Spaces:
Runtime error
Runtime error
File size: 20,724 Bytes
f106663 d6abeb6 f106663 d6abeb6 f106663 d6abeb6 f106663 d6abeb6 f106663 d6abeb6 f106663 d6abeb6 f106663 d6abeb6 f106663 e2e1a8c f106663 e2e1a8c f106663 e2e1a8c f106663 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 |
import numpy as np
import math
from qiskit.circuit import QuantumCircuit, QuantumRegister
from qiskit.circuit.library import StatePreparation, QFTGate, RZGate
from qiskit.quantum_info import Statevector
import pyvista as pv
def create_impulse_state(grid_dims, impulse_pos):
"""
Creates an initial state vector with a single delta impulse at a specified grid position.
The 2D grid is flattened into a 1D vector in row-major order, and this
vector is then padded to match the full simulation state space size (4x).
Args:
grid_dims (tuple): A tuple (width, height) defining the simulation grid dimensions.
For your original code, this would be (nx, nx).
impulse_pos (tuple): A tuple (x, y) for the position of the impulse.
Coordinates are 0-indexed.
Returns:
numpy.ndarray: The full, padded initial state vector with a single 1.
Raises:
ValueError: If the impulse position is outside the grid dimensions.
"""
grid_width, grid_height = grid_dims
impulse_x, impulse_y = impulse_pos
# --- Input Validation ---
# Ensure the requested impulse position is actually on the grid.
if not (0 <= impulse_x < grid_width and 0 <= impulse_y < grid_height):
raise ValueError(f"Impulse position ({impulse_x}, {impulse_y}) is outside the "
f"grid dimensions ({grid_width}x{grid_height}).")
# --- 1. Calculate the 1D Array Index ---
# Convert the (x, y) coordinate to a single index in a flattened 1D array.
# The formula for row-major order is: index = y_coord * width + x_coord
flat_index = impulse_y * grid_width + impulse_x
# --- 2. Create the Full, Padded State Vector ---
grid_size = grid_width * grid_height
total_size = 4 * grid_size # The simulation space is 4x the grid size.
initial_state = np.zeros(total_size)
# --- 3. Set the Delta Impulse ---
initial_state[flat_index] = 1
return initial_state
def create_gaussian_state(grid_dims, mu, sigma):
"""
Creates an initial state vector with a 2D Gaussian distribution.
The state is normalized and padded to match the full simulation state space size (4x).
Args:
grid_dims (tuple): A tuple (width, height) defining the grid dimensions.
mu (tuple): A tuple (mu_x, mu_y) for the center (mean) of the Gaussian.
sigma (tuple): A tuple (sigma_x, sigma_y) for the standard deviation (spread).
Returns:
numpy.ndarray: The full, padded initial state vector for the Gaussian state.
Raises:
ValueError: If sigma values are not positive.
"""
grid_width, grid_height = grid_dims
mu_x, mu_y = mu
sigma_x, sigma_y = sigma
if sigma_x <= 0 or sigma_y <= 0:
raise ValueError("Sigma values (spread) must be positive.")
# --- 1. Create a Coordinate Grid ---
x = np.arange(0, grid_width)
y = np.arange(0, grid_height)
X, Y = np.meshgrid(x, y)
# --- 2. Calculate the 2D Gaussian Function ---
gaussian_2d = np.exp(-((X - mu_x)**2 / (2 * sigma_x**2)) -
((Y - mu_y)**2 / (2 * sigma_y**2)))
# --- 3. Normalize the State Vector ---
# For a valid quantum state, the L2 norm (sum of squares of amplitudes) must be 1.
norm = np.linalg.norm(gaussian_2d)
if norm > 0:
gaussian_2d = gaussian_2d / norm
# --- 4. Flatten and Pad the Vector ---
gaussian_flat = gaussian_2d.flatten()
grid_size = grid_width * grid_height
total_size = 4 * grid_size
initial_state = np.pad(gaussian_flat, (0, total_size - grid_size), mode='constant')
return initial_state
# --- New: Continuous-position helpers for excitation before meshing ---
def _normalize_to_unit(vec: np.ndarray) -> np.ndarray:
n = np.linalg.norm(vec)
return vec / n if n > 0 else vec
def create_impulse_state_from_pos(grid_dims, pos01, snap_to_grid=True):
"""
Create a delta-like initial state from continuous position pos01=(x,y) in [0,1].
