Datasets
Models
Downloads: 1
Diff of
/Code/All Qiskit, PennyLane QML Nov 23/07a QSVM 2xData 16.6s kkawchak.ipynb
[000000]
..
[404218]
Switch to unified view
| a | b/Code/All Qiskit, PennyLane QML Nov 23/07a QSVM 2xData 16.6s kkawchak.ipynb | ||
|---|---|---|---|
| 1 | { |
||
| 2 | "cells": [ |
||
| 3 | { |
||
| 4 | "cell_type": "markdown", |
||
| 5 | "id": "opened-florist", |
||
| 6 | "metadata": {}, |
||
| 7 | "source": [ |
||
| 8 | "import time\n", |
||
| 9 | "seconds = time.time()\n", |
||
| 10 | "print(\"Time in seconds since beginning of run:\", seconds)\n", |
||
| 11 | "local_time = time.ctime(seconds)\n", |
||
| 12 | "print(local_time)\n", |
||
| 13 | "# Pegasos Quantum Support Vector Classifier\n", |
||
| 14 | "\n", |
||
| 15 | "There's another SVM based algorithm that benefits from the quantum kernel method. Here, we introduce an implementation of a another classification algorithm, which is an alternative version to the `QSVC` available in Qiskit Machine Learning and shown in the [\"Quantum Kernel Machine Learning\"](./03_quantum_kernel.ipynb) tutorial. This classification algorithm implements the Pegasos algorithm from the paper \"Pegasos: Primal Estimated sub-GrAdient SOlver for SVM\" by Shalev-Shwartz et al., see: https://home.ttic.edu/~nati/Publications/PegasosMPB.pdf.\n", |
||
| 16 | "\n", |
||
| 17 | "This algorithm is an alternative to the dual optimization from the `scikit-learn` package, benefits from the kernel trick, and yields a training complexity that is independent of the size of the training set. Thus, the `PegasosQSVC` is expected to train faster than QSVC for sufficiently large training sets.\n", |
||
| 18 | "\n", |
||
| 19 | "The algorithm can be used as direct replacement of `QSVC` with some hyper-parameterization." |
||
| 20 | ] |
||
| 21 | }, |
||
| 22 | { |
||
| 23 | "cell_type": "code", |
||
| 24 | "execution_count": 56, |
||
| 25 | "id": "759a01c7-b426-4ebe-8a4e-c9b9a1104603", |
||
| 26 | "metadata": {}, |
||
| 27 | "outputs": [ |
||
| 28 | { |
||
| 29 | "name": "stdout", |
||
| 30 | "output_type": "stream", |
||
| 31 | "text": [ |
||
| 32 | "Time in seconds since beginning of run: 1700640796.3861148\n", |
||
| 33 | "Wed Nov 22 08:13:16 2023\n" |
||
| 34 | ] |
||
| 35 | } |
||
| 36 | ], |
||
| 37 | "source": [ |
||
| 38 | "import time\n", |
||
| 39 | "seconds = time.time()\n", |
||
| 40 | "print(\"Time in seconds since beginning of run:\", seconds)\n", |
||
| 41 | "local_time = time.ctime(seconds)\n", |
||
| 42 | "print(local_time)" |
||
| 43 | ] |
||
| 44 | }, |
||
| 45 | { |
||
| 46 | "cell_type": "markdown", |
||
| 47 | "id": "thirty-painting", |
||
| 48 | "metadata": {}, |
||
| 49 | "source": [ |
||
| 50 | "Let's generate some data:" |
||
| 51 | ] |
||
| 52 | }, |
||
| 53 | { |
||
| 54 | "cell_type": "code", |
||
| 55 | "execution_count": 57, |
||
| 56 | "id": "impressed-laser", |
||
| 57 | "metadata": {}, |
||
| 58 | "outputs": [], |
||
| 59 | "source": [ |
||
| 60 | "from sklearn.datasets import make_blobs\n", |
||
| 61 | "\n", |
||
| 62 | "# example dataset\n", |
||
| 63 | "features, labels = make_blobs(n_samples=40, n_features=2, centers=2, random_state=3, shuffle=True)" |
||
| 64 | ] |
||
| 65 | }, |
||
| 66 | { |
||
| 67 | "cell_type": "markdown", |
||
| 68 | "id": "moderate-yugoslavia", |
||
| 69 | "metadata": {}, |
||
| 70 | "source": [ |
||
