We can optimize so that patterns are generated that maximize the number of times the audience see object dancers cross each other in the site plain.

Latona algorthms can generate movement through random processes. Sometimes these processes run into problems. Maybe all the tracks ended in the corner and cannot move as described in the next phrase of the music. As a result, we created a hierarchal structure for trying new things within limits and failing up to previous versions of the run.

Building on this, we also created a system for optimization where we define criteria we are looking for in the pattern. If we have a successful run, this run is ranked in terms of the criteria we are targeting. The longer we let the algorthmm run, the better is our target value. The better the target value, the more interesting the pattern is likely to be artistically.

As an example, we have a optimation for X and Y axis crosses. We can optimize so that patterns are generated that maximize the number of times the audience see object dancers cross each other in the site plain.