Args:
grid_dims: Tuple (nx, ny) defining the simulation grid dimensions.
pos01: Tuple (x, y) with continuous position in [0, 1] range.
snap_to_grid: If True (default), snaps the impulse to the nearest grid node
for an exact delta with peak value = 1.0. If False, uses
bilinear interpolation which distributes amplitude to 4 nodes.
Why grid_dims?
- Simulation runs on a discrete nx×ny lattice; the continuous position must be
discretized onto that grid to produce the state vector fed into the solver.
- grid_dims provides (nx, ny) so we can map (x,y)∈[0,1]→grid coordinates.
When snap_to_grid=True (default):
- The impulse is placed exactly on the nearest grid node.
- Peak amplitude = 1.0 (exact delta function on the discrete grid).
- This is recommended for visualization and accurate peak value display.
When snap_to_grid=False:
- Uses bilinear interpolation to distribute amplitude to the 4 neighboring nodes.
- Peak amplitude depends on position (e.g., 0.5 when exactly between 4 nodes).
- Total energy is preserved (L2 norm = 1).
The preview uses create_impulse_preview_state(), which renders a smooth bump on a
fixed unit-square grid independent of nx for visualization.
"""
grid_width, grid_height = grid_dims
px, py = pos01
px = float(max(0.0, min(1.0, px)))
py = float(max(0.0, min(1.0, py)))
gx = px * (grid_width - 1)
gy = py * (grid_height - 1)
grid_size = grid_width * grid_height
total_size = 4 * grid_size
if snap_to_grid:
# Snap to nearest grid node for exact delta with peak = 1.0
i0 = int(round(gx))
j0 = int(round(gy))
i0 = max(0, min(i0, grid_width - 1))
j0 = max(0, min(j0, grid_height - 1))
field = np.zeros(grid_size)
field[j0 * grid_width + i0] = 1.0 # Peak value = 1.0
initial_state = np.zeros(total_size)
initial_state[:grid_size] = field
return initial_state
# Bilinear interpolation mode (snap_to_grid=False)
i0, j0 = int(np.floor(gx)), int(np.floor(gy))
i1, j1 = min(i0 + 1, grid_width - 1), min(j0 + 1, grid_height - 1)
dx, dy = gx - i0, gy - j0
w00 = (1 - dx) * (1 - dy)
w10 = dx * (1 - dy)
w01 = (1 - dx) * dy
w11 = dx * dy
field = np.zeros(grid_size)
field[j0 * grid_width + i0] += w00
field[j0 * grid_width + i1] += w10
field[j1 * grid_width + i0] += w01
field[j1 * grid_width + i1] += w11
field = _normalize_to_unit(field)
initial_state = np.zeros(total_size)
initial_state[:grid_size] = field
return initial_state
def create_gaussian_state_from_pos(grid_dims, mu01, sigma01, snap_to_grid=True):
"""
Create a Gaussian initial state with center mu01=(x,y) and spreads sigma01=(sx,sy)
in [0,1] of the domain, then discretize to the solver grid given by grid_dims.
Args:
grid_dims: Tuple (nx, ny) defining the simulation grid dimensions.
mu01: Tuple (x, y) with Gaussian center in [0, 1].
sigma01: Tuple (sx, sy) with Gaussian std dev as fraction of domain in [0, 1].
snap_to_grid: If True (default), snap the Gaussian center to the nearest grid
node before discretization. This makes the simulator behavior consistent
with the "nearest-node" semantics often used elsewhere in the app.
Why grid_dims?
- The quantum solver expects a vector aligned to the chosen nx×ny simulation grid.
We convert normalized μ and σ (fractions of the domain) into grid units using
(nx-1) and (ny-1). This step is necessary for the simulation, not for the preview.
For preview-only rendering, use create_impulse_preview_state() to keep the visuals
continuous and independent of nx.
"""
grid_width, grid_height = grid_dims
mu_x01, mu_y01 = mu01
sig_x01, sig_y01 = sigma01
mu_x01 = float(max(0.0, min(1.0, mu_x01)))
mu_y01 = float(max(0.0, min(1.0, mu_y01)))
sig_x01 = float(sig_x01)
sig_y01 = float(sig_y01)
if sig_x01 <= 0 or sig_y01 <= 0:
raise ValueError("Sigma values (spread) must be positive.")
