{ "cells": [ { "cell_type": "markdown", "id": "3fa07dab", "metadata": { "customInput": null, "originalKey": "06e172a0-2da3-4c90-93c2-be01bf4f6d45", "papermill": { "duration": 0.004158, "end_time": "2023-12-09T19:59:01.073069", "exception": false, "start_time": "2023-12-09T19:59:01.068911", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "This tutorial illustrates use of a Global Stopping Strategy (GSS) in combination with the Service API. For background on the Service API, see the Service API Tutorial: https://ax.dev/tutorials/gpei_hartmann_service.html GSS is also supported in the Scheduler API, where it can be provided as part of `SchedulerOptions`. For more on `Scheduler`, see the Scheduler tutorial: https://ax.dev/tutorials/scheduler.html\n", "\n", "Global Stopping stops an optimization loop when some data-based criteria are met which suggest that future trials will not be very helpful. For example, we might stop when there has been very little improvement in the last five trials. This is as opposed to trial-level early stopping, which monitors the results of expensive evaluations and terminates those that are unlikely to produce promising results, freeing resources to explore more promising configurations. For more on trial-level early stopping, see the tutorial: https://ax.dev/tutorials/early_stopping/early_stopping.html" ] }, { "cell_type": "code", "execution_count": 1, "id": "8976b5b7", "metadata": { "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:01.081649Z", "iopub.status.busy": "2023-12-09T19:59:01.081141Z", "iopub.status.idle": "2023-12-09T19:59:04.270602Z", "shell.execute_reply": "2023-12-09T19:59:04.269630Z" }, "executionStartTime": 1683829335587, "executionStopTime": 1683829339370, "originalKey": "00a04d2c-d990-41c1-9eef-bbb05fba000d", "papermill": { "duration": 3.209035, "end_time": "2023-12-09T19:59:04.285668", "exception": false, "start_time": "2023-12-09T19:59:01.076633", "status": "completed" }, "requestMsgId": "1c560539-1c7d-4c7a-ae55-e87c3b601859", "tags": [] }, "outputs": [ { "data": { "text/html": [ "" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.utils.notebook.plotting: Injecting Plotly library into cell. Do not overwrite or delete cell.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.utils.notebook.plotting: Please see\n", " (https://ax.dev/tutorials/visualizations.html#Fix-for-plots-that-are-not-rendering)\n", " if visualizations are not rendering.\n" ] }, { "data": { "text/html": [ " \n", " " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import numpy as np\n", "\n", "from ax.service.ax_client import AxClient, ObjectiveProperties\n", "from ax.utils.measurement.synthetic_functions import Branin, branin\n", "from ax.utils.notebook.plotting import init_notebook_plotting, render\n", "\n", "init_notebook_plotting()" ] }, { "cell_type": "markdown", "id": "d7d28958", "metadata": { "customInput": null, "originalKey": "8688d729-b402-4a4c-b796-94fdcf5e022c", "papermill": { "duration": 0.034755, "end_time": "2023-12-09T19:59:04.354887", "exception": false, "start_time": "2023-12-09T19:59:04.320132", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "# 1. What happens without global stopping? Optimization can run for too long.\n", "This example uses the Branin test problem. We run 25 trials, which turns out to be far more than needed, because we get close to the optimum quite quickly." ] }, { "cell_type": "code", "execution_count": 2, "id": "051f60f3", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:04.426721Z", "iopub.status.busy": "2023-12-09T19:59:04.426109Z", "iopub.status.idle": "2023-12-09T19:59:04.430666Z", "shell.execute_reply": "2023-12-09T19:59:04.429989Z" }, "executionStartTime": 1683829339516, "executionStopTime": 1683829339531, "originalKey": "320a952b-9e78-43e1-a55b-76a355e90f83", "papermill": { "duration": 0.042151, "end_time": "2023-12-09T19:59:04.432019", "exception": false, "start_time": "2023-12-09T19:59:04.389868", "status": "completed" }, "requestMsgId": "14e3a517-c7d0-4300-92d9-57ceb5afca34", "showInput": true, "tags": [] }, "outputs": [], "source": [ "def evaluate(parameters):\n", " x = np.array([parameters.get(f\"x{i+1}\") for i in range(2)])\n", " return {\"branin\": (branin(x), 0.0)}" ] }, { "cell_type": "code", "execution_count": 3, "id": "8d3afc1e", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:04.504042Z", "iopub.status.busy": "2023-12-09T19:59:04.503316Z", "iopub.status.idle": "2023-12-09T19:59:04.507520Z", "shell.execute_reply": "2023-12-09T19:59:04.506870Z" }, "executionStartTime": 