Monday, October 11, 2010

10/11: Class Notes

We ran the simulator on the 3 light blank map. Most layers
G6: 297
G5: 296
G4: 528
G3: 322
G2: 337
G1: 368
G7: 292

What did we learn from this exercise?
David: When we have limited lights, make sure to cover the area as much as possible, with slower cycles
Nitin: The diagonals are longer than the side, so it makes more sense to put the lights along a diagonal
Neerja: We don't really have enough evidence here to make a lot of conclusions
Andrew: How do we approach timing? We need to make sure that they get all the way across a flow before we turn off the lights. We determined 11.5 was an appropriate timing
Neha: It's most efficient to leave the lights on longer (we looked at 20-30 seconds)
Jesse: Our ideal timing was 30 seconds for a cycle - 18 on, 12 off
Daniel: We used 20/20, but it will be different for different maps and placements

KR: How will you determine these parameters?
Zach: I think it will be constant between maps
EPK: It varies by how much overlap there is between lights

We did the same with 10 lights:
G1: 50
G2: 53
G3: 140
G4: 130
G5: 54
G6: 65
G7: 84

What did we learn from 10 lights, no barriers?
Pilunchana: Only cover 50% of the space, don't worry about getting more.
Daniel: Only invest a light to do a rapid collection if you have a lot of lights
John: Get more than 50% of the mosquitos to the collector as quickly as possible, it's better to have 50.2% than 49.8%


With g2board2, 10 lights:
G1: 181
G2: 141
G3: 963
G4: 899
G6: 125
G7: 132


Deliverable for Weds:
Let's see automatic players. We'll challenge it with one from the collection that we haven't played a lot yet.

No comments:

Post a Comment