mu_x = mu_x01 * (grid_width - 1)
mu_y = mu_y01 * (grid_height - 1)
if snap_to_grid:
mu_x = float(int(round(mu_x)))
mu_y = float(int(round(mu_y)))
mu_x = float(max(0, min(int(mu_x), grid_width - 1)))
mu_y = float(max(0, min(int(mu_y), grid_height - 1)))
sigma_x = sig_x01 * (grid_width - 1)
sigma_y = sig_y01 * (grid_height - 1)
x = np.arange(0, grid_width)
y = np.arange(0, grid_height)
X, Y = np.meshgrid(x, y)
gaussian_2d = np.exp(-((X - mu_x) ** 2) / (2 * sigma_x ** 2) - ((Y - mu_y) ** 2) / (2 * sigma_y ** 2))
field = _normalize_to_unit(gaussian_2d.ravel())
grid_size = grid_width * grid_height
total_size = 4 * grid_size
initial_state = np.zeros(total_size)
initial_state[:grid_size] = field
return initial_state
# --- Simulation Code (from previous context) ---
def Wj_block(j, n, ctrl_state, theta, lam, name='Wj_block', xgate=False):
qc = QuantumCircuit(n + j, name=name)
if j > 1: qc.cx(n + j - 1, range(n, n + j - 1))
if lam != 0: qc.p(lam, n + j - 1)
qc.h(n + j - 1)
if xgate and j > 1:
if isinstance(xgate, (list, tuple)):
for idx, flag in enumerate(xgate):
if flag: qc.x(n + idx)
elif xgate is True: qc.x(range(n, n + j - 1))
if j > 1:
mcrz = RZGate(theta).control(len(ctrl_state) + j - 1, ctrl_state="1" * (j - 1) + ctrl_state)
qc.append(mcrz, range(0, n + j))
else:
mcrz = RZGate(theta).control(len(ctrl_state), ctrl_state=ctrl_state)
qc.append(mcrz, range(0, n + j))
if xgate and j > 1:
if isinstance(xgate, (list, tuple)):
for idx, flag in enumerate(xgate):
if flag: qc.x(n + idx)
elif xgate is True: qc.x(range(n, n + j - 1))
qc.h(n + j - 1)
if lam != 0: qc.p(-lam, n + j - 1)
if j > 1: qc.cx(n + j - 1, range(n, n + j - 1))
return qc.to_gate(label=name)
def V1(nx, dt):
n = int(np.ceil(np.log2(nx)))
derivatives, blocks = QuantumRegister(2 * n), QuantumRegister(2)
qc = QuantumCircuit(derivatives, blocks)
qc.append(Wj_block(2, n, "0" * n, -dt, 0, xgate=True), list(derivatives[0:n]) + list(blocks[:]))
qc.append(Wj_block(3, n - 1, "1" * (n - 1), dt, 0, xgate=[0, 1]), list(derivatives[1:n]) + [derivatives[0]] + list(blocks[:]))
qc.append(Wj_block(1, n + 1, "0" * (n + 1), dt, 0, xgate=True), list(derivatives[n:2 * n]) + list(blocks[:]))
qc.append(Wj_block(2, n, "0" + "1" * (n - 1), -dt, 0, xgate=False), list(derivatives[n + 1:2 * n]) + [blocks[0]] + [derivatives[n]] + [blocks[1]])
return qc
def V2(nx, dt):
n = int(np.ceil(np.log2(nx)))
derivatives, blocks = QuantumRegister(2 * n), QuantumRegister(2)
qc = QuantumCircuit(derivatives, blocks)
qc.append(Wj_block(2, 0, "", -2 * dt, -np.pi / 2, xgate=True), blocks[:])
for j in range(1, n + 1): qc.append(Wj_block(2 + j, 0, "", 2 * dt, -np.pi / 2, xgate=[1] * (j - 1) + [0, 1]), list(derivatives[0:j]) + list(blocks[:]))
qc.append(Wj_block(2, n, "0" * n, -dt, -np.pi / 2, xgate=True), list(derivatives[0:n]) + list(blocks[:]))
qc.append(Wj_block(2, n, "1" * n, 2 * dt, -np.pi / 2, xgate=True), list(derivatives[0:n]) + list(blocks[:]))
qc.append(Wj_block(3, n - 1, "1" * (n - 1), dt, -np.pi / 2, xgate=[0, 1]), list(derivatives[1:n]) + [derivatives[0]] + list(blocks[:]))
qc.append(Wj_block(1, 1, "0", 2 * dt, -np.pi / 2, xgate=False), blocks[:])
for j in range(1, n + 1): qc.append(Wj_block(1 + j, 1, "0", -2 * dt, -np.pi / 2, xgate=[1] * (j - 1)), [blocks[0]] + list(derivatives[n:n + j]) + [blocks[1]])
qc.append(Wj_block(1, n + 1, "0" * (n + 1), dt, -np.pi / 2, xgate=False), list(derivatives[n:2 * n]) + list(blocks[:]))
qc.append(Wj_block(1, n + 1, "0" + "1" * n, -2 * dt, -np.pi / 2, xgate=False), list(derivatives[n:2 * n]) + list(blocks[:]))
qc.append(Wj_block(2, n, "0" + "1" * (n - 1), -dt, -np.pi / 2, xgate=False), list(derivatives[n + 1:2 * n]) + [blocks[0]] + [derivatives[n]] + [blocks[1]])
return qc
def run_sim(nx, na, R, initial_state, T, snapshot_dt=None, stop_check=None, progress_callback=None, print_callback=None):
"""
Runs the quantum simulation for electromagnetic scattering with fixed dt=0.1.