1683829339659, "executionStopTime": 1683829339668, "originalKey": "5740fbc2-97d6-465b-b01c-61e6c34c0220", "papermill": { "duration": 0.041984, "end_time": "2023-12-09T19:59:04.508914", "exception": false, "start_time": "2023-12-09T19:59:04.466930", "status": "completed" }, "requestMsgId": "ff819cc9-ff17-4763-a857-83662b01e955", "showInput": true, "tags": [] }, "outputs": [], "source": [ "params = [\n", " {\n", " \"name\": f\"x{i + 1}\",\n", " \"type\": \"range\",\n", " \"bounds\": [*Branin._domain[i]],\n", " \"value_type\": \"float\",\n", " \"log_scale\": False,\n", " }\n", "\n", " for i in range(2)\n", "]" ] }, { "cell_type": "code", "execution_count": 4, "id": "f6d60455", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:04.579770Z", "iopub.status.busy": "2023-12-09T19:59:04.579289Z", "iopub.status.idle": "2023-12-09T19:59:04.588992Z", "shell.execute_reply": "2023-12-09T19:59:04.588380Z" }, "executionStartTime": 1683829339782, "executionStopTime": 1683829339834, "originalKey": "65667172-14df-437b-bdd0-5a59580e4054", "papermill": { "duration": 0.04706, "end_time": "2023-12-09T19:59:04.590449", "exception": false, "start_time": "2023-12-09T19:59:04.543389", "status": "completed" }, "requestMsgId": "e0bc2847-17a5-43d7-bf49-ed97c90f1d50", "showInput": true, "tags": [] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[WARNING 12-09 19:59:04] ax.service.ax_client: Random seed set to 0. Note that this setting only affects the Sobol quasi-random generator and BoTorch-powered Bayesian optimization models. For the latter models, setting random seed to the same number for two optimizations will make the generated trials similar, but not exactly the same, and over time the trials will diverge more.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.service.utils.instantiation: Created search space: SearchSpace(parameters=[RangeParameter(name='x1', parameter_type=FLOAT, range=[-5.0, 10.0]), RangeParameter(name='x2', parameter_type=FLOAT, range=[0.0, 15.0])], parameter_constraints=[]).\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.core.experiment: The is_test flag has been set to True. This flag is meant purely for development and integration testing purposes. If you are running a live experiment, please set this flag to False\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.modelbridge.dispatch_utils: Using Models.BOTORCH_MODULAR since there are more ordered parameters than there are categories for the unordered categorical parameters.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.modelbridge.dispatch_utils: Calculating the number of remaining initialization trials based on num_initialization_trials=None max_initialization_trials=None num_tunable_parameters=2 num_trials=None use_batch_trials=False\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.modelbridge.dispatch_utils: calculated num_initialization_trials=5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.modelbridge.dispatch_utils: num_completed_initialization_trials=0 num_remaining_initialization_trials=5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.modelbridge.dispatch_utils: `verbose`, `disable_progbar`, and `jit_compile` are not yet supported when using `choose_generation_strategy` with ModularBoTorchModel, dropping these arguments.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:04] ax.modelbridge.dispatch_utils: Using Bayesian Optimization generation strategy: GenerationStrategy(name='Sobol+BoTorch', steps=[Sobol for 5 trials, BoTorch for subsequent trials]). Iterations after 5 will take longer to generate due to model-fitting.\n" ] } ], "source": [ "ax_client = AxClient(random_seed=0, verbose_logging=False)\n", "\n", "ax_client.create_experiment(\n", " name=\"branin_test_experiment\",\n", " parameters=params,\n", " objectives={\"branin\": ObjectiveProperties(minimize=True)},\n", " is_test=True,\n", ")" ] }, { "cell_type": "code", "execution_count": 5, "id": "1edb5630", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:04.662335Z", "iopub.status.busy": "2023-12-09T19:59:04.661872Z", "iopub.status.idle": "2023-12-09T19:59:15.542906Z", "shell.execute_reply": "2023-12-09T19:59:15.542208Z" }, "executionStartTime": 1683829339928, "executionStopTime": 1683829356006, "originalKey": "1f208de3-5189-4847-a779-940795977845", "papermill": { "duration": 10.918806, "end_time": "2023-12-09T19:59:15.544437", "exception": false, "start_time": "2023-12-09T19:59:04.625631", "status": "completed" }, "requestMsgId": "95f327f2-327f-4284-93ae-3053c9b6ec45", "showInput": true, "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "CPU times: user 21.5 s, sys: 42.7 ms, total: 21.5 s\n", "Wall time: 10.9 s\n" ] } ], "source": [ "%%time\n", "for i in range(25):\n", " parameters, trial_index = ax_client.get_next_trial()\n", " # Local evaluation here can be replaced with deployment to external system.