Change Details

DiskPlacer previously loops over the whole map each time regardless of disk size. Even though the performance cost is not That Huge, this seems to be really wasteful. This diff loops over the coordinates that are in the square with sides 2R. So placing smaller disks would be significantly faster than before. A giant new_rms_test with 3 disks placed. Second one is with the diff: create mapSized disk... 0.791s. create halfMapSized disk... 0.208s. create small disk... 0.021s. create mapSized disk... 0.771s. create halfMapSized disk... 0.188s. create small disk... 0.001s.

DiskPlacer previously loops over the whole map each time regardless of disk size. Even though the performance cost is not That Huge, this seems to be really wasteful. This diff loops over the coordinates that are in the square with sides 2R. So placing smaller disks would be significantly faster than before. A giant new_rms_test with 3 disks placed. Second one is with the diff: create mapSized disk... 0.791s. create halfMapSized disk... 0.208s. create small disk... 0.021s. create mapSized disk... 0.771s. create halfMapSized disk... 0.188s. create small disk... 0.001s. There is also a trivial comment fix.

DiskPlacer previously loops over the whole map each time regardless of disk size. Even though the performance cost is not That Huge, this seems to be really wasteful. This diff loops over the coordinates that are in the square with sides 2R. So placing smaller disks would be significantly faster than before. A giant new_rms_test with 3 disks placed. Second one is with the diff: create mapSized disk... 0.791s. create halfMapSized disk... 0.208s. create small disk... 0.021s. create mapSized disk... 0.771s. create halfMapSized disk... 0.188s. create small disk... 0.001s. There is also a trivial comment fix.