Spaces:
Paused
Paused
File size: 9,987 Bytes
5e0b58b | 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 | """Interactive Brain 5D observatory with optional X-Y heatmap."""
from __future__ import annotations
from pathlib import Path
from typing import Any
import matplotlib.pyplot as plt
import numpy as np
from src.core.spatial_index import unpack_coords
from src.visualization.heatmap import HeatmapKind, HeatmapProjector, HeatmapView
def build_raster_points(frames: Any, sample_ids: list[int]) -> list[tuple[int, int]]:
"""Convert spike-history frames to raster plot points."""
id_to_row = {neuron_id: row for row, neuron_id in enumerate(sample_ids)}
points: list[tuple[int, int]] = []
for frame in frames:
for neuron_id in frame.neuron_ids:
row = id_to_row.get(neuron_id)
if row is not None:
points.append((frame.tick, row))
return points
class Observatory:
"""Interactive visual observer for sparse Brain 5D simulations."""
def __init__(
self,
network: Any,
config: dict[str, Any],
spike_history: Any,
history: Any,
probes: Any | None = None,
) -> None:
self.net = network
self.config = config
self.spike_history = spike_history
self.history = history
self.probes = probes
self.dims = tuple(config["dimensions"])
visualization = config["visualization"]
self.raster_n = int(visualization["spike_raster_neurons"])
self.proj_dim = str(visualization.get("projection_4d", "d4"))
self.tau = float(visualization.get("activity_tau_ticks", 50.0))
self.show_heatmap = bool(visualization.get("show_heatmap", True))
self.heatmap_kind: HeatmapKind = self._parse_heatmap_kind(
visualization.get("heatmap_type", "activity")
)
plt.ion() # type: ignore[reportUnknownMemberType]
self.fig = plt.figure(figsize=(16, 10)) # type: ignore[reportUnknownMemberType]
grid = self.fig.add_gridspec(2, 3) # type: ignore[reportUnknownMemberType]
self.ax1 = self.fig.add_subplot(grid[0, 0], projection="3d") # type: ignore[reportUnknownMemberType]
self.ax2 = self.fig.add_subplot(grid[0, 1]) # type: ignore[reportUnknownMemberType]
self.ax_heat = self.fig.add_subplot(grid[0, 2]) # type: ignore[reportUnknownMemberType]
self.ax3 = self.fig.add_subplot(grid[1, 0]) # type: ignore[reportUnknownMemberType]
self.ax4 = self.fig.add_subplot(grid[1, 1:]) # type: ignore[reportUnknownMemberType]
self.scatter_xyz = self.ax1.scatter( # type: ignore[reportUnknownMemberType]
[], [], [], c=[], cmap="hot", vmin=0, vmax=1, s=10, alpha=0.7 # type: ignore[reportArgumentType]
)
self.scatter_xd = self.ax2.scatter( # type: ignore[reportUnknownMemberType]
[], [], c=[], cmap="plasma", vmin=0, vmax=1, s=8, alpha=0.7
)
self.raster_scatter = self.ax3.scatter( # type: ignore[reportUnknownMemberType]
[], [], s=8, marker="|"
)
(self.line_spikes,) = self.ax4.plot([], [], label="Spikes/tick") # type: ignore[reportUnknownMemberType]
(self.line_v,) = self.ax4.plot([], [], label="Mean V") # type: ignore[reportUnknownMemberType]
(self.line_queue,) = self.ax4.plot([], [], label="Queue") # type: ignore[reportUnknownMemberType]
self.ax4.legend(fontsize=8) # type: ignore[reportUnknownMemberType]
self.ax1.set_title("XYZ Activity") # type: ignore[reportUnknownMemberType]
self.ax2.set_title(f"X vs {self.proj_dim.upper()}") # type: ignore[reportUnknownMemberType]
self.ax3.set_title("Real Spike Raster") # type: ignore[reportUnknownMemberType]
self.ax4.set_title("Time Series") # type: ignore[reportUnknownMemberType]
self.heatmap_projector = HeatmapProjector(self.net, self.tau)
self.heatmap_view = HeatmapView(self.ax_heat)
if not self.show_heatmap:
self.ax_heat.set_visible(False)
self.status_text = self.fig.text(0.02, 0.02, "", family="monospace", fontsize=9) # type: ignore[reportUnknownMemberType]
self.probe_text = self.fig.text(0.72, 0.02, "", family="monospace", fontsize=9) # type: ignore[reportUnknownMemberType]
self.snapshot_dir = Path("artifacts/snapshots")
self.snapshot_dir.mkdir(parents=True, exist_ok=True)
self.fig.canvas.mpl_connect("key_press_event", self._on_key)
plt.show(block=False) # type: ignore[reportUnknownMemberType]
@staticmethod
def _parse_heatmap_kind(value: object) -> HeatmapKind:
kind = str(value).lower()
if kind not in {"activity", "weights", "energy"}:
raise ValueError(
"visualization.heatmap_type must be activity, weights or energy"
)
return kind # type: ignore[return-value]
def _on_key(self, event: Any) -> None:
if event.key and str(event.key).lower() == "s":
self.save_snapshot()
def _activity(self, neuron: Any) -> float:
if neuron.last_spike_tick < 0:
return 0.0
age = max(0, self.net.current_tick - neuron.last_spike_tick)
return float(np.exp(-age / self.tau))
def draw(self) -> None:
"""Refresh all Observatory panels from the current network state."""