\n", " ax_client.complete_trial(\n", " trial_index=trial_index, raw_data=evaluate(parameters)\n", " )" ] }, { "cell_type": "code", "execution_count": 6, "id": "d8bfa765", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:15.617970Z", "iopub.status.busy": "2023-12-09T19:59:15.617323Z", "iopub.status.idle": "2023-12-09T19:59:15.780818Z", "shell.execute_reply": "2023-12-09T19:59:15.780155Z" }, "executionStartTime": 1683829356136, "executionStopTime": 1683829356616, "originalKey": "a369aafa-8ee4-4c02-bea6-673271da81ab", "papermill": { "duration": 0.202333, "end_time": "2023-12-09T19:59:15.782529", "exception": false, "start_time": "2023-12-09T19:59:15.580196", "status": "completed" }, "requestMsgId": "b601e1e9-fd2d-4faf-a369-04e5c4a9f8cb", "showInput": true, "tags": [] }, "outputs": [ { "data": { "application/vnd.plotly.v1+json": { "config": { "linkText": "Export to plot.ly", "plotlyServerURL": "https://plot.ly", "showLink": false }, "data": [ { "hoverinfo": "none", "legendgroup": "", "line": { "width": 0 }, "mode": "lines", "showlegend": false, "type": "scatter", "x": [ 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 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811, 0.5928152446626811, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.40071785969624685, 0.40071785969624685, 0.40071785969624685 ] }, { "fill": "tonexty", "fillcolor": "rgba(128,177,211,0.3)", "legendgroup": "objective value", "line": { "color": "rgba(128,177,211,1)" }, "mode": "lines", "name": "objective value", "text": [ "
Parameterization:
x1: 2.126607894897461
x2: 8.887859880924225", "
Parameterization:
x1: 3.681450095027685
x2: 0.5568291060626507", "
Parameterization:
x1: 9.260048242285848
x2: 12.935160705819726", "
Parameterization:
x1: -3.1931319646537304
x2: 3.921633088029921", "
Parameterization:
x1: -1.7758215544745326
x2: 14.612732883542776", "
Parameterization:
x1: 6.232787563351421
x2: 0.0", "
Parameterization:
x1: 3.828656443042684
x2: 3.054375210082606", "
Parameterization:
x1: 2.5135536215481267
x2: 2.036740458858706", "
Parameterization:
x1: 3.275043554719373
x2: 1.8422413342357475", "
Parameterization:
x1: 2.21021545751365
x2: 0.0", "
Parameterization:
x1: 3.1413664148063116
x2: 2.3859949684133146", "
Parameterization:
x1: -5.0
x2: 15.0", "
Parameterization:
x1: 10.0
x2: 0.0", "
Parameterization:
x1: 10.0
x2: 3.44339449425057", "
Parameterization:
x1: 8.954890878695284
x2: 2.906342749771447", "
Parameterization:
x1: 9.903128549987203
x2: 2.3115712966201185", "
Parameterization:
x1: 7.757262617920592
x2: 4.63590439057401", "
Parameterization:
x1: 9.476359778673913
x2: 2.8200750691668395", "
Parameterization:
x1: 9.266267042160726
x2: 2.1272365361611985", "
Parameterization:
x1: -5.0
x2: 10.642874390875647", "
Parameterization:
x1: 2.6310529917106287
x2: 3.6341174771781297", "
Parameterization:
x1: 3.311003670211786
x2: 2.211513685770969", "
Parameterization:
x1: 3.129855359867232
x2: 2.2375976733084926", "
Parameterization:
x1: 9.392837921507155
x2: 2.4807083475177327", "
Parameterization:
x1: 4.59589426359728
x2: 15.0" ], "type": "scatter", "x": [ 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 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811, 0.5928152446626811, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.40071785969624685, 0.40071785969624685, 0.40071785969624685 ] }, { "fill": "tonexty", "fillcolor": "rgba(128,177,211,0.3)", "hoverinfo": "none", "legendgroup": "", "line": { "width": 0 }, "mode": "lines", "showlegend": false, "type": "scatter", "x": [ 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 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811, 0.5928152446626811, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923, 0.40071785969624685, 0.40071785969624685, 0.40071785969624685 ] }, { "line": { "color": "rgba(141,211,199,1)", "dash": "dash" }, "mode": "lines", "name": "model