Solves a Linear Programming problem in "canonical" form, i.e.,
with equations only and xi ≥ 0.
The constraint matrix, A, must be given ending (each row)
with the right-hand side (RHS) constant ('return' at end of line).
So, e.g.,
− x1 + 4 x2 ≤ 78 would
become −1 4 … 78 .
The program finds the number of constraints.
This Problem follows the manual resolution by the matrix method
(revised simplex). For a "commercial" resolution:
NAG
version.
'Delta' is: (a) [V. Tavares, 1996] the reduced cost
(rc) vector; (b) [WinQSB, 1996] the rc vector for
the structural basic variables, and minus the
shadow prices for the constraints, according to the slack
variables. ('Lindo' [2002] gives symmetrical rc.) |
|
References:
• Tavares, L. Valadares,
Rui Carvalho Oliveira,
Isabel Hall Themido,
F. Nunes Correia,
1996,
"Investigação Operacional" (Operational Research), McGraw-Hill,
Amadora (Portugal).
• WinQSB ↓
(see instructions !)
by Yih-Long Chang in Lawrence, Jr., John A. and
Barry A. Pasternack, 2.nd ed., 2002,
"Applied Management Science: modeling,
spreadsheet analysis, and communication for decision making",
John Wiley, New York, NY (USA).
• Lindo
↓, Lindo Systems, Inc., Chicago, IL (USA).
• Wagner, Harvey M., 1972+,
"Principles of Operations
Research, with applications to managerial decisions",
John Wiley, New York, NY (USA). |
