[Iter 320] Code modification in PDE_D_DensityDragCIL.py

[Automated commit by Claude]

Author: allierc

src/ParticleGraph/generators/PDE_D_DensityDragCIL.py

@@ -0,0 +1,308 @@
+import torch
+import torch_geometric as pyg
+import torch_geometric.utils as pyg_utils
+from ParticleGraph.utils import to_numpy
+
+class PDE_D_DensityDragCIL(pyg.nn.MessagePassing):
+    """
+    Density-dependent mobility (CIL) + velocity-dependent fp drag.
+
+    Combines two proven convergence mechanisms that have NEVER been used together:
+    1. Contact Inhibition of Locomotion (Mayor & Carmona-Fontaine 2010):
+       Particles reduce mobility when local density is high.
+       f(rho) = 1 / (1 + (rho/rho_0)^n)
+    2. Velocity-dependent fp drag (Tranquillo & Lauffenburger 1987):
+       Fast-moving particles reduce chemotactic sensitivity.
+       drag = 1 / (1 + fp_drag * |v| / v_ref)
+
+    Rationale: CIL provides density-based self-limiting aggregation, but
+    clusters can oscillate as particles overshoot the density equilibrium.
+    Adding velocity drag penalizes fast oscillatory motion, potentially
+    stabilizing the CIL mechanism and improving convergence.
+
+    Physics:
+    1. fp: v = M * f(rho) * nabla_C / (1 + fp_drag * |vel|/v_ref)
+       - f(rho) = 1/(1+(rho/rho_0)^n) is Hill function CIL
+       - drag attenuates response at high speed
+    2. pf: Standard consumption/production
+    3. pp: Standard attraction-repulsion (density computed from pp graph)
+
+    Literature:
+    - Mayor, R. & Carmona-Fontaine, C. (2010) Trends in Cell Biology 20:319-328
+      "Keeping in touch with contact inhibition of locomotion"
+    - Cates, M. E. & Tailleur, J. (2015) ARCMP 6:219-244
+      "Motility-induced phase separation"
+    - Tranquillo, R. T. & Lauffenburger, D. A. (1987) J Math Biol 25:229-262
+      "Stochastic model of leukocyte chemosensory movement"
+
+    Per-type params layout: [M1, M2, consumption, production, ar_p1, ar_p2, ar_p3, ar_p4]
+    """
+
+    PARAMS_DOC = {
+        "model_name": "DensityDragCIL",
+        "literature": "Mayor & Carmona-Fontaine (2010); Cates & Tailleur (2015); Tranquillo (1987)",
+        "description": "Density-dependent CIL + velocity drag: density limits aggregation, speed limits oscillation",
+        "equations": {
+            "field_to_particle": "v = M * f(rho) * nabla_C / (1 + fp_drag * |vel|/v_ref)",
+            "density_function": "f(rho) = 1 / (1 + (rho/rho_0)^n)",
+            "particle_to_field": "dC1 = -consumption * w(r), dC2 = production * w(r)",
+            "particle_to_particle": "f = (p1*exp(-d^(2p2)/(2sigma^2)) - p3*exp(-d^(2p4)/(2sigma^2))) * dir"
+        },
+        "params_mesh": [
+            {
+                "row": 0, "description": "C1 field parameters",
+                "slots": [
+                    {"index": 0, "name": "D1", "description": "Diffusion coeff for C1"},
+                    {"index": 1, "name": "Da_c", "description": "Damkohler number"},
+                    {"index": 2, "name": "A", "description": "Brusselator A"},
+                    {"index": 3, "name": "B", "description": "Brusselator B"},
+                    {"index": 4, "name": "mu", "description": "Morphological param"},
+                    {"index": 5, "name": "M1", "description": "Mobility for C1 gradients"},
+                    {"index": 6, "name": "unused_6", "description": "Unused (pad)"},
+                    {"index": 7, "name": "unused_7", "description": "Unused (pad)"}
+                ]
+            },
+            {
+                "row": 1, "description": "C2 field parameters",
+                "slots": [
+                    {"index": 0, "name": "D2", "description": "Diffusion coeff for C2"},