change", "type": "scatter", "x": [ 5, 5 ], "y": [ 0.40071785969624685, 37.28956500488602 ] } ], "layout": { "showlegend": true, "template": { "data": { "bar": [ { "error_x": { "color": "#2a3f5f" }, "error_y": { "color": "#2a3f5f" }, "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "bar" } ], "barpolar": [ { "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "barpolar" } ], "carpet": [ { "aaxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "baxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "type": "carpet" } ], "choropleth": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "choropleth" } ], "contour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "contour" } ], "contourcarpet": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "contourcarpet" } ], "heatmap": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "heatmap" } ], "heatmapgl": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "heatmapgl" } ], "histogram": [ { "marker": { "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "histogram" } ], "histogram2d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "histogram2d" } ], "histogram2dcontour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "histogram2dcontour" } ], "mesh3d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "mesh3d" } ], "parcoords": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "parcoords" } ], "pie": [ { "automargin": true, "type": "pie" } ], "scatter": [ { "fillpattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 }, "type": "scatter" } ], "scatter3d": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatter3d" } ], "scattercarpet": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattercarpet" } ], "scattergeo": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergeo" } ], "scattergl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergl" } ], "scattermapbox": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattermapbox" } ], "scatterpolar": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolar" } ], "scatterpolargl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolargl" } ], "scatterternary": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterternary" } ], "surface": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "surface" } ], "table": [ { "cells": { "fill": { "color": "#EBF0F8" }, "line": { "color": "white" } }, "header": { "fill": { "color": "#C8D4E3" }, "line": { "color": "white" } }, "type": "table" } ] }, "layout": { "annotationdefaults": { "arrowcolor": "#2a3f5f", "arrowhead": 0, "arrowwidth": 1 }, "autotypenumbers": "strict", "coloraxis": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "colorscale": { "diverging": [ [ 0, "#8e0152" ], [ 0.1, "#c51b7d" ], [ 0.2, "#de77ae" ], [ 0.3, "#f1b6da" ], [ 0.4, "#fde0ef" ], [ 0.5, "#f7f7f7" ], [ 0.6, "#e6f5d0" ], [ 0.7, "#b8e186" ], [ 0.8, "#7fbc41" ], [ 0.9, "#4d9221" ], [ 1, "#276419" ] ], "sequential": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "sequentialminus": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ] }, "colorway": [ "#636efa", "#EF553B", "#00cc96", "#ab63fa", "#FFA15A", "#19d3f3", "#FF6692", "#B6E880", "#FF97FF", "#FECB52" ], "font": { "color": "#2a3f5f" }, "geo": { "bgcolor": "white", "lakecolor": "white", "landcolor": "#E5ECF6", "showlakes": true, "showland": true, "subunitcolor": "white" }, "hoverlabel": { "align": "left" }, "hovermode": "closest", "mapbox": { "style": "light" }, "paper_bgcolor": "white", "plot_bgcolor": "#E5ECF6", "polar": { "angularaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "radialaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "scene": { "xaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "yaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "zaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" } }, "shapedefaults": { "line": { "color": "#2a3f5f" } }, "ternary": { "aaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "baxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "caxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "title": { "x": 0.05 }, "xaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 }, "yaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 } } }, "title": { "text": "Model performance vs. # of iterations" }, "xaxis": { "title": { "text": "Iteration" } }, "yaxis": { "title": { "text": "Branin" } } } }, "text/html": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "render(ax_client.get_optimization_trace())" ] }, { "cell_type": "markdown", "id": "5ad94360", "metadata": { "customInput": null, "originalKey": "ca391462-4695-44f1-bc53-070a947c5648", "papermill": { "duration": 