Bug 2210608

I see an issue and PR have been filed already:

https://github.com/ljvmiranda921/pyswarms/issues/508

https://github.com/ljvmiranda921/pyswarms/pull/509

Iztok: could we carry the patch in the meantime to fix this FTBFS? 

Please let us know if there's anything we can do to help :)

Cheers,

I've applied the matplotlib patch, but it still fails:

_ ERROR at setup of TestGeneralOptimizer.test_train_history[optimizer_history0-cost_history-expected_shape0] _
self = <tests.optimizers.test_general_optimizer.TestGeneralOptimizer object at 0xffff6f8dde10>
request = <SubRequest 'optimizer_history' for <Function test_train_history[optimizer_history0-cost_history-expected_shape0]>>
options = {'c1': 0.3, 'c2': 0.7, 'k': 2, 'p': 2, ...}
    @pytest.fixture(params=topologies)
    def optimizer_history(self, request, options):
        opt = GeneralOptimizerPSO(
            n_particles=10,
            dimensions=2,
            options=options,
            topology=request.param,
        )
>       opt.optimize(sphere, 1000)
tests/optimizers/test_general_optimizer.py:51: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
pyswarms/single/general_optimizer.py:252: in optimize
    self.swarm.best_pos, self.swarm.best_cost = self.top.compute_gbest(
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
self = <pyswarms.backend.topology.pyramid.Pyramid object at 0xffff6f89ffd0>
swarm = Swarm(position=array([[0.99029359, 0.76531587],
       [0.6685371 , 0.85310907],
       [0.36142089, 0.14341208],
    ...6, 1.17473693, 0.15119209, 0.54211531, 0.53914621,
       0.4226627 , 1.00655433, 0.39733392, 1.45670983, 0.97363041]))
kwargs = {'c1': 0.3, 'c2': 0.7, 'k': 2, 'p': 2, ...}
pyramid = <scipy.spatial._qhull.Delaunay object at 0xffff6f893290>
    def compute_gbest(self, swarm, **kwargs):
        """Update the global best using a pyramid neighborhood approach
    
        This topology uses the :code:`Delaunay` class from :code:`scipy`. To
        prevent precision errors in the Delaunay class, custom
        :code:`qhull_options` were added. Namely, :code:`QJ0.001 Qbb Qc Qx`.
        The meaning of those options is explained in [qhull]. This method is
        used to triangulate N-dimensional space into simplices. The vertices of
        the simplicies consist of swarm particles. This method is adapted from
        the work of Lane et al.[SIS2008]
    
        [SIS2008] J. Lane, A. Engelbrecht and J. Gain, "Particle swarm optimization with spatially
        meaningful neighbours," 2008 IEEE Swarm Intelligence Symposium, St. Louis, MO, 2008,
        pp. 1-8. doi: 10.1109/SIS.2008.4668281
        [qhull] http://www.qhull.org/html/qh-optq.htm
    
        Parameters
        ----------
        swarm : pyswarms.backend.swarms.Swarm
            a Swarm instance
    
        Returns
        -------
        numpy.ndarray
            Best position of shape :code:`(n_dimensions, )`
        float
            Best cost
        """
        try:
            # If there are less than (swarm.dimensions + 1) particles they are all connected
            if swarm.n_particles < swarm.dimensions + 1:
                self.neighbor_idx = np.tile(
                    np.arange(swarm.n_particles), (swarm.n_particles, 1)
                )
                best_pos = swarm.pbest_pos[np.argmin(swarm.pbest_cost)]
                best_cost = np.min(swarm.pbest_cost)
            else:
                # Check if the topology is static or dynamic and assign neighbors
                if (
                    self.static and self.neighbor_idx is None
                ) or not self.static:
                    pyramid = Delaunay(
                        swarm.position, qhull_options="QJ0.001 Qbb Qc Qx"
                    )
                    indices, index_pointer = pyramid.vertex_neighbor_vertices
                    # Insert all the neighbors for each particle in the idx array
>                   self.neighbor_idx = np.array(
                        [
                            index_pointer[indices[i] : indices[i + 1]]
                            for i in range(swarm.n_particles)
                        ]
                    )
E                   ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (10,) + inhomogeneous part.
pyswarms/backend/topology/pyramid.py:81: ValueError


I've filed https://github.com/ljvmiranda921/pyswarms/issues/516

This bug appears to have been reported against 'rawhide' during the Fedora Linux 39 development cycle.
Changing version to 39.