+                    {"index": 1, "name": "M2", "description": "Mobility for C2 gradients"},
+                    {"index": 2, "name": "unused_2", "description": "Unused (pad)"},
+                    {"index": 3, "name": "unused_3", "description": "Unused (pad)"},
+                    {"index": 4, "name": "unused_4", "description": "Unused (pad)"},
+                    {"index": 5, "name": "unused_5", "description": "Unused (pad)"},
+                    {"index": 6, "name": "unused_6", "description": "Unused (pad)"},
+                    {"index": 7, "name": "unused_7", "description": "Unused (pad)"}
+                ]
+            },
+            {
+                "row": 2, "description": "Particle-field coupling + CIL + drag params",
+                "slots": [
+                    {"index": 0, "name": "Pe", "description": "Peclet number"},
+                    {"index": 1, "name": "consumption", "description": "Consumption rate of C1"},
+                    {"index": 2, "name": "production", "description": "Production rate of C2"},
+                    {"index": 3, "name": "influence_radius", "description": "Gaussian pf influence radius"},
+                    {"index": 4, "name": "rho_0", "description": "CIL critical density threshold"},
+                    {"index": 5, "name": "hill_n", "description": "CIL Hill coefficient"},
+                    {"index": 6, "name": "fp_drag", "description": "Velocity-dependent fp drag (0=off)"},
+                    {"index": 7, "name": "unused_7", "description": "Unused (pad)"}
+                ]
+            }
+        ],
+        "width_constraint": "ALL rows of params_mesh MUST have same number of columns (8)."
+    }
+
+    def __init__(self, aggr_type='mean', p=None, particle_params=None, bc_dpos=None, dimension=2, sigma=0.005):
+        super(PDE_D_DensityDragCIL, self).__init__(aggr=aggr_type)
+
+        self.p = p
+        self.particle_params = particle_params
+        self.bc_dpos = bc_dpos
+        self.dimension = dimension
+        self.sigma = sigma
+
+        # Global parameters from mesh
+        self.M1 = p[0, 5]
+        self.M2 = p[1, 1]
+        self.consumption_rate = p[2, 1]
+        self.production_rate = p[2, 2]
+        self.influence_radius = p[2, 3]
+        self.Pe = p[2, 0]
+        self.repulsion_strength = 50
+        self.repulsion_range = 0.04
+
+        # CIL parameters (Mayor & Carmona-Fontaine 2010)
+        self.rho_0 = p[2, 4] if p.shape[1] > 4 and p[2, 4] != 0 else 34.0
+        self.hill_n = p[2, 5] if p.shape[1] > 5 and p[2, 5] != 0 else 2.0
+        self.sensing_radius = 0.05
+
+        # Velocity-dependent fp drag (Tranquillo & Lauffenburger 1987)
+        self.fp_drag = p[2, 6] if p.shape[1] > 6 and p[2, 6] != 0 else 0.0
+        self.v_ref = 0.01
+
+        # Convert to proper tensors if needed
+        if not isinstance(self.rho_0, torch.Tensor):
+            self.rho_0 = torch.tensor(float(self.rho_0), device=p.device)
+        if not isinstance(self.hill_n, torch.Tensor):
+            self.hill_n = torch.tensor(float(self.hill_n), device=p.device)
+        if not isinstance(self.fp_drag, torch.Tensor):
+            self.fp_drag = torch.tensor(float(self.fp_drag), device=p.device)
+
+        # Storage for local density (computed in pp pass, used in fp pass)
+        self.local_density = None
+
+        # Report configuration
+        rho0_val = self.rho_0.item() if hasattr(self.rho_0, 'item') else self.rho_0
+        hill_val = self.hill_n.item() if hasattr(self.hill_n, 'item') else self.hill_n
+        drag_val = self.fp_drag.item() if hasattr(self.fp_drag, 'item') else self.fp_drag