0.03933, "end_time": "2023-12-09T19:59:15.861545", "exception": false, "start_time": "2023-12-09T19:59:15.822215", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "# 2. Optimization with global stopping, with the Service API" ] }, { "cell_type": "markdown", "id": "73fdf612", "metadata": { "customInput": null, "originalKey": "5a2690ef-0990-4cbd-9bc9-529b1455a4c3", "papermill": { "duration": 0.11884, "end_time": "2023-12-09T19:59:16.021644", "exception": false, "start_time": "2023-12-09T19:59:15.902804", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "Rather than running a fixed number of trials, we can use a GlobalStoppingStrategy (GSS), which checks whether some stopping criteria have been met when `get_next_trial` is called. Here, we use an `ImprovementGlobalStoppingStrategy`, which checks whether the the last `window_size` trials have improved by more than some threshold amount.\n", "\n", "For single-objective optimization, which we are doing here, `ImprovementGlobalStoppingStrategy` checks if an improvement is \"significant\" by comparing it to the inter-quartile range (IQR) of the objective values attained so far. \n", "\n", "`ImprovementGlobalStoppingStrategy` also supports multi-objective optimization (MOO), in which case it checks whether the percentage improvement in hypervolume over the last `window_size` trials exceeds `improvement_bar`." ] }, { "cell_type": "code", "execution_count": 7, "id": "112b606c", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:16.102016Z", "iopub.status.busy": "2023-12-09T19:59:16.101436Z", "iopub.status.idle": "2023-12-09T19:59:16.105317Z", "shell.execute_reply": "2023-12-09T19:59:16.104714Z" }, "executionStartTime": 1683829356716, "executionStopTime": 1683829356725, "originalKey": "a6634232-448a-4b84-98cd-399c755537df", "papermill": { "duration": 0.045838, "end_time": "2023-12-09T19:59:16.106753", "exception": false, "start_time": "2023-12-09T19:59:16.060915", "status": "completed" }, "requestMsgId": "7e428336-eeeb-4e5b-91c4-fcf5a671773d", "showInput": true, "tags": [] }, "outputs": [], "source": [ "from ax.global_stopping.strategies.improvement import ImprovementGlobalStoppingStrategy\n", "from ax.exceptions.core import OptimizationShouldStop" ] }, { "cell_type": "code", "execution_count": 8, "id": "589aef81", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:16.187022Z", "iopub.status.busy": "2023-12-09T19:59:16.186375Z", "iopub.status.idle": "2023-12-09T19:59:16.190022Z", "shell.execute_reply": "2023-12-09T19:59:16.189479Z" }, "executionStartTime": 1683829356822, "executionStopTime": 1683829356829, "originalKey": "c313de63-03ee-4a65-aa5c-5e7b6f436480", "papermill": { "duration": 0.045806, "end_time": "2023-12-09T19:59:16.191370", "exception": false, "start_time": "2023-12-09T19:59:16.145564", "status": "completed" }, "requestMsgId": "953b064b-8db6-430f-909d-872469bc1e16", "showInput": true, "tags": [] }, "outputs": [], "source": [ "# Start considering stopping only after the 5 initialization trials + 5 GPEI trials.\n", "# Stop if the improvement in the best point in the past 5 trials is less than\n", "# 1% of the IQR thus far.\n", "stopping_strategy = ImprovementGlobalStoppingStrategy(\n", " min_trials=5 + 5, window_size=5, improvement_bar=0.01\n", ")" ] }, { "cell_type": "code", "execution_count": 9, "id": "e27bb412", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:16.272129Z", "iopub.status.busy": "2023-12-09T19:59:16.271481Z", "iopub.status.idle": "2023-12-09T19:59:16.281392Z", "shell.execute_reply": "2023-12-09T19:59:16.280742Z" }, "executionStartTime": 1683829356961, "executionStopTime": 1683829356997, "originalKey": "a2c6c699-f0d2-4001-9bee-3964594e435c", "papermill": { "duration": 0.052875, "end_time": "2023-12-09T19:59:16.283606", "exception": false, "start_time": "2023-12-09T19:59:16.230731", "status": "completed" }, "requestMsgId": "2ba6f82b-1443-4274-83d1-03c56f0190d0", "showInput": true, "tags": [] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[WARNING 12-09 19:59:16] ax.service.ax_client: Random seed set to 0. Note that this setting only affects the Sobol quasi-random generator and BoTorch-powered Bayesian optimization models. For the latter models, setting random seed to the same number for two optimizations will make the generated trials similar, but not exactly the same, and over time the trials will diverge more.