Pyomo.DoE: Add symbolic differentiation gradient calculation option as an alternative to finite different #7

adowling2 · 2025-06-04T13:39:01Z

Fixes # .

Summary/Motivation:

Instead of computing Jacobian (sensitivity of measurements/model outputs to model unknown parameters) via finite difference, use symbolic differentiation features in PyNumero
Development history and additional context:

Changes proposed in this PR:

Replaced FiniteDifferenceMethod with GradientMethod
Plan to deprecate fd_method in favor of gradient_method as an input for Pyomo.DoE
Added symbolic as an option in GradientMethod
Updated tests

Legal Acknowledgement

By contributing to this software project, I have read the contribution guide and agree to the following terms and conditions for my contribution:

I agree my contributions are submitted under the BSD license.
I represent I am authorized to make the contributions and grant the license. If my employer has rights to intellectual property that includes these contributions, I represent that I have received permission to make contributions and grant the required license on behalf of that employer.

… users from modifying it

…Created new temporary example for testing.

…sions.

adowling2 · 2025-06-06T13:14:53Z

Here is the current error message:

  File "/Users/adowling/GitHub/pyomo/pyomo/contrib/doe/doe.py", line 777, in _kaug_FIM
    self.kaug_FIM = self.kaug_jac.T @ cov_y @ self.kaug_jac + self.prior_FIM
ValueError: matmul: Input operand 1 has a mismatch in its core dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?) (size 27 is different from 4)

Here is output from some extra print statements:

Dimensions of kaug_jac =  (4, 25)
Dimensions of cov_y (27, 27)
Dimensions of prior_FIM (4, 4)

Why does my new PyNumero code use 25 measurements but the old code considers 27? I think the old code is counting the ODE initial conditions as both a measurement (output) and experiment input. But this needs to be investigated.

adowling2 · 2025-06-06T13:16:46Z

From my "shunk-works" branch dowlinglab/pyomo-doe#8 ...

Experiment model size:
  179 total variables
  27 outputs (measurements)
  10 inputs
  4 unknown parameters
  175 constraints

New thought: something is not getting fixed properly before calling my AD/SD code.

adowling2 · 2025-06-06T20:24:46Z

Good news: most recent change (fixing input parameters and variables)

Bad news: solver failure on a machine without good Ipopt:

WARNING: Loading a SolverResults object with a warning status into
model.name="unknown";
    - termination condition: other
    - message from solver: <undefined>
WARNING: :::::::::::Warning: Cannot converge this run.::::::::::::
--- Logging error ---
Traceback (most recent call last):
  File "/Users/adowling/DowlingLab/pyomo/pyomo/contrib/doe/doe.py", line 605, in _sequential_FIM
    pyo.assert_optimal_termination(res)
  File "/Users/adowling/DowlingLab/pyomo/pyomo/opt/results/solver.py", line 196, in assert_optimal_termination
    raise RuntimeError(msg)
RuntimeError: Solver failed to return an optimal solution. Solution status: warning, Termination condition: other

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/Users/adowling/DowlingLab/pyomo/pyomo/contrib/doe/doe.py", line 1636, in compute_FIM_full_factorial
    self.compute_FIM(model=model, method=method)
  File "/Users/adowling/DowlingLab/pyomo/pyomo/contrib/doe/doe.py", line 515, in compute_FIM
    self._sequential_FIM(model=model)
  File "/Users/adowling/DowlingLab/pyomo/pyomo/contrib/doe/doe.py", line 609, in _sequential_FIM
    raise RuntimeError(
RuntimeError: Model from experiment did not solve appropriately. Make sure the model is well-posed.

adowling2 · 2025-06-06T20:26:57Z

Actually, there is still a dimension mismatch error:

Dimensions of kaug_jac =  (4, 25)
Dimensions of cov_y (27, 27)
Dimensions of prior_FIM (4, 4)

adowling2 · 2025-06-06T22:52:05Z

Confirmed the issue: my AD code is not considering CA and CB at time zero are measurements. I suspect that is because these are experiment inputs and thus fixed.

adowling2 · 2025-06-06T23:00:58Z

Interesting... CA[0] is an experiment input, but CB[0] is not. By construction, for this problem, we fixed CB[0] = 0.