| 71 | "We pre-process the data to ensure compatibility with the rotation encoding and split it into the training and test datasets." |
||
| 72 | ] |
||
| 73 | }, |
||
| 74 | { |
||
| 75 | "cell_type": "code", |
||
| 76 | "execution_count": 58, |
||
| 77 | "id": "adolescent-composer", |
||
| 78 | "metadata": {}, |
||
| 79 | "outputs": [], |
||
| 80 | "source": [ |
||
| 81 | "import numpy as np\n", |
||
| 82 | "\n", |
||
| 83 | "from sklearn.model_selection import train_test_split\n", |
||
| 84 | "from sklearn.preprocessing import MinMaxScaler\n", |
||
| 85 | "\n", |
||
| 86 | "features = MinMaxScaler(feature_range=(0, np.pi)).fit_transform(features)\n", |
||
| 87 | "\n", |
||
| 88 | "train_features, test_features, train_labels, test_labels = train_test_split(\n", |
||
| 89 | " features, labels, train_size=30, shuffle=False\n", |
||
| 90 | ")" |
||
| 91 | ] |
||
| 92 | }, |
||
| 93 | { |
||
| 94 | "cell_type": "markdown", |
||
| 95 | "id": "central-poverty", |
||
| 96 | "metadata": {}, |
||
| 97 | "source": [ |
||
| 98 | "We have two features in the dataset, so we set a number of qubits to the number of features in the dataset.\n", |
||
| 99 | "\n", |
||
| 100 | "Then we set $\\tau$ to the number of steps performed during the training procedure. Please note that, there is no early stopping criterion in the algorithm. The algorithm iterates over all $\\tau$ steps.\n", |
||
| 101 | "\n", |
||
| 102 | "And the last one is the hyperparameter $C$. This is a positive regularization parameter. The strength of the regularization is inversely proportional to $C$. Smaller $C$ induce smaller weights which generally helps preventing overfitting. However, due to the nature of this algorithm, some of the computation steps become trivial for larger $C$. Thus, larger $C$ improve the performance of the algorithm drastically. If the data is linearly separable in feature space, $C$ should be chosen to be large. If the separation is not perfect, $C$ should be chosen smaller to prevent overfitting." |
||
| 103 | ] |
||
| 104 | }, |
||
| 105 | { |
||
| 106 | "cell_type": "code", |
||
| 107 | "execution_count": 59, |
||
| 108 | "id": "dying-dispatch", |
||
| 109 | "metadata": {}, |
||
| 110 | "outputs": [], |
||
| 111 | "source": [ |
||
| 112 | "# number of qubits is equal to the number of features\n", |
||
| 113 | "num_qubits = 2\n", |
||
| 114 | "\n", |
||
| 115 | "# number of steps performed during the training procedure\n", |
||
| 116 | "tau = 50\n", |
||
| 117 | "\n", |
||
| 118 | "# regularization parameter\n", |
||
| 119 | "C = 1000" |
||
| 120 | ] |
||
| 121 | }, |
||
| 122 | { |
||
| 123 | "cell_type": "markdown", |
||
| 124 | "id": "improving-wilderness", |
||
| 125 | "metadata": {}, |
||
| 126 | "source": [ |
||
| 127 | "The algorithm will run using:\n", |
||
| 128 | "\n", |
||
| 129 | "- The default fidelity instantiated in `FidelityQuantumKernel`\n", |
||
| 130 | "- A quantum kernel created from `ZFeatureMap`" |
||
| 131 | ] |
||
| 132 | }, |
||
| 133 | { |
||
| 134 | "cell_type": "code", |
||
| 135 | "execution_count": 60, |
||
| 136 | "id": "automated-allergy", |
||
| 137 | "metadata": {}, |
||
| 138 | "outputs": [], |
||
| 139 | "source": [ |
||
| 140 | "from qiskit import BasicAer\n", |
||
| 141 | "from qiskit.circuit.library import ZFeatureMap\n", |
||
| 142 | "from qiskit.utils import algorithm_globals\n", |