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.service.utils.instantiation: Created search space: SearchSpace(parameters=[RangeParameter(name='x1', parameter_type=FLOAT, range=[-5.0, 10.0]), RangeParameter(name='x2', parameter_type=FLOAT, range=[0.0, 15.0])], parameter_constraints=[]).\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.core.experiment: The is_test flag has been set to True. This flag is meant purely for development and integration testing purposes. If you are running a live experiment, please set this flag to False\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.modelbridge.dispatch_utils: Using Models.BOTORCH_MODULAR since there are more ordered parameters than there are categories for the unordered categorical parameters.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.modelbridge.dispatch_utils: Calculating the number of remaining initialization trials based on num_initialization_trials=None max_initialization_trials=None num_tunable_parameters=2 num_trials=None use_batch_trials=False\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.modelbridge.dispatch_utils: calculated num_initialization_trials=5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.modelbridge.dispatch_utils: num_completed_initialization_trials=0 num_remaining_initialization_trials=5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.modelbridge.dispatch_utils: `verbose`, `disable_progbar`, and `jit_compile` are not yet supported when using `choose_generation_strategy` with ModularBoTorchModel, dropping these arguments.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:16] ax.modelbridge.dispatch_utils: Using Bayesian Optimization generation strategy: GenerationStrategy(name='Sobol+BoTorch', steps=[Sobol for 5 trials, BoTorch for subsequent trials]). Iterations after 5 will take longer to generate due to model-fitting.\n" ] } ], "source": [ "ax_client_gss = AxClient(\n", " global_stopping_strategy=stopping_strategy, random_seed=0, verbose_logging=False\n", ")\n", "\n", "ax_client_gss.create_experiment(\n", " name=\"branin_test_experiment\",\n", " parameters=params,\n", " objectives={\"branin\": ObjectiveProperties(minimize=True)},\n", " is_test=True,\n", ")" ] }, { "cell_type": "markdown", "id": "52fee1b3", "metadata": { "customInput": null, "originalKey": "7ff170a1-e885-429f-9695-8b64b5b8e209", "papermill": { "duration": 0.039505, "end_time": "2023-12-09T19:59:16.362435", "exception": false, "start_time": "2023-12-09T19:59:16.322930", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "If there has not been much improvement, `ImprovementGlobalStoppingStrategy` will raise an exception. If the exception is raised, we catch it and terminate optimization." ] }, { "cell_type": "code", "execution_count": 10, "id": "a5616355", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:16.443256Z", "iopub.status.busy": "2023-12-09T19:59:16.442723Z", "iopub.status.idle": "2023-12-09T19:59:20.519120Z", "shell.execute_reply": "2023-12-09T19:59:20.518497Z" }, "executionStartTime": 1683829357114, "executionStopTime": 1683829363866, "originalKey": "3db097cb-1e6e-4320-806a-981dcef6bade", "papermill": { "duration": 4.118799, "end_time": "2023-12-09T19:59:20.520698", "exception": false, "start_time": "2023-12-09T19:59:16.401899", "status": "completed" }, "requestMsgId": "fd039109-2a23-4287-8935-b74274405e56", "showInput": true, "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "The improvement in best objective in the past 5 trials (=0.005) is less than 0.01 times the interquartile range (IQR) of objectives attained so far (IQR=33.434).\n" ] } ], "source": [ "for i in range(25):\n", " try:\n", " parameters, trial_index = ax_client_gss.get_next_trial()\n", " except OptimizationShouldStop as exc:\n", " print(exc.message)\n", " break\n", " ax_client_gss.complete_trial(trial_index=trial_index, raw_data=evaluate(parameters))" ] }, { "cell_type": "code", "execution_count": 11, "id": "1ff1dc0c", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:20.601562Z", "iopub.status.busy": "2023-12-09T19:59:20.600886Z", "iopub.status.idle": "2023-12-09T19:59:20.643450Z", "shell.execute_reply": "2023-12-09T19:59:20.642770Z" }, "executionStartTime": 1683829363988, "executionStopTime": 1683829364103, "originalKey": "ffb53ed2-8775-492d-a357-348957637454", "papermill": { "duration": 