1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223 | """Generate a mixed-field layout that blends multiple distributions."""
import random
import math
from ..config import load_distribution
from ..patterns.csr import _point_in_area, _check_min_distance, ProximityGrid
from ..patterns.clustered import _scatter_gaussian, _SCATTER_FNS
def _sigmoid(x, center, width):
"""Compute a smooth sigmoid weight in the range [0, 1]."""
t = (x - center) / max(width, 1e-6)
return 1.0 / (1.0 + math.exp(-t))
def _clamp01(v):
if v <= 0.0:
return 0.0
if v >= 1.0:
return 1.0
return v
def _field_value(kind, x, y, K, cfg):
if kind == "sigmoid_x":
v = _sigmoid(x, cfg.get("center_x", 0.0), cfg.get("width", 1.0))
elif kind == "sigmoid_y":
v = _sigmoid(y, cfg.get("center_y", 0.0), cfg.get("width", 1.0))
elif kind == "radial":
cx = cfg.get("center_x", 0.0)
cy = cfg.get("center_y", 0.0)
radius = cfg.get("radius", K * 0.25)
width = cfg.get("width", max(radius * 0.25, 1e-6))
dist = math.hypot(x - cx, y - cy)
v = _sigmoid(radius - dist, 0.0, width)
elif kind == "linear_x":
x_min = cfg.get("x_min", -K / 2)
x_max = cfg.get("x_max", K / 2)
denom = max(x_max - x_min, 1e-6)
v = (x - x_min) / denom
elif kind == "linear_y":
y_min = cfg.get("y_min", -K / 2)
y_max = cfg.get("y_max", K / 2)
denom = max(y_max - y_min, 1e-6)
v = (y - y_min) / denom
else:
v = 1.0
if cfg.get("invert"):
v = 1.0 - v
v = v * cfg.get("scale", 1.0) + cfg.get("bias", 0.0)
return _clamp01(v)
def _build_candidate_sampler(dist_type, params, K):
"""Build a candidate point sampler for mixture-field placement.
For ``regular`` we still return uniform candidates - the hard min-distance
constraint is already enforced by the outer placement loop in
``generate_mixed_field``. The spatial inhibition in the standalone
``sample_regular`` (Bridson Poisson-disc) is used when the regular
pattern owns the whole zone / world; inside a mixture field the outer
rejection loop handles it.
"""
if dist_type in ("csr", "regular"):
return lambda: _point_in_area(K)
if dist_type in ("clustered", "scale_dependent"):
cluster_count = max(1, int(params.get("cluster_count", 5)))
cluster_radius = float(params.get("cluster_radius", 3.0))
scatter_shape = params.get("scatter_shape", "gaussian")
allow_overlap = bool(params.get("allow_cluster_overlap", False))
min_parent_dist = (
0.0 if allow_overlap else float(params.get("min_parent_distance", cluster_radius * 0.5))
)
scatter_fn = _SCATTER_FNS.get(scatter_shape, _scatter_gaussian)
parents = []
for _ in range(cluster_count):
for _ in range(60):
cx, cy = _point_in_area(K)
if min_parent_dist <= 0 or _check_min_distance(cx, cy, parents, min_parent_dist):
parents.append((cx, cy))
break
else:
parents.append(_point_in_area(K))
def _sample_clustered():
if parents:
cx, cy = random.choice(parents)
x, y = scatter_fn(cx, cy, cluster_radius)
x = max(-K / 2, min(K / 2, x))
y = max(-K / 2, min(K / 2, y))
return x, y
return _point_in_area(K)
return _sample_clustered
raise ValueError(f"Unknown distribution type: {dist_type}")
def _weighted_choice(weights, total_weight):
r = random.random() * total_weight
acc = 0.0
for idx, w in enumerate(weights):
acc += w
if r <= acc:
return idx
return len(weights) - 1
def generate_mixed_field(layout_config, world_config, project_root):
"""Blend multiple distributions across the world using a spatial field."""
# World-level params always come from world.default.yaml (single source of truth)
K = world_config["generation"]["area_size"]
min_distance = world_config["generation"]["min_distance"]
total_count = world_config["generation"]["object_count"]
components = layout_config["layout"].get("components", [])
field_cfg = layout_config["layout"].get("field", {}) or {}
max_attempts = int(layout_config["layout"].get("max_attempts", 200))
dists = []
for comp in components:
dc = load_distribution(comp["distribution_ref"], project_root)
dist_type = dc["distribution"]["type"]
dist_params = dc["distribution"].get("params", {}) or {}
dists.append(
{
"name": comp["name"],
"type": dist_type,
"params": dist_params,
"weight": float(comp.get("weight", 1.0)),
"field": comp.get("field"),
"sampler": _build_candidate_sampler(dist_type, dist_params, K),
}
)
if not dists:
return []
name_to_index = {d["name"]: i for i, d in enumerate(dists)}
component_fields = {}
for fc in field_cfg.get("components", []) or []:
ref = fc.get("component")
if isinstance(ref, int) and 0 <= ref < len(dists):
component_fields[ref] = fc
elif ref in name_to_index:
component_fields[name_to_index[ref]] = fc
def _component_field_weight(i, x, y):
if dists[i].get("field") is not None:
fc = dists[i]["field"]
return _field_value(fc.get("kind", "constant"), x, y, K, fc)
if i in component_fields:
fc = component_fields[i]
return _field_value(fc.get("kind", "constant"), x, y, K, fc)
if "kind" in field_cfg:
base = _field_value(field_cfg.get("kind", "constant"), x, y, K, field_cfg)
if len(dists) == 1:
return 1.0
if i == 0:
return base
return _clamp01((1.0 - base) / max(len(dists) - 1, 1))
return 1.0
base_weights = [max(0.0, d["weight"]) for d in dists]
total_weight = sum(base_weights)
if total_weight <= 0.0:
base_weights = [1.0] * len(dists)
total_weight = float(len(dists))
all_positions = []
counts = [0] * len(dists)
warned = False
grid = ProximityGrid(min_distance) if min_distance > 0 else None
forced_relaxations = 0
for placed_count in range(total_count):
progress = placed_count / max(total_count, 1)
if progress >= 0.95:
attempt_budget = min(max_attempts, 20)
elif progress >= 0.90:
attempt_budget = min(max_attempts, 30)
elif progress >= 0.80:
attempt_budget = min(max_attempts, 60)
else:
attempt_budget = max_attempts
if forced_relaxations >= 8:
attempt_budget = min(attempt_budget, 25)
placed = False
for _ in range(attempt_budget):
chosen = _weighted_choice(base_weights, total_weight)
x, y = dists[chosen]["sampler"]()
if random.random() > _component_field_weight(chosen, x, y):
continue
if grid is not None and not grid.check(x, y, min_distance):
continue
all_positions.append((x, y))
if grid is not None:
grid.insert(x, y)
counts[chosen] += 1
placed = True
forced_relaxations = max(0, forced_relaxations - 1)
break
if not placed:
chosen = _weighted_choice(base_weights, total_weight)
x, y = dists[chosen]["sampler"]()
all_positions.append((x, y))
if grid is not None:
grid.insert(x, y)
counts[chosen] += 1
forced_relaxations += 1
if not warned:
print("Warning: relaxed placement constraints in mixture_field")
warned = True
summary = ", ".join(f"{counts[i]} '{d['name']}'" for i, d in enumerate(dists))
print(f" mixture_field: {summary}")
return all_positions
|