||
| 143 | "\n", |
||
| 144 | "from qiskit_machine_learning.kernels import FidelityQuantumKernel\n", |
||
| 145 | "\n", |
||
| 146 | "algorithm_globals.random_seed = 12345\n", |
||
| 147 | "\n", |
||
| 148 | "feature_map = ZFeatureMap(feature_dimension=num_qubits, reps=1)\n", |
||
| 149 | "\n", |
||
| 150 | "qkernel = FidelityQuantumKernel(feature_map=feature_map)" |
||
| 151 | ] |
||
| 152 | }, |
||
| 153 | { |
||
| 154 | "cell_type": "markdown", |
||
| 155 | "id": "attractive-stationery", |
||
| 156 | "metadata": {}, |
||
| 157 | "source": [ |
||
| 158 | "The implementation `PegasosQSVC` is compatible with the `scikit-learn` interfaces and has a pretty standard way of training a model. In the constructor we pass parameters of the algorithm, in this case there are a regularization hyper-parameter $C$ and a number of steps.\n", |
||
| 159 | "\n", |
||
| 160 | "Then we pass training features and labels to the `fit` method, which trains a models and returns a fitted classifier.\n", |
||
| 161 | "\n", |
||
| 162 | "Afterwards, we score our model using test features and labels." |
||
| 163 | ] |
||
| 164 | }, |
||
| 165 | { |
||
| 166 | "cell_type": "code", |
||
| 167 | "execution_count": 61, |
||
| 168 | "id": "representative-thumb", |
||
| 169 | "metadata": {}, |
||
| 170 | "outputs": [ |
||
| 171 | { |
||
| 172 | "name": "stdout", |
||
| 173 | "output_type": "stream", |
||
| 174 | "text": [ |
||
| 175 | "PegasosQSVC classification test score: 1.0\n" |
||
| 176 | ] |
||
| 177 | } |
||
| 178 | ], |
||
| 179 | "source": [ |
||
| 180 | "from qiskit_machine_learning.algorithms import PegasosQSVC\n", |
||
| 181 | "\n", |
||
| 182 | "pegasos_qsvc = PegasosQSVC(quantum_kernel=qkernel, C=C, num_steps=tau)\n", |
||
| 183 | "\n", |
||
| 184 | "# training\n", |
||
| 185 | "pegasos_qsvc.fit(train_features, train_labels)\n", |
||
| 186 | "\n", |
||
| 187 | "# testing\n", |
||
| 188 | "pegasos_score = pegasos_qsvc.score(test_features, test_labels)\n", |
||
| 189 | "print(f\"PegasosQSVC classification test score: {pegasos_score}\")" |
||
| 190 | ] |
||
| 191 | }, |
||
| 192 | { |
||
| 193 | "cell_type": "markdown", |
||
| 194 | "id": "sustainable-empire", |
||
| 195 | "metadata": {}, |
||
| 196 | "source": [ |
||
| 197 | "For visualization purposes we create a mesh grid of a predefined step that spans our minimum and maximum values we applied in MinMaxScaler. We also add some margin to the grid for better representation of the training and test samples." |
||
| 198 | ] |
||
| 199 | }, |
||
| 200 | { |
||
| 201 | "cell_type": "code", |
||
| 202 | "execution_count": 62, |
||
| 203 | "id": "judicial-pottery", |
||
| 204 | "metadata": {}, |
||
| 205 | "outputs": [], |
||
| 206 | "source": [ |
||
| 207 | "grid_step = 0.2\n", |
||
| 208 | "margin = 0.2\n", |
||
| 209 | "grid_x, grid_y = np.meshgrid(\n", |
||
| 210 | " np.arange(-margin, np.pi + margin, grid_step), np.arange(-margin, np.pi + margin, grid_step)\n", |
||
| 211 | ")" |
||
| 212 | ] |
||
| 213 | }, |
||
| 214 | { |
||
| 215 | "cell_type": "markdown", |
||
| 216 | "id": "marine-constitution", |
||
| 217 | "metadata": {}, |
||
| 218 | "source": [ |
||
| 219 | "We convert the grid to the shape compatible with the model, the shape should be `(n_samples, n_features)`.\n", |
||
| 220 | "Then for each grid point we predict a label. In our case predicted labels will be used for coloring the grid." |