sample = list(self.net.neurons.items())[:2000]
x_values: list[int] = []
y_values: list[int] = []
z_values: list[int] = []
activity: list[float] = []
xd_values: list[int] = []
yd_values: list[int] = []
activity_2d: list[float] = []
for neuron_id, neuron in sample:
x_coord, y_coord, z_coord, d4_coord, d5_coord = unpack_coords(neuron_id)
value = self._activity(neuron)
x_values.append(x_coord)
y_values.append(y_coord)
z_values.append(z_coord)
activity.append(value)
xd_values.append(x_coord)
yd_values.append(d4_coord if self.proj_dim == "d4" else d5_coord)
activity_2d.append(value)
self.scatter_xyz._offsets3d = (x_values, y_values, z_values) # type: ignore[reportAttributeAccessIssue]
self.scatter_xyz.set_array(np.asarray(activity)) # type: ignore[reportUnknownMemberType]
points_2d = (
np.column_stack((xd_values, yd_values)) if xd_values else np.empty((0, 2))
)
self.scatter_xd.set_offsets(points_2d) # type: ignore[reportUnknownMemberType]
self.scatter_xd.set_array(np.asarray(activity_2d)) # type: ignore[reportUnknownMemberType]
self.ax1.set( # type: ignore[reportUnknownMemberType]
xlim=(0, self.dims[0] - 1),
ylim=(0, self.dims[1] - 1),
zlim=(0, self.dims[2] - 1),
)
projection_size = self.dims[3] if self.proj_dim == "d4" else self.dims[4]
self.ax2.set( # type: ignore[reportUnknownMemberType]
xlim=(0, self.dims[0] - 1),
ylim=(0, projection_size - 1),
)
frames = self.spike_history.get_frames()[-100:]
sample_ids = list(self.net.neurons)[: self.raster_n]
raster_points = build_raster_points(frames, sample_ids)
self.raster_scatter.set_offsets( # type: ignore[reportUnknownMemberType]
np.asarray(raster_points) if raster_points else np.empty((0, 2))
)
if frames:
self.ax3.set_xlim( # type: ignore[reportUnknownMemberType]
frames[0].tick,
max(frames[0].tick + 1, frames[-1].tick + 1),
)
self.ax3.set_ylim(0, max(1, len(sample_ids))) # type: ignore[reportUnknownMemberType]
history = self.history.get_all()
if history:
ticks = [item["tick"] for item in history]
self.line_spikes.set_data( # type: ignore[reportUnknownMemberType]
ticks, [item["spikes_this_tick"] for item in history]
)
self.line_v.set_data( # type: ignore[reportUnknownMemberType]
ticks, [item["mean_v"] for item in history]
)
self.line_queue.set_data( # type: ignore[reportUnknownMemberType]
ticks, [item["queued_events"] for item in history]
)
self.ax4.relim() # type: ignore[reportUnknownMemberType]
self.ax4.autoscale_view() # type: ignore[reportUnknownMemberType]
if self.show_heatmap:
self.heatmap_view.render(self.heatmap_projector.build(self.heatmap_kind))
self.status_text.set_text( # type: ignore[reportUnknownMemberType]
f"Tick {self.net.current_tick} | neurons {len(self.net.neurons)} | "
f"synapses {self.net.synapse_count} | queue {self.net.queued_event_count} | "
f"spikes {self.net.total_spikes}"
)
if self.probes and self.probes.probes:
rows: list[str] = []
for probe_id in self.probes.probes[:3]:
data = self.probes.get_probe_data(probe_id)
rows.append(
f"{probe_id}: v={data.get('v', 0):.1f} "
f"spk={data.get('spike_counter', 0)}"
)
self.probe_text.set_text("Probes\n" + "\n".join(rows)) # type: ignore[reportUnknownMemberType]
self.fig.canvas.draw_idle() # type: ignore[reportUnknownMemberType]
plt.pause(0.001) # type: ignore[reportUnknownMemberType]
def save_snapshot(self) -> Path:
"""Save the current Observatory figure and return its path."""
path = self.snapshot_dir / f"brain5d_tick_{self.net.current_tick:06d}.png"
self.fig.savefig(path, dpi=150, bbox_inches="tight") # type: ignore[reportUnknownMemberType]
return path
@staticmethod
def block_until_closed() -> None:
"""Switch Matplotlib to blocking mode until the window is closed."""
plt.ioff() # type: ignore[reportUnknownMemberType]
plt.show(block=True) # type: ignore[reportUnknownMemberType]
|