State estimation converging and observable when determined unobservable.

[titou1del]

Describe the bug
The state estimation test: test_observability_analysis_and_pseudo_meas.py is not passing.

To Reproduce
Steps to reproduce the behavior:

1. Run test_observability_analysis_and_pseudo_meas.py from tests folder
2. Fails with error

============================== 1 failed in 3.28s ==============================
FAILED [100%]
src\tests\StateEstimation\test_observability_analysis_and_pseudo_meas.py:8 (test_se_with_and_without_pseudo_measurements)
def test_se_with_and_without_pseudo_measurements():
grid = gce.MultiCircuit()

    # --- Buses ---
    b1 = gce.Bus(name='B1', is_slack=True)
    b2 = gce.Bus(name='B2')
    b3 = gce.Bus(name='B3')
    grid.add_bus(b1)
    grid.add_bus(b2)
    grid.add_bus(b3)

    # --- Lines ---
    br1 = gce.Line(bus_from=b1, bus_to=b2, r=0.01, x=0.03)
    br2 = gce.Line(bus_from=b1, bus_to=b3, r=0.02, x=0.05)
    br3 = gce.Line(bus_from=b2, bus_to=b3, r=0.03, x=0.08)
    grid.add_line(br1)
    grid.add_line(br2)
    grid.add_line(br3)

    Sb = 100.0

    # --- Sparse measurements (insufficient) ---
    grid.add_pf_measurement(gce.PfMeasurement(0.888 * Sb, 0.008 * Sb, br1))
   # grid.add_pf_measurement(PfMeasurement(1.173 * Sb, 0.008 * Sb, br2))
    #grid.add_pi_measurement(PiMeasurement(-0.501 * Sb, 0.01 * Sb, b2))

    grid.add_qf_measurement(gce.QfMeasurement(0.568 * Sb, 0.008 * Sb, br1))
    grid.add_qf_measurement(gce.QfMeasurement(0.663 * Sb, 0.008 * Sb, br2))
    #grid.add_qi_measurement(QiMeasurement(-0.286 * Sb, 0.01 * Sb, b2))
    grid.add_vm_measurement(gce.VmMeasurement(1.006, 0.004, b1))
    grid.add_vm_measurement(gce.VmMeasurement(0.968, 0.004, b2))


    # --- SE without pseudo-measurements ---
    se_options_no_pseudo = gce.StateEstimationOptions(
        fixed_slack=False,
        run_observability_analyis=True,
        add_pseudo_measurements=False,  # no pseudo
        pseudo_meas_std=1.0,
        solver=gce.SolverType.NR, # NR used because in LM and GN we have relaxations that tend the
        # solution to converge
        verbose=0
    )
    se_no_pseudo = gce.StateEstimationDriver(circuit=grid, options=se_options_no_pseudo)
    with warnings.catch_warnings():
        warnings.filterwarnings("ignore", category=MatrixRankWarning)
        warnings.filterwarnings("ignore", message="invalid value encountered in divide", category=RuntimeWarning)
        se_no_pseudo.run()
    report_no_pseudo = se_no_pseudo.results.convergence_reports[0]
    assert report_no_pseudo

  assert not report_no_pseudo.is_observable

E assert not True
E + where True = <VeraGridEngine.Simulations.StateEstimation.state_stimation_driver.StateEstimationConvergenceReport object at 0x000002025EA0CA10>.is_observable

StateEstimation\test_observability_analysis_and_pseudo_meas.py:58: AssertionError

Expected behavior
There should be no observability or convergence when no measurements are provided.

The problematic lines, found in run method of state estimation driver are:
converged=solution.converged if solution else False,
is_observable=solution.is_observable if solution else is_observable,

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: Windows 11
• Version 6.0.20