+        print(f"initialized PDE_D_DensityDragCIL with parameters:")
+        print(f"  mobility: M1={self.M1.item()}, M2={self.M2.item()}")
+        print(f"  CIL: rho_0={rho0_val}, hill_n={hill_val}, sensing_radius={self.sensing_radius} (Mayor 2010)")
+        print(f"  fp_drag={drag_val:.3f}, v_ref={self.v_ref:.4f} (Tranquillo 1987)")
+        print(f"  Pe={self.Pe.item():.3f}, sigma={self.sigma}")
+        print(f"  particle->field: consumption={self.consumption_rate.item()}, production={self.production_rate.item()}, influence_radius={self.influence_radius.item():.3f}")
+        if particle_params is not None:
+            print(f"  multi-type support: {particle_params.shape[0]} particle types")
+
+    def forward(self, data, direction='fp'):
+        x, edge_index = data.x, data.edge_index
+        edge_index, _ = pyg_utils.remove_self_loops(edge_index)
+
+        if self.particle_params is not None:
+            particle_type = x[:, 1 + 2*self.dimension].long()
+            max_type = particle_type.max().item()
+            n_param_rows = self.particle_params.shape[0]
+            if max_type >= n_param_rows:
+                raise ValueError(
+                    f"PDE_D_DensityDragCIL: particle_params has {n_param_rows} rows but found "
+                    f"particle type {max_type}. Need {max_type + 1} rows in simulation.params."
+                )
+            parameters = self.particle_params[to_numpy(particle_type), :]
+        else:
+            parameters = None
+
+        if direction == 'interpolate':
+            result = self.propagate(edge_index, x=x, mode='interpolate', parameters=parameters)
+            pos = x[:, 1:self.dimension+1]
+            in_box = ((pos >= 0) & (pos <= 1)).all(dim=1, keepdim=True)
+            result = result * in_box.float()
+            return result
+        elif direction == 'fp':
+            result = self.propagate(edge_index, x=x, mode='fp', parameters=parameters)
+
+            # Apply density-dependent + velocity drag modulation
+            if self.local_density is not None:
+                n_total = x.size(0)
+                n_particles = self.local_density.size(0)
+                n_nodes = n_total - n_particles
+
+                # CIL Hill function modulation
+                ratio = self.local_density / self.rho_0
+                cil_modulation = 1.0 / (1.0 + ratio ** self.hill_n)
+
+                # Apply to particle portion only
+                mod_full = torch.ones(n_total, 1, device=x.device)
+                mod_full[n_nodes:, 0] = cil_modulation
+                result = result * mod_full
+
+            # Apply velocity-dependent drag at the aggregate level
+            if self.fp_drag > 0:
+                n_total = x.size(0)
+                vel = x[:, 1+self.dimension:1+2*self.dimension]
+                speed = torch.sqrt(torch.sum(vel**2, dim=1, keepdim=True))
+                drag_factor = 1.0 / (1.0 + self.fp_drag * speed / self.v_ref)
+                result = result * drag_factor
+
+            pos = x[:, 1:self.dimension+1]
+            in_box = ((pos >= 0) & (pos <= 1)).all(dim=1, keepdim=True)
+            result = result * in_box.float()
+            return result
+        elif direction == 'pf':
+            result = self.propagate(edge_index, x=x, mode='pf', parameters=parameters)
+            return result
+        else:  # direction == 'pp'
+            self._compute_local_density(x, edge_index)
+            result = self.propagate(edge_index, x=x, mode='pp', parameters=parameters)
+            return result
+
+    def _compute_local_density(self, x, edge_index):
+        """Count particle neighbors within sensing_radius for CIL."""
