Riffs & Rotes & A061396 & A062504?
Marc LeBrun
mlb at well.com
Mon Jun 25 08:27:45 CEST 2001
It might be worth noting that these trees themselves can be encoded as
numbers, thus forming a kind of Godelization of the integers.
Let me sketch an example for one of the notations I think I saw go by here
recently:
A given number corresponds to an infinite sequence of finite trees, all but
a finite number of which are empty. We wish to transcribe these into an
infinite string of bits, all but a finite number of which are 0.
First we assign each tree (ie each prime) its own infinite subset of
bits. To the tree rooted at 2 we assign the even bit positions, to the
tree rooted at 3 we assign every other of the remaining positions (ie the
bit positions congruent to 1 mod 4), to the tree rooted at 5 we interleave
again (ie 3 mod 8) and so on, sort of giving each successive prime "half"
the remaining positions.
Next we transcribe the tree rooted at a given prime into these
positions. If the tree is empty, o, we set all the bits to 0. If the tree
is one, (o), we set the least significant bit to a 1 and the rest to 0.
For the more complex trees we just recurse (I think the exact details of
the recursion you pick gives you a choice of the different structures
discussed in this thread, but I'm not sure I understand all that yet!).
Anyway, the point is that this allows you to map the integers through
the trees and back into numbers, which opens up many new possibilities.
We needn't be limited to just taking censi of various populations (eg trees
of various types). We can also directly access the images of single
specific concrete cases numerically.
For example if f is the mapping (either as we've constructed above, or some
other way) then there's sequences such as f(n+1) and tables of operations
such as f(x+y) or f(xy), and for all of these there's also their images
under the inverse of f. All of which might be entered in the EIS, I imagine.
And so on...
More information about the SeqFan
mailing list