problem.getpivots
problem.getpivots |
Purpose
Returns a list of potential leaving variables if a specified variable enters the basis. The return value is a tuple containing the objective function value that would result if
in entered the basis; and an integer where the actual number of potential leaving variables will be returned.
Synopsis
dobj, npiv = problem.getpivots (in, outlist, x, maxpiv)
Arguments
|
in
|
Index of the specified row or column to enter basis.
|
|
outlist
|
Array of length at least
maxpiv to hold list of potential leaving variables. May be
None if not required.
|
|
x
|
Array of length
problem.attributes.rows + problem.attributes.sparerows + problem.attributes.cols to hold the values of all the variables that would result if
in entered the basis. May be
None if not required.
|
|
maxpiv
|
Maximum number of potential leaving variables to return.
|
Example
The following retrieves a list of up to 5 potential leaving variables if variable 6 enters the basis:
outlist = [] x = [] obj, npiv = p.getpivots (2, outlist, x, 10)
Further information
1. If the variable
in enters the basis and the problem is degenerate then several basic variables are candidates for leaving the basis, and the number of potential candidates is returned in
npiv. A list of at most
maxpiv of these candidates is returned in
outlist which must be at least
maxpiv long. If variable
in were to be pivoted in, then because the problem is degenerate, the resulting values of the objective function and all the variables do not depend on which of the candidates from
outlist is chosen to leave the basis. The value of the objective is returned in
dobj and the values of the variables into
x.
2. Row indices are in the range
0 to
ROWS-
1, whilst columns are in the range
ROWS+
SPAREROWS to
ROWS+
SPAREROWS+
COLS-
1.
Related topics