Okay! the extension I wanna talk about
today is the network extension.
We already played around
with a little bit in unit 4.
But I mentioned a little bit
about how you could add
a calculation that would be much
harder to calculate in
that logo but putting
then into extensions.And
this model really exemplifies that.
That's why I wanna this talk about it
So, this is another viral
marking model of... you know...
it's ..in neuro we cal it diffusion
information. This is taking into the
the point - "I wanna find who is
the best person to see... what particular
messages if I want that message
to diffuse quickly thorugh .. the system.
and I have a budget, so, I can determine
how many different people
I can see.. I have a different
types of networks named in
preferential attachment like we saw
before. But now,I am also gonna have
a setting that isn't random or based upon
betweenness and betweenness is exactly
what I am talking about when
I am saying sometime it's better
to have a particular calculation
umm.. be written in another language
and then just have netlogo
use an extension to call it.
And that's because betweenness is a
really complex calculation.
What betweenness is calculating ..
is it's calculating how often a
node appears on the shortest path between
any of the other two nodes in
the entire network.right..
that involves comparing every node
in the network, finding each shortest
path to every other node in
the network and then figuring out. which
..how many ...and what nodes are on
those paths. As you might imagine
as we increase the size of the
network this calculation absolutely
explodes and in many cases
we don't waana put code like that
directly into a net logo model.right...
Instead, it's better to, spend some
time, develop a Java vesrion of
that.. another language version of that
particular code and then have net logo
simply call that version on the particula
err.. umm.. dataset that we are looking
at. That's exactly what the model
does.So, what it does is set up network