| 1 |
Yes, this is partially considered by generalized norms for the energy vector (in your example: infinity norm) explained in the [url=http://zero-k.info/Forum/Thread/11335#114110]first paragraph of this post[/url] and the [url=http://zero-k.info/Forum/Thread/7478#114117]first two paragraphs of this post[/url]. Further calculations can be done on this. Note that finding the ideal solution for concave gamma with s-normed energy vectors requires equality of derivations with respect to (e_i to the power of s).
|
1 |
Yes, this is partially considered by generalized norms for the energy vector (in your example: infinity norm) explained in the [url=http://zero-k.info/Forum/Thread/11335#114110]first paragraph of this post[/url] and the [url=http://zero-k.info/Forum/Thread/7478#114117]first two paragraphs of this post[/url]. Further calculations can be done on this. Note that finding the ideal solution for concave gamma with s-normed energy vectors requires equality of derivations with respect to (e_i to the power of s).
|
| 2 |
\n
|
2 |
\n
|
| 3 |
For
the
infinity
norm
E=||evector||_infinity
however,
math
becomes
much
easier,
maybe
too
easy.
Then
every
mex
would
get
the
whole
grid
energy
and
no
optimization
would
be
needed.
But
then
gamma
would
have
be
changed
and
also
depend
on
m=||mvector||_1
(
the
grid's
total
base
metal
rate)
and
be
not
only
concave
in
e_i,
but
also
in
m
so
that
connecting
doesn't
become
too
powerfull.
I
have
to
think
about
grid
connection
monotony.
|
3 |
For
the
infinity
norm
E=||evector||_infinity
however,
math
becomes
much
easier,
maybe
too
easy.
Then
every
mex
would
get
the
whole
grid
energy
and
no
optimization
would
be
needed.
But
then
gamma
would
have
to
be
changed
and
also
depend
on
m=||mvector||_1
(
the
grid's
total
base
metal
rate)
and
be
not
only
concave
in
e_i,
but
also
in
m
so
that
connecting
doesn't
become
too
powerfull.
I
have
to
think
about
grid
connection
monotony.
|
| 4 |
\n
|
4 |
\n
|
| 5 |
Maybe you want to try it with other norms. You don't even need to change the energy norm. You can also make gamma concave in ||mvector||_s.
|
5 |
Maybe you want to try it with other norms. You don't even need to change the energy norm. You can also make gamma concave in ||mvector||_s.
|
| 6 |
\n
|
6 |
\n
|
| 7 |
But implementing this will be more work than 1 line for symmetrization.
|
7 |
But implementing this will be more work than 1 line for symmetrization.
|
Overdrive, Underdrive and Quantum Mechanics