Captures frames only at user-defined snapshot times: [0, Δt, 2Δt, ..., ≤ T_eff],
always including t=0 and the final solver-aligned T (T_eff = floor(T/dt)*dt).
Returns:
frames (np.ndarray), snapshot_times (np.ndarray)
"""
def _log(msg):
if print_callback:
print_callback(msg)
else:
print(msg)
dt = 0.1
# Validate total time and compute solver-aligned end time
try:
T_val = float(T)
except Exception:
return np.array([]), np.array([])
if T_val <= 0:
return np.array([]), np.array([])
steps = int(np.floor(T_val / dt))
if steps <= 0:
return np.array([]), np.array([])
T_eff = steps * dt
# Determine snapshot Δt on solver grid
tol = 1e-12
if snapshot_dt is None:
snapshot_dt_val = dt
else:
try:
snapshot_dt_val = float(snapshot_dt)
except Exception:
snapshot_dt_val = dt
if snapshot_dt_val < dt - tol:
snapshot_dt_val = dt
k = max(1, int(round(snapshot_dt_val / dt)))
snapshot_dt_eff = k * dt
# Build requested snapshot times on solver grid
target_times = [0.0]
t = 0.0
while t + snapshot_dt_eff <= T_eff + tol:
t = round(t + snapshot_dt_eff, 12)
if t <= T_eff + tol:
target_times.append(min(t, T_eff))
if abs(target_times[-1] - T_eff) > tol:
target_times.append(T_eff)
# Setup circuit
nq = int(np.ceil(np.log2(nx)))
dp = 2 * R * np.pi / 2 ** na
p = np.arange(-R * np.pi, R * np.pi, step=dp)
fp = np.exp(-np.abs(p))
system, ancilla = QuantumRegister(2 * nq + 2), QuantumRegister(na)
qc = QuantumCircuit(system, ancilla)
qc.append(StatePreparation(initial_state), system)
qc.append(StatePreparation(fp / np.linalg.norm(fp)), ancilla)
expA1 = V1(nx, dt).to_gate()
expA2 = V2(nx, dt)
frames = []
# Capture initial frame at t=0
sv0 = np.real(Statevector(qc)).reshape(2 ** na, 2 ** (2 * nq + 2))
frames.append(sv0[2 ** (na - 1)])
next_idx = 1 # next target_times index to capture
_log(f"Starting simulation: T={T_eff:.2f}s, steps={steps}, snapshot_dt={snapshot_dt_eff:.2f}s")
for i in range(steps):
if stop_check and stop_check():
_log(f"Simulation interrupted at step {i}/{steps}")
break
# One solver step
qc.append(QFTGate(na), ancilla)
qc.x(ancilla[-1])
for j in range(na - 1):
qc.append(expA1.control().repeat(2 ** j), [ancilla[j]] + system[:])
qc.append(expA1.inverse().control(ctrl_state="0").repeat(2 ** (na - 1)), [ancilla[na - 1]] + system[:])
qc.append(expA2, system[:])
qc.x(ancilla[-1])
qc.append(QFTGate(na).inverse(), ancilla)
current_time = (i + 1) * dt
if next_idx < len(target_times) and abs(current_time - target_times[next_idx]) <= tol:
u = np.real(Statevector(qc)).reshape(2 ** na, 2 ** (2 * nq + 2))
frames.append(u[2 ** (na - 1)])
next_idx += 1
if progress_callback:
try:
progress = ((i + 1) / steps) * 100
progress_callback(progress)
except Exception:
pass
if progress_callback:
try:
progress_callback(100.0)
except Exception:
pass
_log("Simulation completed.")