Should we allow the user to specify quantities that are fixed as measurements? If they do, the gradient (sensitivity) should be zero. I should be able to handle this zero gradient.

…t_fixed.

adowling2 · 2025-06-11T13:24:36Z

Results from symbolic sensitivity analysis (reactor example):

adowling2 · 2025-06-11T13:29:46Z

Results with central finite difference and sequential:

adowling2 · 2025-06-11T13:32:16Z

Results generated with kaug:

adowling2 · 2025-06-11T13:39:22Z

Comparison of trace:

Same shape, but symbolic values are lower

Comparison of determinate:

Same shape, smallest value is lowest with symbolic

Ideas:

Check that symbolic is assembling the measurement errors correctly
Compare Jacobians at a single point
Symbolic explicitly adds a zero row for fixed measurements... there might be a small gradient with the others due to solver convergence tolerances

adowling2 · 2025-06-11T13:50:00Z

Results for a single point using kaug:

kaug_jac
: [[ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]
 [-1.63468975e+00 -7.20926079e-15  3.09547195e+00 -3.79355202e-14]
 [-1.33002472e+00 -1.79960585e-14  2.89441710e+00  1.24206956e-14]
 [-1.02460753e+00 -1.21392639e-14  2.41080622e+00  3.24985417e-15]
 [-7.61613008e-01 -8.14617661e-15  1.88410535e+00 -7.98770880e-16]
 [-5.52026118e-01 -5.44195302e-15  1.41418067e+00 -2.27373609e-15]
 [-3.92682429e-01 -3.62105040e-15  1.03222972e+00 -2.53378302e-15]
 [-2.75314860e-01 -2.40099454e-15  7.38175156e-01 -2.28132668e-15]
 [-1.90812791e-01 -1.58704881e-15  5.19693626e-01 -1.86332539e-15]
 [ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]
 [ 5.33726838e-01 -1.24246100e+00 -1.03703316e+00  4.53010295e+00]
 [ 1.88837814e-01 -1.33332768e+00 -7.33381777e-01  5.39096253e+00]
 [-1.19831767e-01 -1.42212773e+00 -1.97331587e-01  6.20642210e+00]
 [-3.55637395e-01 -1.49674329e+00  3.21685806e-01  6.90683175e+00]
 [-5.16043971e-01 -1.55195047e+00  7.31799560e-01  7.46377434e+00]
 [-6.12063742e-01 -1.58648627e+00  1.01535861e+00  7.87240556e+00]
 [-6.58324208e-01 -1.60126116e+00  1.18593874e+00  8.14070357e+00]
 [-6.68686268e-01 -1.59828246e+00  1.26716070e+00  8.28300293e+00]
 [ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]
 [ 1.10096291e+00  1.24246100e+00 -2.05843879e+00 -4.53010295e+00]
 [ 1.14118691e+00  1.33332768e+00 -2.16103533e+00 -5.39096253e+00]
 [ 1.14443929e+00  1.42212773e+00 -2.21347463e+00 -6.20642210e+00]
 [ 1.11725040e+00  1.49674329e+00 -2.20579115e+00 -6.90683175e+00]
 [ 1.06807009e+00  1.55195047e+00 -2.14598023e+00 -7.46377434e+00]
 [ 1.00474617e+00  1.58648627e+00 -2.04758834e+00 -7.87240556e+00]
 [ 9.33639068e-01  1.60126116e+00 -1.92411390e+00 -8.14070357e+00]
 [ 8.59499060e-01  1.59828246e+00 -1.78685433e+00 -8.28300293e+00]]


cov_y
: [[100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
  100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.]]