||
| 221 | ] |
||
| 222 | }, |
||
| 223 | { |
||
| 224 | "cell_type": "code", |
||
| 225 | "execution_count": 63, |
||
| 226 | "id": "competitive-outdoors", |
||
| 227 | "metadata": {}, |
||
| 228 | "outputs": [], |
||
| 229 | "source": [ |
||
| 230 | "meshgrid_features = np.column_stack((grid_x.ravel(), grid_y.ravel()))\n", |
||
| 231 | "meshgrid_colors = pegasos_qsvc.predict(meshgrid_features)" |
||
| 232 | ] |
||
| 233 | }, |
||
| 234 | { |
||
| 235 | "cell_type": "markdown", |
||
| 236 | "id": "former-constraint", |
||
| 237 | "metadata": {}, |
||
| 238 | "source": [ |
||
| 239 | "Finally, we plot our grid according to the labels/colors we obtained from the model. We also plot training and test samples." |
||
| 240 | ] |
||
| 241 | }, |
||
| 242 | { |
||
| 243 | "cell_type": "code", |
||
| 244 | "execution_count": 64, |
||
| 245 | "id": "monetary-knife", |
||
| 246 | "metadata": { |
||
| 247 | "tags": [ |
||
| 248 | "nbsphinx-thumbnail" |
||
| 249 | ] |
||
| 250 | }, |
||
| 251 | "outputs": [ |
||
| 252 | { |
||
| 253 | "data": { |
||
| 254 | "image/png": "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\n", |
||
| 255 | "text/plain": [ |
||
| 256 | "<Figure size 500x500 with 1 Axes>" |
||
| 257 | ] |
||
| 258 | }, |
||
| 259 | "metadata": {}, |
||
| 260 | "output_type": "display_data" |
||
| 261 | } |
||
| 262 | ], |
||
| 263 | "source": [ |
||
| 264 | "import matplotlib.pyplot as plt\n", |
||
| 265 | "\n", |
||
| 266 | "plt.figure(figsize=(5, 5))\n", |
||
| 267 | "meshgrid_colors = meshgrid_colors.reshape(grid_x.shape)\n", |
||
| 268 | "plt.pcolormesh(grid_x, grid_y, meshgrid_colors, cmap=\"RdBu\", shading=\"auto\")\n", |
||
| 269 | "\n", |
||
| 270 | "plt.scatter(\n", |
||
| 271 | " train_features[:, 0][train_labels == 0],\n", |
||
| 272 | " train_features[:, 1][train_labels == 0],\n", |
||
| 273 | " marker=\"s\",\n", |
||
| 274 | " facecolors=\"w\",\n", |
||
| 275 | " edgecolors=\"r\",\n", |
||
| 276 | " label=\"A train\",\n", |
||
| 277 | ")\n", |
||
| 278 | "plt.scatter(\n", |
||
| 279 | " train_features[:, 0][train_labels == 1],\n", |
||
| 280 | " train_features[:, 1][train_labels == 1],\n", |
||
| 281 | " marker=\"o\",\n", |
||
| 282 | " facecolors=\"w\",\n", |
||
| 283 | " edgecolors=\"b\",\n", |
||
| 284 | " label=\"B train\",\n", |
||
| 285 | ")\n", |
||
| 286 | "\n", |
||
| 287 | "plt.scatter(\n", |
||
| 288 | " test_features[:, 0][test_labels == 0],\n", |
||
| 289 | " test_features[:, 1][test_labels == 0],\n", |
||
| 290 | " marker=\"s\",\n", |
||
| 291 | " facecolors=\"r\",\n", |
||
| 292 | " edgecolors=\"r\",\n", |
||
| 293 | " label=\"A test\",\n", |
||
| 294 | ")\n", |
||
| 295 | "plt.scatter(\n", |
||
| 296 | " test_features[:, 0][test_labels == 1],\n", |
||
| 297 | " test_features[:, 1][test_labels == 1],\n", |
||
| 298 | " marker=\"o\",\n", |
||
| 299 | " facecolors=\"b\",\n", |
||
| 300 | " edgecolors=\"b\",\n", |
||
| 301 | " label=\"B test\",\n", |
||
| 302 | ")\n", |
||
| 303 | "\n", |
||
| 304 | "plt.legend(bbox_to_anchor=(1.05, 1), loc=\"upper left\", borderaxespad=0.0)\n", |
||
| 305 | "plt.title(\"Pegasos Classification\")\n", |
||
| 306 | "plt.show()" |
||
| 307 | ] |
||
| 308 | }, |
||
| 309 | { |
||
| 310 | "cell_type": "code", |
||
| 311 | "execution_count": 65, |
||
| 312 | "id": "fddca436-0531-4f77-9c54-71a0ed4dc462", |
