| Name |
Update |
Plans for tomorrow |
| Alex M |
I replaced AraSim on Eliot's user with the one we have on the shared space since I wasn't able to compile his yesterday. It compiled correctly and I ran the loop but got the same results from last week (effective volumes that differed from the ones we got in the original and database test runs). I found one potential error in the seed we were passing to AraSim from the loop, so I'm currently running with that fix. I'm hopeful, but not convinved it will fix things because the setup.txt files on each version of AraSim looked the same before I made this fix. Regardless of whether or not it fixes the problem, it was a problem so it was important to run with that fix. This also gives me another thing to add to the list of changes to make before doing a real run. You can find the list below. |
I'll continue this run for 4-5 generations if it looks like it's working just to confirm that we're getting agreement with the previous runs. If it agrees, I can turn my attention to either merging or doing a real run. I can do a real run without merging, but it depends on if Amy wants me to merge first or not. Eliot gave me his account info for OSC and github so I should be able to merge correctly (hopefully I can meet with Julie over this so we can make sure we don't have a problem) but it might take me some time to resolve the conflicts. |
| Alex P |
|
|
| Eliot |
|
|
| Leo |
|
|
| Evelyn |
|
|
| Ryan |
Worked on rewriting certain parts of the paperclips code so that it runs better. |
Fix the bugs in the code. |
| Ben |
|
|
| Ethan |
|
|
Here's the list of changes to make when doing a real run (undo some of these when doing a test run):
1. Change the separation distance from 0.05 cm to 2-3 cm.
2. Remove the seed in AraSim by changing the setup.txt file so that Random_Mode = 1 instead of 0.
3. Add noise to AraSim (unless Amy says not to?) by changing the setup.txt file so that Trig_Analysis_Mode = 0 (or whatever the setting is--see the comments).
4. Make sure all of the frequency lists are the same! I think this has largely been resolved, but basically there was a time when the frequencies we were using differed from those used by AraSim for the actual bicone. We changed our lists to agree with the AraSim one, but I think there could be a few lists floaitng around (potentially just commented out) that I'm going to want to take our or correct.
5. Remove the seeds from the roulette algortihm. To do this, search for where "seed" is used and comment out the one that has 1 as an argument and comment in the one that uses time(0). Also, do the same for srand().
6. Fix the exponent in the calculation of the standard deviation of the genes in the roulette algorithm. Right now it has an integer division error, so the exponent becomes 0 leading to all the sigmas being 1. This is an easy fix--I'll just change 1/2 to 1.0/2.0.
I haven't made these changes yet because we are testing and want to confirm that we get the same results from the asymmetric version with the original and database versions. Since those didn't have these fixes, I haven't put them into the asymmetric version yet either. |