kaug_FIM
: [[  1745.81343391   1599.21859987  -3512.47892155  -7589.26220445]
 [  1599.21859987   3525.63856364  -2900.94638673 -16465.46338508]
 [ -3512.47892155  -2900.94638673   7190.13048958  13849.96839993]
 [ -7589.26220445 -16465.46338508  13849.96839993  77674.03976715]]


prior_FIM
: [[0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]]

adowling2 · 2025-06-11T13:51:00Z

Results using symbolic:

kaug_jac
: [[ 0.          0.          0.          0.        ]
 [-0.01927928  0.          0.39787557  0.        ]
 [-0.0156861   0.          0.37203305  0.        ]
 [-0.01208406  0.          0.30987226  0.        ]
 [-0.00898234  0.          0.24217292  0.        ]
 [-0.00651051  0.          0.18177129  0.        ]
 [-0.00463124  0.          0.13267734  0.        ]
 [-0.00324702  0.          0.09488113  0.        ]
 [-0.00225042 -0.          0.06679867 -0.        ]
 [ 0.          0.          0.          0.        ]
 [ 0.00629469 -0.00334246 -0.13329475  0.30100352]
 [ 0.00222712 -0.00358691 -0.09426501  0.35820349]
 [-0.00141328 -0.0038258  -0.02536396  0.41238685]
 [-0.00419433 -0.00402653  0.04134779  0.4589257 ]
 [-0.00608614 -0.00417505  0.09406164  0.49593185]
 [-0.00721858 -0.00426796  0.13050882  0.52308343]
 [-0.00776417 -0.00430771  0.15243429  0.54091054]
 [-0.00788638 -0.00429969  0.16287413  0.55036564]
 [ 0.          0.         -0.         -0.        ]
 [ 0.01298458  0.00334246 -0.26458082 -0.30100352]
 [ 0.01345898  0.00358691 -0.27776804 -0.35820349]
 [ 0.01349734  0.0038258  -0.28450831 -0.41238685]
 [ 0.01317668  0.00402653 -0.28352071 -0.4589257 ]
 [ 0.01259665  0.00417505 -0.27583293 -0.49593185]
 [ 0.01184982  0.00426796 -0.26318616 -0.52308343]
 [ 0.01101119  0.00430771 -0.24731541 -0.54091054]
 [ 0.0101368   0.00429969 -0.22967279 -0.55036564]]


cov_y
: [[100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
  100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.]]


kaug_FIM
: [[ 2.42833478e-01  5.07396268e-02 -5.32463073e+00 -5.94728050e+00]
 [ 5.07396268e-02  2.55156086e-02 -1.00309988e+00 -2.94321189e+00]
 [-5.32463073e+00 -1.00309988e+00  1.18789370e+02  1.18285820e+02]
 [-5.94728050e+00 -2.94321189e+00  1.18285820e+02  3.42927958e+02]]


prior_FIM
: [[0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]]

adowling2 · 2025-06-11T13:55:13Z

I should make a much easier test example without ODEs/DAEs.

adowling2 · 2025-06-11T15:01:59Z

Set scaling to False; symbolic results did not change.

Results using kaug without scaling:

Dimensions of kaug_jac =  (27, 4)
Dimensions of cov_y (27, 27)
Dimensions of prior_FIM (4, 4)
kaug_jac
: [[ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]
 [-1.92792752e-02 -1.93943312e-17  3.97875573e-01 -2.52063257e-15]
 [-1.56861036e-02 -4.84129411e-17  3.72033047e-01  8.25295390e-16]
 [-1.20840609e-02 -3.26570103e-17  3.09872265e-01  2.15937154e-16]
 [-8.98234471e-03 -2.19148192e-17  2.42172924e-01 -5.30744771e-17]
 [-6.51050971e-03 -1.46399253e-17  1.81771294e-01 -1.51078810e-16]
 [-4.63123516e-03 -9.74133864e-18  1.32677342e-01 -1.68357675e-16]
 [-3.24702040e-03 -6.45914811e-18  9.48811255e-02 -1.51583168e-16]
 [-2.25041622e-03 -4.26947383e-18  6.67986666e-02 -1.23808996e-16]
 [ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]
 [ 6.29469086e-03 -3.34246476e-03 -1.33294751e-01  3.01003518e-01]
 [ 2.22712365e-03 -3.58691402e-03 -9.42650099e-02  3.58203491e-01]
 [-1.41327712e-03 -3.82580363e-03 -2.53639571e-02  4.12386850e-01]
 [-4.19433182e-03 -4.02653418e-03  4.13477900e-02  4.58925698e-01]
 [-6.08614189e-03 -4.17505238e-03  9.40616402e-02  4.95931850e-01]
 [-7.21858405e-03 -4.26796047e-03  1.30508819e-01  5.23083426e-01]
 [-7.76417276e-03 -4.30770785e-03  1.52434286e-01  5.40910537e-01]
 [-7.88638128e-03 -4.29969455e-03  1.62874127e-01  5.50365643e-01]
 [ 0.00000000e+00  0.00000000e+00  0.00000000e+00  0.00000000e+00]
 [ 1.29845844e-02  3.34246476e-03 -2.64580821e-01 -3.01003518e-01]
 [ 1.34589799e-02  3.58691402e-03 -2.77768037e-01 -3.58203491e-01]
 [ 1.34973380e-02  3.82580363e-03 -2.84508308e-01 -4.12386850e-01]
 [ 1.31766765e-02  4.02653418e-03 -2.83520714e-01 -4.58925698e-01]
 [ 1.25966516e-02  4.17505238e-03 -2.75832935e-01 -4.95931850e-01]
 [ 1.18498192e-02  4.26796047e-03 -2.63186162e-01 -5.23083426e-01]
 [ 1.10111932e-02  4.30770785e-03 -2.47315411e-01 -5.40910537e-01]
 [ 1.01367975e-02  4.29969455e-03 -2.29672793e-01 -5.50365643e-01]]


