Tell ya' what:

castironpi at gmail.com castironpi at gmail.com
Sat Mar 29 05:08:20 EDT 2008


On Mar 28, 3:59 pm, castiro... at gmail.com wrote:
> I want to regression a signal.  It's NN, not necessarily, sine wave.
> What kind of numbers of harmonic simultaneous sounds are we looking
> at?  How close to an A 440 can a human make?
>
> I want recognition to alert me.  As a sine wave is coming in:
>
> >>> for t in [ ( 200* sin( x*2 ), 200* cos( x*2 ) ) for x in range( 10 ) ]: print( *t )
>
> ...
> 0.0 200.0
> 181.859485365 -83.2293673094
> -151.360499062 -130.728724173
> -55.8830996398 192.03405733
> 197.871649325 -29.1000067617
> -108.804222178 -167.814305815
> -107.3145836 168.770791746
> 198.121471139 27.3474436416
> -57.580663333 -191.531896065
> -150.197449354 132.063341649
>
> Did everyone take the course on computer architecture?

E.g. sum() is not an operation you do in world operations.  (i
question that 'walk' is.)  you don't ( add ) drop to puddle, you move
drop.  we've all got graphics sub-primitives but no modelling ones.
you don't want to decode sense cell inputs, do you?  i'd say in
complexity, one cell adds up to an entire input line-trail (wake).  I
would be proposing multiplexcating every CPU unit onto every input
gate: off keystrokes and camera photovoltaics.  (Light takes six -
complexity- muscles, plus squinting, while not six calory-muscles.)
So, taking outputs to be common combinations of inputs, unless you
want symbological inputs to map to anything constructive.  I think we
design architecture utility-first.  Ste p1.  Big box?  Would
calculations include utility transponders (tabulation)?  Time-scale is
discretized.  Can you be more productive continuously?  Do you need to
recognize users discretely, each an own interaction on file, due to
discrete measures (discrete times)?  It's more robotic than our
interface with dogs.  Does anyone have names for syllables?
Superlatives are always theoretical.  What can we learn from an
executed time-space translation?  Point is, sum() is not real, add is.

I think event-driven stuff.  It's non-destructive.  Are there other
tables after D.J.?  He's predictable alright.

Broadcast dropoff "sighted" (recognized).  Is that bandwidth or baud?
How quickly can you get total human decible out of a room?



More information about the Python-list mailing list