||
| 313 | "metadata": {}, |
||
| 314 | "outputs": [ |
||
| 315 | { |
||
| 316 | "name": "stdout", |
||
| 317 | "output_type": "stream", |
||
| 318 | "text": [ |
||
| 319 | "Time in seconds since end of run: 1700640813.010587\n", |
||
| 320 | "Wed Nov 22 08:13:33 2023\n" |
||
| 321 | ] |
||
| 322 | } |
||
| 323 | ], |
||
| 324 | "source": [ |
||
| 325 | "seconds = time.time()\n", |
||
| 326 | "print(\"Time in seconds since end of run:\", seconds)\n", |
||
| 327 | "local_time = time.ctime(seconds)\n", |
||
| 328 | "print(local_time)" |
||
| 329 | ] |
||
| 330 | }, |
||
| 331 | { |
||
| 332 | "cell_type": "code", |
||
| 333 | "execution_count": 66, |
||
| 334 | "id": "imperial-promise", |
||
| 335 | "metadata": { |
||
| 336 | "tags": [] |
||
| 337 | }, |
||
| 338 | "outputs": [ |
||
| 339 | { |
||
| 340 | "data": { |
||
| 341 | "text/html": [ |
||
| 342 | "<h3>Version Information</h3><table><tr><th>Software</th><th>Version</th></tr><tr><td><code>qiskit</code></td><td>0.44.1</td></tr><tr><td><code>qiskit-terra</code></td><td>0.25.1</td></tr><tr><td><code>qiskit_machine_learning</code></td><td>0.6.1</td></tr><tr><th colspan='2'>System information</th></tr><tr><td>Python version</td><td>3.10.8</td></tr><tr><td>Python compiler</td><td>GCC 10.4.0</td></tr><tr><td>Python build</td><td>main, Nov 22 2022 08:26:04</td></tr><tr><td>OS</td><td>Linux</td></tr><tr><td>CPUs</td><td>8</td></tr><tr><td>Memory (Gb)</td><td>31.142810821533203</td></tr><tr><td colspan='2'>Wed Nov 22 08:13:33 2023 UTC</td></tr></table>" |
||
| 343 | ], |
||
| 344 | "text/plain": [ |
||
| 345 | "<IPython.core.display.HTML object>" |
||
| 346 | ] |
||
| 347 | }, |
||
| 348 | "metadata": {}, |
||
| 349 | "output_type": "display_data" |
||
| 350 | }, |
||
| 351 | { |
||
| 352 | "data": { |
||
| 353 | "text/html": [ |
||
| 354 | "<div style='width: 100%; background-color:#d5d9e0;padding-left: 10px; padding-bottom: 10px; padding-right: 10px; padding-top: 5px'><h3>This code is a part of Qiskit</h3><p>© Copyright IBM 2017, 2023.</p><p>This code is licensed under the Apache License, Version 2.0. You may<br>obtain a copy of this license in the LICENSE.txt file in the root directory<br> of this source tree or at http://www.apache.org/licenses/LICENSE-2.0.<p>Any modifications or derivative works of this code must retain this<br>copyright notice, and modified files need to carry a notice indicating<br>that they have been altered from the originals.</p></div>" |
||
| 355 | ], |
||
| 356 | "text/plain": [ |
||
| 357 | "<IPython.core.display.HTML object>" |
||
| 358 | ] |
||
| 359 | }, |
||
| 360 | "metadata": {}, |
||
| 361 | "output_type": "display_data" |
||
| 362 | } |
||
| 363 | ], |
||
| 364 | "source": [ |
||
| 365 | "import qiskit.tools.jupyter\n", |
||
| 366 | "\n", |
||
| 367 | "%qiskit_version_table\n", |
||
| 368 | "%qiskit_copyright" |
||
| 369 | ] |
||
| 370 | } |
||
| 371 | ], |
||
| 372 | "metadata": { |
||
| 373 | "celltoolbar": "Tags", |
||
| 374 | "kernelspec": { |
||
| 375 | "display_name": "Python 3 (ipykernel)", |
||
| 376 | "language": "python", |
||
| 377 | "name": "python3" |
||
| 378 | }, |
||
| 379 | "language_info": { |
||
| 380 | "codemirror_mode": { |
||
| 381 | "name": "ipython", |
||
| 382 | "version": 3 |
||
| 383 | }, |
||
| 384 | "file_extension": ".py", |
||
| 385 | "mimetype": "text/x-python", |
||
