9 January 2009 10:23:14.753 AM UNCMIN_PRB FORTRAN77 version. Test the routines in the UNCMIN library. TEST01 Test OPTIF0, the simple interface to UNCMIN. RESULT - Iterate 0 Step 0.00000 0.00000 X 1.00000 1.00000 Function value = 783.98157 Gradient vector 759.888 2403.20 Hessian matrix 1.00000 0.00000 0.00000 1.00000 RESULT - Iterate 3 X 20.0000 -20.2152 Function value = 91.000000 Gradient vector 0.198364E-01 0.00000 Output from UNCMIN: Error code= 1 EXPLAIN: UNCMIN returned a termination code of 1 This code has the following meaning: The gradient is relatively close to zero. The current iterate is probably a good solution. F(X*) = 91.000000 X* = 19.999983 -20.215185 (Partial reference results for comparison:) 19.9145 -20.6011 -5.26250 19.9900 -20.6230 19.9145 20.0100 -20.6230 19.9145 19.9900 -20.6023 19.9145 Error code = 1 F(X*) = 91.0001 X* = 19.9900 -20.6230 uncprb function of 3 variables. solve problem using line search. RESULT - Iterate 0 Step 0.00000 0.00000 0.00000 X -1.00000 0.00000 0.00000 Function value = 95579.453 Gradient vector -0.00000 -61831.9 -6183.19 Hessian matrix 199.997 -61851.3 0.00000 -61851.3 20058.6 1998.05 0.00000 1998.05 201.660 RESULT - Iterate 25 X 1.00002 -0.825389E-03 -0.825292E-02 Function value = 6.81394886E-05 Gradient vector 0.338417E-02 -0.166425E-02 -0.163397E-01 termination code itrmcd= 4 EXPLAIN: UNCMIN returned a termination code of 4 This code has the following meaning: The iteration limit was exceeded. The current iterate does not satisfy the requirements. return code msg= 0 x* = 1.0000166 -8.25389405E-04 -8.25292431E-03 f(x*) = 6.81394886E-05 gradient= 3.38417129E-03 -1.66424504E-03 -1.63397007E-02 uncprb function of 3 variables. solve problem using double dogleg method. RESULT - Iterate 0 Step 0.00000 0.00000 0.00000 X -1.00000 0.00000 0.00000 Function value = 95579.453 Gradient vector -0.00000 -61831.9 -6183.19 Hessian matrix 199.997 -61851.3 0.00000 -61851.3 20058.6 1998.05 0.00000 1998.05 201.660 RESULT - Iterate 12 X 0.999995 0.246142E-03 0.246118E-02 Function value = 6.05992318E-06 Gradient vector -0.100148E-02 0.507327E-03 0.487160E-02 termination code itrmcd= 1 EXPLAIN: UNCMIN returned a termination code of 1 This code has the following meaning: The gradient is relatively close to zero. The current iterate is probably a good solution. return code msg= 0 x* = 0.99999499 2.46142212E-04 2.46118056E-03 f(x*) = 6.05992318E-06 gradient= -1.00148295E-03 5.07326855E-04 4.87160403E-03 uncprb function of 3 variables. solve problem using more-hebdon method. RESULT - Iterate 0 Step 0.00000 0.00000 0.00000 X -1.00000 0.00000 0.00000 Function value = 95579.453 Gradient vector -0.00000 -61831.9 -6183.19 Hessian matrix 199.997 -61851.3 0.00000 -61851.3 20058.6 1998.05 0.00000 1998.05 201.660 RESULT - Iterate 15 X 1.00000 -0.237842E-03 -0.237817E-02 Function value = 5.65806477E-06 Gradient vector 0.977401E-03 -0.485449E-03 -0.470781E-02 termination code itrmcd= 1 EXPLAIN: UNCMIN returned a termination code of 1 This code has the following meaning: The gradient is relatively close to zero. The current iterate is probably a good solution. return code msg= 0 x* = 1.0000049 -2.37841945E-04 -2.37816526E-03 f(x*) = 5.65806477E-06 gradient= 9.77400807E-04 -4.85449185E-04 -4.70780861E-03 UNCMIN_PRB Normal end of execution. 9 January 2009 10:23:14.757 AM