cov_y
: [[100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
  100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.]]


kaug_FIM
: [[ 2.42833478e-01  5.07396268e-02 -5.32463073e+00 -5.94728050e+00]
 [ 5.07396268e-02  2.55156086e-02 -1.00309988e+00 -2.94321189e+00]
 [-5.32463073e+00 -1.00309988e+00  1.18789370e+02  1.18285820e+02]
 [-5.94728050e+00 -2.94321189e+00  1.18285820e+02  3.42927958e+02]]


prior_FIM
: [[0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]]

Conclusion: I need to implement scaling!

adowling2 · 2025-06-11T15:31:56Z

Results with scaling for symbolic:

kaug_jac
: [[ 0.          0.          0.          0.        ]
 [-1.63468975  0.          3.09547195  0.        ]
 [-1.33002472  0.          2.8944171   0.        ]
 [-1.02460753  0.          2.41080622  0.        ]
 [-0.76161301  0.          1.88410535  0.        ]
 [-0.55202612  0.          1.41418067  0.        ]
 [-0.39268243  0.          1.03222972  0.        ]
 [-0.27531486  0.          0.73817516  0.        ]
 [-0.19081279 -0.          0.51969363 -0.        ]
 [ 0.          0.          0.          0.        ]
 [ 0.53372684 -1.242461   -1.03703316  4.53010295]
 [ 0.18883781 -1.33332768 -0.73338178  5.39096253]
 [-0.11983177 -1.42212773 -0.19733159  6.2064221 ]
 [-0.35563739 -1.49674329  0.32168581  6.90683175]
 [-0.51604397 -1.55195047  0.73179956  7.46377434]
 [-0.61206374 -1.58648627  1.01535861  7.87240556]
 [-0.65832421 -1.60126116  1.18593874  8.14070357]
 [-0.66868627 -1.59828246  1.2671607   8.28300293]
 [ 0.          0.         -0.         -0.        ]
 [ 1.10096291  1.242461   -2.05843879 -4.53010295]
 [ 1.14118691  1.33332768 -2.16103533 -5.39096253]
 [ 1.14443929  1.42212773 -2.21347463 -6.2064221 ]
 [ 1.1172504   1.49674329 -2.20579115 -6.90683175]
 [ 1.06807009  1.55195047 -2.14598023 -7.46377434]
 [ 1.00474617  1.58648627 -2.04758834 -7.87240556]
 [ 0.93363907  1.60126116 -1.9241139  -8.14070357]
 [ 0.85949906  1.59828246 -1.78685433 -8.28300293]]


cov_y
: [[100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
  100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0. 100.]]


kaug_FIM
: [[  1745.81343391   1599.21859987  -3512.47892155  -7589.26220445]
 [  1599.21859987   3525.63856364  -2900.94638673 -16465.46338508]
 [ -3512.47892155  -2900.94638673   7190.13048958  13849.96839993]
 [ -7589.26220445 -16465.46338508  13849.96839993  77674.03976715]]


prior_FIM
: [[0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]
 [0. 0. 0. 0.]]

adowling2 · 2025-06-11T15:35:32Z

More results using symbolic with scaling. The results are now consistent!

adowling2 · 2025-06-13T00:14:07Z

Original Pyomo.DoE for the reaction kinetics example:

This is Ipopt version 3.13.2, running with linear solver ma27.