| 386 | "name": "python", |
||
| 387 | "nbconvert_exporter": "python", |
||
| 388 | "pygments_lexer": "ipython3", |
||
| 389 | "version": "3.10.8" |
||
| 390 | }, |
||
| 391 | "widgets": { |
||
| 392 | "application/vnd.jupyter.widget-state+json": { |
||
| 393 | "state": { |
||
| 394 | "1b3456f682144beeaec61de127255f23": { |
||
| 395 | "model_module": "@jupyter-widgets/controls", |
||
| 396 | "model_module_version": "2.0.0", |
||
| 397 | "model_name": "HTMLStyleModel", |
||
| 398 | "state": { |
||
| 399 | "description_width": "", |
||
| 400 | "font_size": null, |
||
| 401 | "text_color": null |
||
| 402 | } |
||
| 403 | }, |
||
| 404 | "22ac8e820eb543138cfebf0dafe77136": { |
||
| 405 | "model_module": "@jupyter-widgets/controls", |
||
| 406 | "model_module_version": "2.0.0", |
||
| 407 | "model_name": "HTMLModel", |
||
| 408 | "state": { |
||
| 409 | "layout": "IPY_MODEL_b3bcc1a52ab64fa283a8a642b4777df2", |
||
| 410 | "style": "IPY_MODEL_f00034b9b85f49d0a754f396a37354ed", |
||
| 411 | "value": "<h5>Status</h5>" |
||
| 412 | } |
||
| 413 | }, |
||
| 414 | "41fd22798e464ef7996d6a6ad6e13d52": { |
||
| 415 | "model_module": "@jupyter-widgets/controls", |
||
| 416 | "model_module_version": "2.0.0", |
||
| 417 | "model_name": "HTMLStyleModel", |
||
| 418 | "state": { |
||
| 419 | "description_width": "", |
||
| 420 | "font_size": null, |
||
| 421 | "text_color": null |
||
| 422 | } |
||
| 423 | }, |
||
| 424 | "4308592f5e0a411c80bd75ca8247cdc4": { |
||
| 425 | "model_module": "@jupyter-widgets/base", |
||
| 426 | "model_module_version": "2.0.0", |
||
| 427 | "model_name": "LayoutModel", |
||
| 428 | "state": { |
||
| 429 | "width": "70px" |
||
| 430 | } |
||
| 431 | }, |
||
| 432 | "446272bfa6a048389eb5efa3a6ef0af8": { |
||
| 433 | "model_module": "@jupyter-widgets/controls", |
||
| 434 | "model_module_version": "2.0.0", |
||
| 435 | "model_name": "HTMLModel", |
||
| 436 | "state": { |
||
| 437 | "layout": "IPY_MODEL_6f0b6cb82e7e470c943a33461af8bc16", |
||
| 438 | "style": "IPY_MODEL_41fd22798e464ef7996d6a6ad6e13d52", |
||
| 439 | "value": "<h5>Backend</h5>" |
||
| 440 | } |
||
| 441 | }, |
||
| 442 | "4b9c0b1d4326437e86e9e30db1d5fea6": { |
||
| 443 | "model_module": "@jupyter-widgets/controls", |
||
| 444 | "model_module_version": "2.0.0", |
||
| 445 | "model_name": "HTMLModel", |
||
| 446 | "state": { |
||
| 447 | "layout": "IPY_MODEL_4308592f5e0a411c80bd75ca8247cdc4", |
||
| 448 | "style": "IPY_MODEL_6039025fb08c4be5a9191a654dec032d", |
||
| 449 | "value": "<h5>Queue</h5>" |
||
| 450 | } |
||
| 451 | }, |
||
| 452 | "4d2cce1675aa4f2b9c1cf23412aac9a7": { |
||
| 453 | "model_module": "@jupyter-widgets/base", |
||
| 454 | "model_module_version": "2.0.0", |
||
| 455 | "model_name": "LayoutModel", |
||
| 456 | "state": { |
||
| 457 | "width": "190px" |
||
| 458 | } |
||
| 459 | }, |
||
| 460 | "54c195dc23cb459890244746aba5053a": { |
||
| 461 | "model_module": "@jupyter-widgets/controls", |
||
| 462 | "model_module_version": "2.0.0", |
||
| 463 | "model_name": "HTMLStyleModel", |
||
| 464 | "state": { |
||
| 465 | "description_width": "", |
||
| 466 | "font_size": null, |
||
| 467 | "text_color": null |
||
| 468 | } |
||
| 469 | }, |
||
| 470 | "6039025fb08c4be5a9191a654dec032d": { |
||
| 471 | "model_module": "@jupyter-widgets/controls", |
||
| 472 | "model_module_version": "2.0.0", |
||
| 473 | "model_name": "HTMLStyleModel", |