# Ensure snapshot_times align with number of captured frames (covers early stop)
frames_arr = np.asarray(frames)
times_arr = np.asarray(target_times[: len(frames_arr)])
return frames_arr, times_arr
def create_impulse_preview_state(preview_n: int, pos01, sigma01: float = 0.02):
"""
Smooth delta-like preview on a unit square using a narrow Gaussian (sigma in [0,1]).
Preview-only helper, independent of simulation grid size (nx). Use this for the
Excitation preview; use the *_from_pos() variants for the actual simulation.
"""
try:
sx = float(sigma01) if sigma01 and sigma01 > 0 else 0.02
except Exception:
sx = 0.02
return create_gaussian_state_from_pos((int(preview_n), int(preview_n)), (float(pos01[0]), float(pos01[1])), (sx, sx))
##### Statevector Estimator Simulation Code Below #####
from .base_functions import *
def create_time_frames(total_time, snapshot_interval):
dt = 0.1
tol = 1e-9
try:
T_val = float(total_time)
except (ValueError, TypeError):
return []
if T_val <= 0:
return []
steps = int(np.floor(T_val / dt))
if steps <= 0:
return [0.0]
T_eff = steps * dt
try:
snapshot_dt_val = float(snapshot_interval)
except (ValueError, TypeError):
snapshot_dt_val = dt
if snapshot_dt_val < dt:
snapshot_dt_val = dt
k = max(1, int(round(snapshot_dt_val / dt)))
snapshot_dt_eff = k * dt
times = np.arange(0, T_eff + tol, snapshot_dt_eff)
if abs(times[-1] - T_eff) > tol:
times = np.append(times, T_eff)
times = np.round(times, 12)
unique_times = []
for t in times:
if not unique_times or abs(t - unique_times[-1]) > tol:
unique_times.append(float(t))
return unique_times
def run_sve(field, x, y, T, snapshot_time, nx, initial_state, impulse_pos, progress_callback=None, print_callback=None):
"""Statevector Estimator for time-series field values.
Supports both single-point and multi-point modes.
- Single-point (backward compatible): x, y are integers; returns list[float].
- Multi-point: x is a list/tuple of (ix, iy) integer pairs and y is None; returns dict[(ix,iy) -> list[float]].
"""
def _log(msg):
if print_callback:
print_callback(msg)
else:
print(msg)
na = 1
dt = 0.1
R = 4
nq = int(np.ceil(np.log2(nx)))
# Normalize monitor points input
if isinstance(x, (list, tuple)) and y is None:
points = [tuple(map(int, pt)) for pt in x]
multi = True
else:
points = [(int(x), int(y))]
multi = False
xref, yref = impulse_pos
offset = 0
grid_dims = (nx, nx)
initial_state = create_impulse_state(grid_dims, impulse_pos)
dp = 2 * R * np.pi / 2**na
p = np.arange(- R * np.pi, R * np.pi, step=dp)
fp = np.exp(-np.abs(p))
norm = np.linalg.norm(fp)
time_frames = create_time_frames(T, snapshot_time)
total_frames = len(time_frames)
_log(f"Starting QPU simulation: T={T}s, frames={total_frames}, points={len(points)}")
# Prepare outputs
if multi:
series_by_point = { (px, py): [] for (px, py) in points }
else:
series_single = []
for idx, time in enumerate(time_frames):
steps = int(math.ceil(time / dt))
# Reference Ez field at impulse location for sign
Eref = Eref_value(nx, nq, R, dt, na, steps, xref, yref, field_ref='Ez')
for (px, py) in points:
circ_magnitude = circ_for_magnitude(field, px, py, nx, na, R, dt, initial_state, steps)
magnitude = get_absolute_field_value(circ_magnitude, nq, na, offset, norm)
if field == 'Ez' and px == xref and py == yref:
Field_value = -magnitude if Eref < 0 else magnitude
else:
circsum, circdiff = circuits_for_sign(field, px, py, nx, na, dt, R, initial_state, steps, xref, yref, field_ref='Ez')
sign = get_relative_sign(circsum, circdiff, nq, na)
if (sign == 'same' and Eref > 0) or (sign == 'different' and Eref < 0):
Field_value = magnitude
else:
Field_value = -magnitude
if multi:
series_by_point[(px, py)].append(Field_value)
else:
series_single.append(Field_value)
# Calculate and report progress
pct = (idx + 1) / total_frames * 100
if progress_callback:
progress_callback(pct)
_log(f"SVE Progress: {int(pct)}% (frame {idx + 1}/{total_frames})")
_log("Statevector Estimator simulation completed.")
return series_by_point if multi else series_single |