Number of nonzeros in equality constraint Jacobian...:     6370
Number of nonzeros in inequality constraint Jacobian.:        0
Number of nonzeros in Lagrangian Hessian.............:     1026

Total number of variables............................:     1608
                     variables with only lower bounds:      732
                variables with lower and upper bounds:      256
                     variables with only upper bounds:        0
Total number of equality constraints.................:     1598
Total number of inequality constraints...............:        0
        inequality constraints with only lower bounds:        0
   inequality constraints with lower and upper bounds:        0
        inequality constraints with only upper bounds:        0

iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
   0 -9.9386618e+00 1.54e-01 1.18e+00  -1.0 0.00e+00    -  0.00e+00 0.00e+00   0
   1 -9.9420153e+00 4.81e-02 1.31e-01  -1.0 1.26e+02    -  9.90e-01 9.90e-01h  1
   2 -9.9949580e+00 7.14e-02 9.90e+00  -1.0 8.07e+01    -  1.00e+00 9.90e-01H  1
   3 -1.0165394e+01 4.24e-03 3.03e+00  -1.0 2.57e+02    -  1.00e+00 1.00e+00H  1
   4 -1.0287458e+01 1.97e+01 3.62e+00  -1.7 1.83e+02    -  1.00e+00 1.00e+00f  1
   5 -1.0543793e+01 1.83e+02 1.09e+01  -1.7 5.22e+02    -  1.00e+00 1.00e+00h  1
   6 -1.0533052e+01 4.34e+00 1.74e+00  -1.7 5.95e+02    -  1.00e+00 1.00e+00h  1
   7 -1.0532473e+01 4.53e-02 1.64e-02  -1.7 2.58e+01    -  1.00e+00 1.00e+00h  1
   8 -1.0782349e+01 2.24e+02 3.76e+05  -3.8 5.33e+02    -  8.11e-01 1.00e+00f  1
   9 -1.1083263e+01 5.07e+02 1.21e+05  -3.8 1.89e+03    -  6.78e-01 1.00e+00h  1
iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
  10 -1.1226667e+01 2.45e+01 1.24e+04  -3.8 4.39e+02    -  8.97e-01 1.00e+00h  1
  11 -1.1288911e+01 2.11e+01 8.72e-01  -3.8 4.28e+02    -  1.00e+00 1.00e+00h  1
  12 -1.1290305e+01 1.11e+00 4.54e-02  -3.8 3.14e+02    -  1.00e+00 1.00e+00h  1
  13 -1.1290370e+01 3.44e-04 2.32e-04  -3.8 3.58e+00    -  1.00e+00 1.00e+00h  1
  14 -1.1315467e+01 1.66e+00 1.26e+03  -5.7 1.03e+02    -  9.15e-01 1.00e+00h  1
  15 -1.1317504e+01 9.09e-02 3.82e-02  -5.7 2.33e+01    -  1.00e+00 1.00e+00h  1
  16 -1.1317664e+01 1.84e-03 2.72e-03  -5.7 5.37e+00    -  1.00e+00 1.00e+00h  1
  17 -1.1317669e+01 2.04e-05 6.62e-05  -5.7 3.82e-01    -  1.00e+00 1.00e+00h  1
  18 -1.1317669e+01 1.10e-08 3.54e-08  -5.7 7.69e-03    -  1.00e+00 1.00e+00h  1
  19 -1.1318050e+01 8.16e-04 2.57e-01  -8.6 2.28e+00    -  9.99e-01 1.00e+00h  1
iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
  20 -1.1318051e+01 2.84e-07 9.25e-07  -8.6 4.52e-02    -  1.00e+00 1.00e+00h  1
  21 -1.1318051e+01 2.91e-11 9.57e-12  -8.6 1.27e-04    -  1.00e+00 1.00e+00h  1