||
| 474 | "state": { |
||
| 475 | "description_width": "", |
||
| 476 | "font_size": null, |
||
| 477 | "text_color": null |
||
| 478 | } |
||
| 479 | }, |
||
| 480 | "6eb42a5a8aa7449b9977c6288066562f": { |
||
| 481 | "model_module": "@jupyter-widgets/controls", |
||
| 482 | "model_module_version": "2.0.0", |
||
| 483 | "model_name": "HTMLStyleModel", |
||
| 484 | "state": { |
||
| 485 | "description_width": "", |
||
| 486 | "font_size": null, |
||
| 487 | "text_color": null |
||
| 488 | } |
||
| 489 | }, |
||
| 490 | "6f0b6cb82e7e470c943a33461af8bc16": { |
||
| 491 | "model_module": "@jupyter-widgets/base", |
||
| 492 | "model_module_version": "2.0.0", |
||
| 493 | "model_name": "LayoutModel", |
||
| 494 | "state": { |
||
| 495 | "width": "145px" |
||
| 496 | } |
||
| 497 | }, |
||
| 498 | "7aa4f6cd229542e7a6497e0248615f55": { |
||
| 499 | "model_module": "@jupyter-widgets/controls", |
||
| 500 | "model_module_version": "2.0.0", |
||
| 501 | "model_name": "HTMLModel", |
||
| 502 | "state": { |
||
| 503 | "layout": "IPY_MODEL_f1653f08dcdf410e97c8b441d1ddbe9a", |
||
| 504 | "style": "IPY_MODEL_54c195dc23cb459890244746aba5053a", |
||
| 505 | "value": "<h5>Message</h5>" |
||
| 506 | } |
||
| 507 | }, |
||
| 508 | "a3d541cb70c34d85b34fdc709f7e7af4": { |
||
| 509 | "model_module": "@jupyter-widgets/controls", |
||
| 510 | "model_module_version": "2.0.0", |
||
| 511 | "model_name": "HBoxModel", |
||
| 512 | "state": { |
||
| 513 | "children": [ |
||
| 514 | "IPY_MODEL_aecc303bf5fe454a81dbb1a3ba7ac3d5", |
||
| 515 | "IPY_MODEL_446272bfa6a048389eb5efa3a6ef0af8", |
||
| 516 | "IPY_MODEL_22ac8e820eb543138cfebf0dafe77136", |
||
| 517 | "IPY_MODEL_4b9c0b1d4326437e86e9e30db1d5fea6", |
||
| 518 | "IPY_MODEL_7aa4f6cd229542e7a6497e0248615f55" |
||
| 519 | ], |
||
| 520 | "layout": "IPY_MODEL_f02c3ce2ffd94b5abe9fa5128064baf7" |
||
| 521 | } |
||
| 522 | }, |
||
| 523 | "a59a211db01941b6b5bb22fd4d0e8bf0": { |
||
| 524 | "model_module": "@jupyter-widgets/base", |
||
| 525 | "model_module_version": "2.0.0", |
||
| 526 | "model_name": "LayoutModel", |
||
| 527 | "state": { |
||
| 528 | "grid_template_areas": "\n \". . . . right \"\n ", |
||
| 529 | "grid_template_columns": "20% 20% 20% 20% 20%", |
||
| 530 | "width": "100%" |
||
| 531 | } |
||
| 532 | }, |
||
| 533 | "a71e1bfeecfa4a298bdf3fe6cffad1b7": { |
||
| 534 | "model_module": "@jupyter-widgets/base", |
||
| 535 | "model_module_version": "2.0.0", |
||
| 536 | "model_name": "LayoutModel", |
||
| 537 | "state": { |
||
| 538 | "grid_area": "right", |
||
| 539 | "padding": "0px 0px 0px 0px", |
||
| 540 | "width": "70px" |
||
| 541 | } |
||
| 542 | }, |
||
| 543 | "aecc303bf5fe454a81dbb1a3ba7ac3d5": { |
||
| 544 | "model_module": "@jupyter-widgets/controls", |
||
| 545 | "model_module_version": "2.0.0", |
||
| 546 | "model_name": "HTMLModel", |
||
| 547 | "state": { |
||
| 548 | "layout": "IPY_MODEL_4d2cce1675aa4f2b9c1cf23412aac9a7", |
||
| 549 | "style": "IPY_MODEL_1b3456f682144beeaec61de127255f23", |
||
| 550 | "value": "<h5>Job ID</h5>" |
||
| 551 | } |
||
| 552 | }, |
||
| 553 | "b3bcc1a52ab64fa283a8a642b4777df2": { |
||
| 554 | "model_module": "@jupyter-widgets/base", |
||
| 555 | "model_module_version": "2.0.0", |
||
| 556 | "model_name": "LayoutModel", |
||
| 557 | "state": { |
||
| 558 | "width": "95px" |
||
| 559 | } |
||
| 560 | }, |
