[ExI] What if humans were twice as intelligent?

[rex]

Anders Sandberg <anders at aleph.se> [2013-10-21 15:11]:
>On 21/10/2013 10:44, rex wrote:
>>I suspect it's the plot(ff) statement that is causing the problem. No
>>plot is produced.
>
>Yes, I would suspect that is the risky part. Octave has never had any 
>native plot functions. And my code certainly doesn't touch any fonts.

It was working -- the plot window was apparently under another window,
and the warnings were only warnings.

Your code is very compact and clean. :)

>>Rather than try to track down the plot() incompatibility, I'm going
>>to try to translate your code to R. There may be some sticky spots,
>>and I hope you don't mind my asking about them.
>
>Happy to explain. I ought to learn R too - Matlab is getting a bit 
>old anyway. R and SciPy look like the future.

I've a version that works in R now, but I only partially understand
it. In particular, I don't understand how selectrand() works. I
can see what it does, but it's not clear to me why it works. 

Once I understand the code better I'm going to look into vectorizing
some of it. 

I've attached a plot of 3 runs with a fitness change at gen 50.

I stumbled across the Prestion-Campbell paper a few years ago and
was so surprised I put up a web page about it.

http://www.nosyntax.net/cfwiki/index.php/Differential_Breeding
http://www.nosyntax.net/cfwiki/index.php/Preston_Campbell_1993

BTW, Python+NumPy+SciPy+Matplotlib was my main language for several
years, but R seduced me, and about 95% of my code is written in R now.

-rex
-- 
"In the beginning there was nothing, which exploded." -- Terry Pratchett

-------------- next part --------------
A non-text attachment was scrubbed...
Name: anders-0.01.png
Type: image/png
Size: 50281 bytes
Desc: not available
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20131021/ca9e353e/attachment.png>