Number of Iterations....: 21

                                   (scaled)                 (unscaled)
Objective...............:  -1.1318050921093612e+01   -1.1318050921093612e+01
Dual infeasibility......:   9.5672042699536117e-12    9.5672042699536117e-12
Constraint violation....:   3.0004887463519481e-12    2.9103830456733704e-11
Complementarity.........:   2.5059065524033082e-09    2.5059065524033082e-09
Overall NLP error.......:   2.5059065524033082e-09    2.5059065524033082e-09


Number of objective function evaluations             = 26
Number of objective gradient evaluations             = 22
Number of equality constraint evaluations            = 26
Number of inequality constraint evaluations          = 0
Number of equality constraint Jacobian evaluations   = 22
Number of inequality constraint Jacobian evaluations = 0
Number of Lagrangian Hessian evaluations             = 21
Total CPU secs in IPOPT (w/o function evaluations)   =      0.086
Total CPU secs in NLP function evaluations           =      0.003

EXIT: Optimal Solution Found.
Optimal experiment values: 
	Initial concentration: 5.00
	Temperature values: [481.88, 300.00, 300.00, 300.00, 300.00, 300.00, 300.00, 300.00, 300.00]
FIM at optimal design:
 [[  1731.79089055   1442.18099318  -3647.74691348  -7115.51970487]
 [  1442.18099318   3479.17598013  -2710.24342004 -16836.77882308]
 [ -3647.74691348  -2710.24342004   7809.41772632  13457.72531516]
 [ -7115.51970487 -16836.77882308  13457.72531516  82181.28341606]]
Objective value at optimal design: 11.32

***Optimal Experiment Design***
  Decision Var   Value
0        CA[0]    5.00
1         T[0]  481.88
2     T[0.125]  300.00
3      T[0.25]  300.00
4     T[0.375]  300.00
5       T[0.5]  300.00
6     T[0.625]  300.00
7      T[0.75]  300.00
8     T[0.875]  300.00
9         T[1]  300.00

adowling2 · 2025-06-13T00:14:32Z

With symbolic derivatives:

This is Ipopt version 3.13.2, running with linear solver ma27.

Number of nonzeros in equality constraint Jacobian...:     4341
Number of nonzeros in inequality constraint Jacobian.:        0
Number of nonzeros in Lagrangian Hessian.............:      886

Total number of variables............................:     1013
                     variables with only lower bounds:       95
                variables with lower and upper bounds:       32
                     variables with only upper bounds:        0
Total number of equality constraints.................:     1003
Total number of inequality constraints...............:        0
        inequality constraints with only lower bounds:        0
   inequality constraints with lower and upper bounds:        0
        inequality constraints with only upper bounds:        0

iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
   0 -1.1297038e+01 2.65e+00 1.10e+00  -1.0 0.00e+00    -  0.00e+00 0.00e+00   0
   1 -1.1242568e+01 3.72e+00 6.29e-01  -1.0 1.83e+02    -  9.89e-01 9.90e-01h  1
   2 -1.1179453e+01 1.80e+02 1.59e+01  -1.0 3.96e+03    -  9.74e-01 9.90e-01f  1
   3 -1.0834854e+01 9.82e+02 1.40e+01  -1.0 6.32e+03    -  1.00e+00 1.00e+00h  1
   4 -1.0698726e+01 3.29e+02 2.57e+00  -1.0 6.92e+03    -  1.00e+00 1.00e+00h  1
   5 -1.0688962e+01 4.53e+00 8.60e-01  -1.0 1.77e+02    -  1.00e+00 1.00e+00h  1
   6 -1.0874882e+01 1.41e+02 6.54e+00  -1.7 4.91e+02    -  1.00e+00 1.00e+00h  1
   7 -1.1055990e+01 1.88e+02 2.75e+00  -1.7 2.25e+03    -  1.00e+00 1.00e+00h  1
   8 -1.1047059e+01 1.87e+01 2.17e-01  -1.7 1.43e+03    -  1.00e+00 1.00e+00h  1
   9 -1.1046237e+01 4.14e-02 3.55e-03  -1.7 2.89e+01    -  1.00e+00 1.00e+00h  1
iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
  10 -1.1213820e+01 8.69e+01 5.06e+05  -3.8 6.72e+02    -  7.45e-01 1.00e+00f  1
  11 -1.1290645e+01 1.29e+01 6.48e+04  -3.8 2.89e+02    -  8.72e-01 1.00e+00h  1
  12 -1.1312328e+01 4.70e+00 1.84e-01  -3.8 1.55e+02    -  1.00e+00 1.00e+00h  1
  13 -1.1314140e+01 1.40e-01 2.21e-02  -3.8 2.84e+01    -  1.00e+00 1.00e+00h  1
  14 -1.1314255e+01 1.58e-03 1.10e-03  -3.8 2.71e+00    -  1.00e+00 1.00e+00h  1
  15 -1.1317925e+01 6.21e-02 1.54e+02  -5.7 2.00e+01    -  9.90e-01 1.00e+00h  1
  16 -1.1318010e+01 4.24e-04 6.08e-04  -5.7 1.54e+00    -  1.00e+00 1.00e+00h  1
  17 -1.1318012e+01 2.45e-06 5.87e-06  -5.7 1.17e-01    -  1.00e+00 1.00e+00h  1
  18 -1.1318060e+01 1.33e-05 3.48e-06  -8.6 2.91e-01    -  1.00e+00 1.00e+00h  1
  19 -1.1318060e+01 1.08e-10 2.61e-10  -8.6 8.44e-04    -  1.00e+00 1.00e+00h  1

Number of Iterations....: 19

                                   (scaled)                 (unscaled)
Objective...............:  -1.1318059589328048e+01   -1.1318059589328048e+01
Dual infeasibility......:   2.6113792358236406e-10    2.6113792358236406e-10
Constraint violation....:   1.0797918115201810e-10    1.0797918115201810e-10
Complementarity.........:   2.5068103979073895e-09    2.5068103979073895e-09
Overall NLP error.......:   2.5068103979073895e-09    2.5068103979073895e-09


Number of objective function evaluations             = 20
Number of objective gradient evaluations             = 20
Number of equality constraint evaluations            = 20
Number of inequality constraint evaluations          = 0
Number of equality constraint Jacobian evaluations   = 20
Number of inequality constraint Jacobian evaluations = 0
Number of Lagrangian Hessian evaluations             = 19
Total CPU secs in IPOPT (w/o function evaluations)   =      0.035
Total CPU secs in NLP function evaluations           =      0.003

EXIT: Optimal Solution Found.
INFO: Successfully optimized experiment. Solve time: 0.1 seconds
INFO: Total time for build, initialization, and solve: 0.4 seconds
Optimal experiment values: 
	Initial concentration: 5.00
	Temperature values: [481.88, 300.00, 300.00, 300.00, 300.00, 300.00, 300.00, 300.00, 300.00]
FIM at optimal design:
 [[  1731.78924161   1442.17360943  -3647.75456986  -7115.52005827]
 [  1442.17360943   3479.17107758  -2710.23709531 -16836.83659169]
 [ -3647.75456986  -2710.23709531   7809.45694289  13457.76589358]
 [ -7115.52005827 -16836.83659169  13457.76589358  82181.9737957 ]]
Objective value at optimal design: 11.32

***Optimal Experiment Design***
  Decision Var   Value
0        CA[0]    5.00
1         T[0]  481.88
2     T[0.125]  300.00
3      T[0.25]  300.00
4     T[0.375]  300.00
5       T[0.5]  300.00
6     T[0.625]  300.00
7      T[0.75]  300.00
8     T[0.875]  300.00
9         T[1]  300.00

adowling2 · 2025-06-13T00:20:38Z

@djlaky @blnicho @michaelbynum @jsiirola I am excited to share we have a working draft implementation of Pyomo.DoE using symbolic differentiation to build the sensitivity matrix instead of finite difference. I tested this on the reactor example. The model is smaller and solves in Ipopt ~2x faster.

