thx for the feedback, you might be right. I can post the graph on Monday, as I do not have access to it right now (out of office at the moment).
In short, I am trying to explore different placement combinations for 15 buildings in a large plot. The script calculates the height of each one, based on the free area around it. In the end, it calculates the distances between buildings to check if they meet the regulations. So, I have divided the plot in 81 different points, which represent 81 possible location for each of the 15 buildings.
You can clearly see, that even with many additional limitations, the study results in millions of options. I suppose this is the problem. The first ‘broken’ test was a result of a cross product study. I then tried two other tests with the optimize option. Even though the total number of results for them was limited down to 10 and 30, they also brought the tool to a crash. Finally, ‘Randomise’ tests worked perfectly and managed to produce 1000 and even 5000 options within minutes.
I will then use some of the options they produced as a basis to limit the parameters for another cross product study with much more limited options. That’s why I thought the problem was the extremely large amount of options in the first cross product test