Chapter Contents |
Previous |
Next |
The NETFLOW Procedure |
Description | Statement | Option |
Output Data Set Options | ||
unconstrained solution data set | RESET | ARCOUT= |
unconstrained solution data set | RESET | NODEOUT= |
constrained solution data set | RESET | CONOUT= |
constrained solution data set | RESET | DUALOUT= |
Stop Optimization Options | ||
pause after stage 1. Don't start stage 2 | RESET | ENDPAUSE1 |
pause when feasible - stage 1 | RESET | FEASIBLEPAUSE1 |
pause when feasible - stage 2 | RESET | FEASIBLEPAUSE2 |
maximum number of iterations - stage 1 | RESET | MAXIT1= |
maximum number of iterations - stage 2 | RESET | MAXIT2= |
negates ENDPAUSE1 | RESET | NOENDPAUSE1 |
negates FEASIBLEPAUSE1 | RESET | NOFEASIBLEPAUSE1 |
negates FEASIBLEPAUSE2 | RESET | NOFEASIBLEPAUSE2 |
pause every PAUSE1 iterations - stage 1 | RESET | PAUSE1= |
pause every PAUSE2 iterations - stage 2 | RESET | PAUSE2= |
Simplex Options | ||
Use Big M instead of twophase - stage 1 | RESET | BIGM1 |
Use Big M instead of twophase - stage 2 | RESET | BIGM2 |
anti-cycling option | RESET | CYCLEMULT1 |
interchange first eligible nonkey with leaving key variable | RESET | INTFIRST |
invert working basis matrix (2-dim) every INVFREQ= iterations | RESET | INVFREQ= |
maximum number of L row operations allowed before refactorization done instead of factor column update | RESET | MAXL= |
maximum number fo LU factor column updates, otherwise refactorize | RESET | MAXLUUPDATES= |
anticycling option | RESET | MINBLOCK1 |
first eligible leaving variable used, otherwise best is used | RESET | LRATIO1 |
first eligible leaving variable used, otherwise best is used | RESET | LRATIO2 |
otherwise best is used | ||
negates INTFIRST | RESET | NOINTFIRST |
negates LRATIO1 | RESET | NOLRATIO1 |
negates LRATIO2 | RESET | NOLRATIO2 |
negates PERTURB1 | RESET | NOPERTURB1 |
anti-cycling option | RESET | PERTURB1 |
re-factorize working basis matrix every REFACTFREQ= iterations | RESET | REFACTFREQ= |
use twophase instead of Big M - stage 1 | RESET | TWOPHASE1 |
use twophase instead of Big M - stage 2 | RESET | TWOPHASE2 |
pivot element selection parameter | RESET | U= |
zero tolerance - stage 1 | RESET | ZERO1= |
zero tolerance - stage 2 | RESET | ZERO2= |
zero tolerance - real number comparisons | RESET | ZEROTOL= |
Pricing Options | ||
calculate dual values every DUALFREQ= iterations after major iterations | RESET | DUALFREQ= |
pricing strategy - stage 1 | RESET | PRICETYPE1= |
pricing strategy - stage 2 | RESET | PRICETYPE2= |
used when P1SCAN=PARTIAL | RESET | P1NPARTIAL= |
how nonbasis variables (PRICETYPE1=NOQ) or queue candidates (PRICETYPE1=Q) are scanned | RESET | P1SCAN= |
used when P2SCAN=PARTIAL | RESET | P2NPARTIAL= |
how nonbasis variables (PRICETYPE2=NOQ) or queue candidates (PRICETYPE2=Q) are scanned | RESET | P2SCAN= |
initial queue size - stage 1 | RESET | QSIZE1= |
initial queue size - stage 2 | RESET | QSIZE2= |
used when Q1FILLSCAN=PARTIAL | RESET | Q1FILLNPARTIAL= |
how candidates are scanned when filling queue - stage 1 | RESET | Q1FILLSCAN= |
used when Q2FILLSCAN=PARTIAL | RESET | Q2FILLNPARTIAL= |
how candidates are scanned when filling queue - stage 2 | RESET | Q2FILLSCAN= |
queue size reduction factor | RESET | REDUCEQSIZE1= |
queue size reduction factor | RESET | REDUCEQSIZE2= |
when the queue are refreshed - stage 1 | RESET | REFRESHQ1= |
when the queue are refreshed - stage 2 | RESET | REFRESHQ2= |
Miscellaneous Options | ||
output complete basis information to ARCOUT and NODEOUT data sets | RESET | FUTURE1 |
output complete basis information to CONOUT and DUALOUT data sets | RESET | FUTURE2 |
turn off infeasibility or optimality flags as more optimization is to be done | RESET | MOREOPT |
negates FUTURE1 | RESET | NOFUTURE1 |
negates FUTURE2 | RESET | NOFUTURE2 |
negates SCRATCH | RESET | NOSCRATCH |
negates ZTOL1 | RESET | NOZTOL1 |
negates ZTOL2 | RESET | NOZTOL2 |
do not do stage 1 optimization. Do stage 2 optimization | RESET | SCRATCH |
display this number of similar SAS log messages, suppress the rest | RESET | VERBOSE= |
use zero tolerance- stage 1 | RESET | ZTOL1 |
use zero tolerance- stage 2 | RESET | ZTOL2 |
Interior Point algorithm Options | ||
allowed amount of dual infeasibility | RESET | TOLDINF= |
allowed amount of primal infeasibility | RESET | TOLPINF= |
cut-off tolerance for Cholesky factorization | RESET | CHOLTINYTOL= |
density threshold for Cholesky processing | RESET | DENSETHR= |
maximum number of Interior Point algorithm iterations | RESET | MAXITERB= |
Primal-Dual (Duality) gap tolerance | RESET | PDGAPTOL= |
step-length multiplier | RESET | PDSTEPMULT= |
preprocessing type | RESET | PRSLTYPE= |
proc netflow nodedata=noded /* supply and demand data */ arcdata=arcd1 /* the arc descriptions */ condata=cond1 /* the side constraints */ conout=solution; /* output the solution */ run; /* Netflow states that the problem is infeasible. */ /* You suspect that the zero tolerance is too large */ reset zero2=1.0e-10 moreopt; run; /* Netflow will attempt more optimization. */ /* After this, if it reports that the problem is */ /* infeasible, the problem really might be infeasible */
ERROR: The HEAD list variable value in obs i in the ARCDATA is missing, - the TAIL list variable value of this obs is nonmissing. This is an incomplete arc specification.If there are lots of observations that have this fault, messages that are similar are issued for only the first VERBOSE= such observations. After the ARCDATA= data set has been read, PROC NETFLOW will issue the message:
NOTE: More messages similar to the ones immediately above could have been issued but were suppressed as VERBOSE= v.
NOTE: To see all messages, specify VERBOSE=v.is issued. The value of v is the smallest value that should be specified for the VERBOSE= option so that all messages are displayed if PROC NETFLOW is run again with the same data and everything else except that the VERBOSE= option is unchanged. No messages are suppressed.
Chapter Contents |
Previous |
Next |
Top |
Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.