Now that there is a draft implementation, this is ready for a software design review. I appreciate your comments in advance. Also, is there a good way to move this PR from my fork to the main Pyomo repo?

blnicho · 2025-06-13T14:14:52Z

@adowling2 thanks for the update, this is really exciting! You may be able to edit the target branch for this PR to point to the main Pyomo repo (I can't remember if GitHub allows you to edit the target to a different fork). If that doesn't work then you'll have to open a new PR into Pyomo.

adowling2 added 4 commits June 4, 2025 08:46

Replaced FiniteDifferenceStep with GradientMethod and extended

3f32b89

Updated tests. Decided to make _gradient_method provide to discourage…

c35c2e3

… users from modifying it

Updated reactor example to new interface. Improve solution printing. …

7528d7f

…Created new temporary example for testing.

Added comments to plan out changes for symbolic gradients.

b97dbfa

adowling2 mentioned this pull request Jun 4, 2025

Extract Jacobian for equality constraints then embed symbolic expressions in constraints Pyomo/pyomo#3564

Open

adowling2 added 5 commits June 4, 2025 10:00

Updated notes based on conversation with Dan

6df0470

Started to add logic for symbolic gradient

dcce045

Symbolic example "runs", but there are no gradient calculations yet

3ac0a52

Decided to make the kaug function also support AD

21ab51e

Added class for differentiation with PyNumero. Debugging matrix dimen…

cf24135

…sions.

adowling2 added 2 commits June 6, 2025 16:16

Minor changes to examples to run on a machine without HSL installed

f791f2b

Fixed dimensions issue

fdefe5a

Added some debugging lines.

dae1c46

adowling2 marked this pull request as draft June 6, 2025 22:51

More debugging code

54f6769

adowling2 added 5 commits June 6, 2025 19:09

Added some comments about next steps

49e2ddf

Added comments for next step

362fea9

Check in while developing... I think I do not actually need the outpu…

222430b

…t_fixed.

Intermediate check-in before removing debugging lines

aef38dd

Symbolic sensitivity analysis runs!

0eb3b9b

Generating results for comparison.

aed858b

Generating results with kaug

d45c479

Debugging with kaug

8eea498

adowling2 added 2 commits June 11, 2025 11:07

Added simply polynomial example

0d3978b

Added scaling for symbolic

9ea42fc

adowling2 added 10 commits June 11, 2025 11:36

Reran the heat map sensitivity analysis

9f8df60

Left some comments on next steps; enabled logging for simple timing

b8ec21c

Work in progress for symbolic version

410e6fc

Minor formatting change

e3bfcd5

Created suffix to track mapping

ae53caf

Updated to use measurement_mapping

92c3e47

It appears to work!

7dc978b

Need debugging help with ComponentUID

1f0e61f

I think it works now

cc8d579

Added scaling

1456f14

Pyomo.DoE: Add symbolic differentiation gradient calculation option as an alternative to finite different #7

Are you sure you want to change the base?

Pyomo.DoE: Add symbolic differentiation gradient calculation option as an alternative to finite different #7

Uh oh!

Conversation

adowling2 commented Jun 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Fixes # .

Summary/Motivation:

Changes proposed in this PR:

Legal Acknowledgement

Uh oh!

adowling2 commented Jun 6, 2025

Uh oh!

adowling2 commented Jun 6, 2025

Uh oh!

adowling2 commented Jun 6, 2025

Uh oh!

adowling2 commented Jun 6, 2025

Uh oh!

adowling2 commented Jun 6, 2025

Uh oh!

adowling2 commented Jun 6, 2025

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adowling2 commented Jun 11, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 11, 2025

Uh oh!

adowling2 commented Jun 13, 2025

Uh oh!

adowling2 commented Jun 13, 2025

Uh oh!

adowling2 commented Jun 13, 2025

Uh oh!

blnicho commented Jun 13, 2025

Uh oh!

Uh oh!

adowling2 commented Jun 4, 2025 •

edited

Loading

adowling2 commented Jun 11, 2025 •

edited

Loading

adowling2 commented Jun 11, 2025 •

edited

Loading

adowling2 commented Jun 11, 2025 •

edited

Loading