Darwinbots Forum

Code center => Bugs and fixes => Topic started by: shvarz on January 10, 2007, 02:54:59 PM

Title: Mutation rates
Post by: shvarz on January 10, 2007, 02:54:59 PM
Not sure if it's a bug, but I see skewed ratio for types of mutations.  I started with (almost)zero-bot that had just reproduction gene and a bunch of random nonsense and after a while looking at my bots I see in mutation details the following:
Quote
Minor Deletion deleted a run of 1 bps at position 15 during cycle 6582423
Major Deletion deleted a run of 4 bps at position 20 during cycle 6578974
Minor Deletion deleted a run of 1 bps at position 13 during cycle 6553210
Minor Deletion deleted a run of 1 bps at position 14 during cycle 6550116
Major Deletion deleted a run of 4 bps at position 20 during cycle 6547061
Major Deletion deleted a run of 3 bps at position 14 during cycle 6535936
Major Deletion deleted a run of 3 bps at position 18 during cycle 6513640
Delta mutations changed Point Mutation from 1 in 5000 to 1 in 4562.133
Minor Deletion deleted a run of 1 bps at position 35 during cycle 6493622
Minor Deletion deleted a run of 1 bps at position 8 during cycle 6483567
Minor Deletion deleted a run of 1 bps at position 4 during cycle 6469072
Major Deletion deleted a run of 3 bps at position 15 during cycle 6468389
Point Mutation changed number from 8 to 16 at position 33 during cycle 6468055
Minor Deletion deleted a run of 1 bps at position 26 during cycle 6467012
Copy Error changed the number: .robage to the flow command: cond at position 23 during cycle 6460197
Major Deletion deleted a run of 5 bps at position 15 during cycle 6438109
Major Deletion deleted a run of 3 bps at position 12 during cycle 6408242
Delta mutations changed Copy Error from 1 in 5000 to 1 in 4660.702
Major Deletion deleted a run of 6 bps at position 33 during cycle 6382245

As you can see, mostly deletions.  I would understand that if the costs of maintaining DNA were high, but they were actually set pretty low, at around 0.000001.   Weird, huh?
Title: Mutation rates
Post by: EricL on January 10, 2007, 03:00:55 PM
What mutation probabilities are you using?
Title: Mutation rates
Post by: shvarz on January 10, 2007, 04:45:37 PM
They used to be at default, but I recently noticed that the multiplier is at .125.  I just picked a random bot and its mutation frequencies are:
Point: 4562
Delta 5000
Copy error: 4661
Insertion 5000
Reversal: 5000
Minor deletion 5000
Major deletion 5000

So, nothing out of the ordinary.  And I double-checked on mutation details, they are

Major Deletion deleted a run of 1 bps at position 21 during cycle 6621251
Minor Deletion deleted a run of 1 bps at position 15 during cycle 6582423
Major Deletion deleted a run of 4 bps at position 20 during cycle 6578974
Minor Deletion deleted a run of 1 bps at position 13 during cycle 6553210
Minor Deletion deleted a run of 1 bps at position 14 during cycle 6550116
Major Deletion deleted a run of 4 bps at position 20 during cycle 6547061
Major Deletion deleted a run of 3 bps at position 14 during cycle 6535936
Major Deletion deleted a run of 3 bps at position 18 during cycle 6513640
Delta mutations changed Point Mutation from 1 in 5000 to 1 in 4562.133
Minor Deletion deleted a run of 1 bps at position 35 during cycle 6493622
Minor Deletion deleted a run of 1 bps at position 8 during cycle 6483567
Minor Deletion deleted a run of 1 bps at position 4 during cycle 6469072
Major Deletion deleted a run of 3 bps at position 15 during cycle 6468389
Point Mutation changed number from 8 to 16 at position 33 during cycle 6468055
Minor Deletion deleted a run of 1 bps at position 26 during cycle 6467012
Copy Error changed the number: .robage to the flow command: cond at position 23 during cycle 6460197
Major Deletion deleted a run of 5 bps at position 15 during cycle 6438109
Major Deletion deleted a run of 3 bps at position 12 during cycle 6408242
Delta mutations changed Copy Error from 1 in 5000 to 1 in 4660.702
Major Deletion deleted a run of 6 bps at position 33 during cycle 6382245

Do you want the sim?
Title: Mutation rates
Post by: EricL on January 10, 2007, 05:04:39 PM
Quote from: shvarz
Do you want the sim?
Let me take a look at the code first and see if I can spot any biases.  Will probably be a few days.