||
| 561 | "ba447f375c82410cbf25489d7a0ca515": { |
||
| 562 | "model_module": "@jupyter-widgets/controls", |
||
| 563 | "model_module_version": "2.0.0", |
||
| 564 | "model_name": "ButtonStyleModel", |
||
| 565 | "state": { |
||
| 566 | "font_family": null, |
||
| 567 | "font_size": null, |
||
| 568 | "font_style": null, |
||
| 569 | "font_variant": null, |
||
| 570 | "font_weight": null, |
||
| 571 | "text_color": null, |
||
| 572 | "text_decoration": null |
||
| 573 | } |
||
| 574 | }, |
||
| 575 | "c18f1f32fa9a46a992f26adafe4aa32b": { |
||
| 576 | "model_module": "@jupyter-widgets/controls", |
||
| 577 | "model_module_version": "2.0.0", |
||
| 578 | "model_name": "GridBoxModel", |
||
| 579 | "state": { |
||
| 580 | "children": [ |
||
| 581 | "IPY_MODEL_e94a67deff694b75aa9135bba1d838d0" |
||
| 582 | ], |
||
| 583 | "layout": "IPY_MODEL_a59a211db01941b6b5bb22fd4d0e8bf0" |
||
| 584 | } |
||
| 585 | }, |
||
| 586 | "cfce2ae34c3944ec9f8cbbefdc50015a": { |
||
| 587 | "model_module": "@jupyter-widgets/controls", |
||
| 588 | "model_module_version": "2.0.0", |
||
| 589 | "model_name": "HTMLModel", |
||
| 590 | "state": { |
||
| 591 | "layout": "IPY_MODEL_eb4c6152dc4c4ac586f6199d76cbf875", |
||
| 592 | "style": "IPY_MODEL_6eb42a5a8aa7449b9977c6288066562f", |
||
| 593 | "value": "<p style='font-family: IBM Plex Sans, Arial, Helvetica, sans-serif; font-size: 20px; font-weight: medium;'>Circuit Properties</p>" |
||
| 594 | } |
||
| 595 | }, |
||
| 596 | "e94a67deff694b75aa9135bba1d838d0": { |
||
| 597 | "model_module": "@jupyter-widgets/controls", |
||
| 598 | "model_module_version": "2.0.0", |
||
| 599 | "model_name": "ButtonModel", |
||
| 600 | "state": { |
||
| 601 | "button_style": "primary", |
||
| 602 | "description": "Clear", |
||
| 603 | "layout": "IPY_MODEL_a71e1bfeecfa4a298bdf3fe6cffad1b7", |
||
| 604 | "style": "IPY_MODEL_ba447f375c82410cbf25489d7a0ca515", |
||
| 605 | "tooltip": null |
||
| 606 | } |
||
| 607 | }, |
||
| 608 | "eb4c6152dc4c4ac586f6199d76cbf875": { |
||
| 609 | "model_module": "@jupyter-widgets/base", |
||
| 610 | "model_module_version": "2.0.0", |
||
| 611 | "model_name": "LayoutModel", |
||
| 612 | "state": { |
||
| 613 | "margin": "0px 0px 10px 0px" |
||
| 614 | } |
||
| 615 | }, |
||
| 616 | "f00034b9b85f49d0a754f396a37354ed": { |
||
| 617 | "model_module": "@jupyter-widgets/controls", |
||
| 618 | "model_module_version": "2.0.0", |
||
| 619 | "model_name": "HTMLStyleModel", |
||
| 620 | "state": { |
||
| 621 | "description_width": "", |
||
| 622 | "font_size": null, |
||
| 623 | "text_color": null |
||
| 624 | } |
||
| 625 | }, |
||
| 626 | "f02c3ce2ffd94b5abe9fa5128064baf7": { |
||
| 627 | "model_module": "@jupyter-widgets/base", |
||
| 628 | "model_module_version": "2.0.0", |
||
| 629 | "model_name": "LayoutModel", |
||
| 630 | "state": { |
||
| 631 | "margin": "0px 0px 0px 37px", |
||
| 632 | "width": "600px" |
||
| 633 | } |
||
| 634 | }, |
||
| 635 | "f1653f08dcdf410e97c8b441d1ddbe9a": { |
||
| 636 | "model_module": "@jupyter-widgets/base", |
||
| 637 | "model_module_version": "2.0.0", |
||
| 638 | "model_name": "LayoutModel", |
||
| 639 | "state": {} |
||
| 640 | } |
||
| 641 | }, |
||
| 642 | "version_major": 2, |
||
| 643 | "version_minor": 0 |
||
| 644 | } |
||
| 645 | } |
||
| 646 | }, |
||
| 647 | "nbformat": 4, |
||
| 648 | "nbformat_minor": 5 |
||
| 649 | } |