+        n_particles = x.size(0)
+        target_nodes = edge_index[1]
+
+        pos_i = x[edge_index[1], 1:self.dimension+1]
+        pos_j = x[edge_index[0], 1:self.dimension+1]
+        d_pos = self.bc_dpos(pos_j - pos_i)
+        dist = torch.sqrt(torch.sum(d_pos**2, dim=1))
+
+        within_radius = dist < self.sensing_radius
+        counts = torch.zeros(n_particles, device=x.device)
+        counts.scatter_add_(0, target_nodes[within_radius],
+                           torch.ones(within_radius.sum(), device=x.device))
+
+        self.local_density = counts
+
+    def message(self, edge_index_i, edge_index_j, x_i, x_j, mode=None, parameters_i=None):
+        pos_i = x_i[:, 1:self.dimension+1]
+        pos_j = x_j[:, 1:self.dimension+1]
+
+        d_pos = self.bc_dpos(pos_j - pos_i)
+        dist = torch.sqrt(torch.sum(d_pos**2, dim=1))
+        dist_safe = torch.clamp(dist, min=1e-6)
+
+        if mode == 'interpolate':
+            C1_mesh = x_j[:, 6:7]
+            C2_mesh = x_j[:, 7:8]
+            weight = torch.exp(-dist / 0.01).unsqueeze(1)
+            return torch.cat([C1_mesh * weight, C2_mesh * weight, weight], dim=1)
+
+        elif mode == 'fp':
+            fields_i = x_i[:, 6:8]
+            fields_j = x_j[:, 6:8]
+
+            dC1 = fields_j[:, 0:1] - fields_i[:, 0:1]
+            dC2 = fields_j[:, 1:2] - fields_i[:, 1:2]
+
+            kernel = torch.exp(-dist / 0.05)
+            dir_norm = d_pos / dist_safe.unsqueeze(1)
+            domain_scale = 32.0
+            grad_C1 = (dC1 * kernel.unsqueeze(1)) / (dist_safe.unsqueeze(1) * domain_scale)
+            grad_C2 = (dC2 * kernel.unsqueeze(1)) / (dist_safe.unsqueeze(1) * domain_scale)
+
+            if parameters_i is not None:
+                M1 = parameters_i[:, 0:1]
+                M2 = parameters_i[:, 1:2]
+            else:
+                M1 = self.M1
+                M2 = self.M2
+
+            velocities = (M1 * grad_C1 + M2 * grad_C2) * dir_norm
+            return velocities
+
+        elif mode == 'pf':
+            weights = torch.exp(-dist**2 / (2 * (self.influence_radius/3)**2))
+
+            if parameters_i is not None:
+                consumption = parameters_i[:, 2]
+                production = parameters_i[:, 3]
+            else:
+                consumption = self.consumption_rate
+                production = self.production_rate
+
+            field_updates = torch.zeros((pos_i.size(0), 2), device=pos_i.device)
+            field_updates[:, 0] = -consumption * weights
+            field_updates[:, 1] = production * weights
+            return field_updates
+
+        else:  # mode == 'pp'
+            if parameters_i is not None:
+                p1 = parameters_i[:, 4]
+                p2 = parameters_i[:, 5]
+                p3 = parameters_i[:, 6]
+                p4 = parameters_i[:, 7]
+
+                f = (p1 * torch.exp(-dist ** (2 * p2) / (2 * self.sigma ** 2))
+                     - p3 * torch.exp(-dist ** (2 * p4) / (2 * self.sigma ** 2)))
+
+                forces = f[:, None] * d_pos / dist_safe.unsqueeze(1)
+            else:
+                forces = torch.zeros_like(pos_i)
+                in_range = dist < self.repulsion_range
+                if in_range.any():
+                    dir_norm = d_pos / dist_safe.unsqueeze(1)
+                    repulsion_mag = self.repulsion_strength * torch.exp(
+                        -5.0 * dist[in_range] / self.repulsion_range
+                    )
+                    forces[in_range] = -dir_norm[in_range] * repulsion_mag.unsqueeze(1)
+
+            return forces
+
+    def update(self, aggr_out, mode=None):
+        if mode == 'interpolate':
+            C1_weighted = aggr_out[:, 0:1]
+            C2_weighted = aggr_out[:, 1:2]
+            weight_sum = aggr_out[:, 2:3]
+            weight_sum = torch.clamp(weight_sum, min=1e-10)
+            return torch.cat([C1_weighted / weight_sum, C2_weighted / weight_sum], dim=1)
+        else:
+            return aggr_out