Readme Simplex Method
FIRST:
sorry for my english!
2nd: i'm
sorry, i don't know why, but if i try to print a pdf docs from this one, pdf
haven't the equations, so i export it in others popular and free formats: .odt
(that is more free than .pdf), and html (that is more popular than pdf).
License
Gnu
free documentation license.
Version 0.3
Input required
A tableau
matrix and a base.
If you
have the problem of linear programming 
in standard canonical form:
where:
The tableau
matrix is:
The base
is:
An ordered
set of index that imply an inner identity matrix of tableau.
Output
The
program don't do smart things (at least in this version), it do some checks for
canonical form and after these checks it try to obtain optimal tableau from
input. Optimal tableau is the first tableau with 
even if there are others solutions (for
example in a unlimited ).
Moreover
it put on stack the partial results as tableau.
After it
try to obtain others optimal tableaus using 
, but it can find all solutions (if there are
more than two ) see the code for details.
Then it
put on stack the others optimal tableaus (if any) and a matrix where each row
is an optimal canonical base.
So you can
use it only for boring operations, not for "smart" steps.
Example
Inputs
(tableau & base)
Outputs
(this is
an edit view)
On 3rd
level, there are partial tableaus and the first optimal tableau (as list)
On 2nd
level there are the others optimal tableaus (as list)
On 1st
level there is the matrix where each row of it is an optimal canonical base.
All 3
levels are show below as list (the program don't do it automatically)
As you can
see, the optimal tableau (2nd matrix in the first sublist) allow a lot of
others optimal tableaus, using . The program, because isn't simple (mainly
for computation cost) at least for me, can't find all optimal tableaus if there
are many zeroes.
The next
versions focus on efficiency and some others point of effectiveness (the best
is: make a C ARM code, so it will be ~1000x faster on hp50g but anyway it will be much less portable
because userRPL runs on emulators and others calulators).
Version 0.1
Input required
A tableau
matrix and a base.
If you
have the problem of linear programming in standard canonical form:
where:
The tableau
matrix is:
The base
is:
An ordered
set of index that imply an inner identity matrix of tableau.
Output
The
program don't do smart things (at least in this version), it do some checks for
canonical form and after these checks it try to obtain optimal tableau from
input. Optimal tableau is the first tableau with even if there are others solutions (for
example in a unlimited ).
Moreover
it put on stack the partial results as tableau.
So you can
use it only for boring operations, not for "smart" steps.
Example
Inputs
(tableau & base)
Outputs
First is
the partial tableau, last is the optimal tableau.