0.084487, "end_time": "2023-12-09T19:59:20.644908", "exception": false, "start_time": "2023-12-09T19:59:20.560421", "status": "completed" }, "requestMsgId": "f0f765dd-85db-4519-90d0-064a1bf64b6d", "showInput": true, "tags": [] }, "outputs": [ { "data": { "application/vnd.plotly.v1+json": { "config": { "linkText": "Export to plot.ly", "plotlyServerURL": "https://plot.ly", "showLink": false }, "data": [ { "hoverinfo": "none", "legendgroup": "", "line": { "width": 0 }, "mode": "lines", "showlegend": false, "type": "scatter", "x": [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811, 0.5928152446626811, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923 ] }, { "fill": "tonexty", "fillcolor": "rgba(128,177,211,0.3)", "legendgroup": "objective value", "line": { "color": "rgba(128,177,211,1)" }, "mode": "lines", "name": "objective value", "text": [ "
Parameterization:
x1: 2.126607894897461
x2: 8.887859880924225", "
Parameterization:
x1: 3.681450095027685
x2: 0.5568291060626507", "
Parameterization:
x1: 9.260048242285848
x2: 12.935160705819726", "
Parameterization:
x1: -3.1931319646537304
x2: 3.921633088029921", "
Parameterization:
x1: -1.7758215544745326
x2: 14.612732883542776", "
Parameterization:
x1: 6.232787563351421
x2: 0.0", "
Parameterization:
x1: 3.828656443042684
x2: 3.054375210082606", "
Parameterization:
x1: 2.5135536215481267
x2: 2.036740458858706", "
Parameterization:
x1: 3.275043554719373
x2: 1.8422413342357475", "
Parameterization:
x1: 2.21021545751365
x2: 0.0", "
Parameterization:
x1: 3.1413664148063116
x2: 2.3859949684133146", "
Parameterization:
x1: -5.0
x2: 15.0", "
Parameterization:
x1: 10.0
x2: 0.0" ], "type": "scatter", "x": [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811, 0.5928152446626811, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923 ] }, { "fill": "tonexty", "fillcolor": "rgba(128,177,211,0.3)", "hoverinfo": "none", "legendgroup": "", "line": { "width": 0 }, "mode": "lines", "showlegend": false, "type": "scatter", "x": [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811, 0.5928152446626811, 0.41016834948196923, 0.41016834948196923, 0.41016834948196923 ] }, { "line": { "color": "rgba(141,211,199,1)", "dash": "dash" }, "mode": "lines", "name": "model change", "type": "scatter", "x": [ 5, 5 ], "y": [ 0.41016834948196923, 37.28956500488602 ] } ], "layout": { "showlegend": true, "template": { "data": { "bar": [ { "error_x": { "color": "#2a3f5f" }, "error_y": { "color": "#2a3f5f" }, "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "bar" } ], "barpolar": [ { "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "barpolar" } ], "carpet": [ { "aaxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "baxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "type": "carpet" } ], "choropleth": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "choropleth" } ], "contour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "contour" } ], "contourcarpet": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "contourcarpet" } ], "heatmap": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "heatmap" } ], "heatmapgl": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "heatmapgl" } ], "histogram": [ { "marker": { "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "histogram" } ], "histogram2d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "histogram2d" } ], "histogram2dcontour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "histogram2dcontour" } ], "mesh3d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "mesh3d" } ], "parcoords": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "parcoords" } ], "pie": [ { "automargin": true, "type": "pie" } ], "scatter": [ { "fillpattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 }, "type": "scatter" } ], "scatter3d": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatter3d" } ], "scattercarpet": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattercarpet" } ], "scattergeo": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergeo" } ], "scattergl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergl" } ], "scattermapbox": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattermapbox" } ], "scatterpolar": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolar" } ], "scatterpolargl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolargl" } ], "scatterternary": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterternary" } ], "surface": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "surface" } ], "table": [ { "cells": { "fill": { "color": "#EBF0F8" }, "line": { "color": "white" } }, "header": { "fill": { "color": "#C8D4E3" }, "line": { "color": "white" } }, "type": "table" } ] }, "layout": { "annotationdefaults": { "arrowcolor": "#2a3f5f", "arrowhead": 0, "arrowwidth": 1 }, "autotypenumbers": "strict", "coloraxis": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "colorscale": { "diverging": [ [ 0, "#8e0152" ], [ 0.1, "#c51b7d" ], [ 0.2, "#de77ae" ], [ 0.3, "#f1b6da" ], [ 0.4, "#fde0ef" ], [ 0.5, "#f7f7f7" ], [ 0.6, "#e6f5d0" ], [ 0.7, "#b8e186" ], [ 0.8, "#7fbc41" ], [ 0.9, "#4d9221" ], [ 1, "#276419" ] ], "sequential": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "sequentialminus": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ] }, "colorway": [ "#636efa", "#EF553B", "#00cc96", "#ab63fa", "#FFA15A", "#19d3f3", "#FF6692", "#B6E880", "#FF97FF", "#FECB52" ], "font": { "color": "#2a3f5f" }, "geo": { "bgcolor": "white", "lakecolor": "white", "landcolor": "#E5ECF6", "showlakes": true, "showland": true, "subunitcolor": "white" }, "hoverlabel": { "align": "left" }, "hovermode": "closest", "mapbox": { "style": "light" }, "paper_bgcolor": "white", "plot_bgcolor": "#E5ECF6", "polar": { "angularaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "radialaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "scene": { "xaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "yaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "zaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" } }, "shapedefaults": { "line": { "color": "#2a3f5f" } }, "ternary": { "aaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "baxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "caxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "title": { "x": 0.05 }, "xaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 }, "yaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 } } }, "title": { "text": "Model performance vs. # of iterations" }, "xaxis": { "title": { "text": "Iteration" } }, "yaxis": { "title": { "text": "Branin" } } } }, "text/html": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "render(ax_client_gss.get_optimization_trace())" ] }, { "cell_type": "markdown", "id": "3ff3fd18", "metadata": { "customInput": null, "originalKey": "b01707f3-0bbf-4003-9222-29ba5e3c77b2", "papermill": { "duration": 0.043709, "end_time": "2023-12-09T19:59:20.732040", "exception": false, "start_time": "2023-12-09T19:59:20.688331", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "# 3. Write your own custom Global Stopping Strategy" ] }, { "cell_type": "markdown", "id": "4dd6bebf", "metadata": { "customInput": null, "originalKey": "23b8372b-0067-4934-b599-210b994e06f1", "papermill": { "duration": 0.043634, "end_time": "2023-12-09T19:59:20.818999", "exception": false, "start_time": "2023-12-09T19:59:20.775365", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "You can write a custom Global Stopping Strategy by subclassing `BaseGlobalStoppingStrategy` and use it where `ImprovementGlobalStoppingStrategy` was used above." ] }, { "cell_type": "code", "execution_count": 12, "id": "894d315e", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:20.908255Z", "iopub.status.busy": "2023-12-09T19:59:20.907711Z", "iopub.status.idle": "2023-12-09T19:59:20.911686Z", "shell.execute_reply": "2023-12-09T19:59:20.911144Z" }, "executionStartTime": 1683829364214, "executionStopTime": 1683829364222, "originalKey": "2e5512a9-82ed-43a0-8616-6cee7f648b0f", "papermill": { "duration": 0.049942, "end_time": "2023-12-09T19:59:20.912944", "exception": false, "start_time": "2023-12-09T19:59:20.863002", "status": "completed" }, "requestMsgId": "d5c268a1-fefe-49d5-8ff4-a2cb40fe278b", "showInput": true, "tags": [] }, "outputs": [], "source": [ "from ax.global_stopping.strategies.base import BaseGlobalStoppingStrategy\n", "from typing import Tuple\n", "from ax.core.experiment import Experiment\n", "from ax.core.base_trial import TrialStatus\n", "from ax.global_stopping.strategies.improvement import constraint_satisfaction" ] }, { "cell_type": "markdown", "id": "cce960c5", "metadata": { "customInput": null, "originalKey": "584df5ac-c0f6-4c48-8cec-f9765a04e635", "papermill": { "duration": 0.04314, "end_time": "2023-12-09T19:59:20.999448", "exception": false, "start_time": "2023-12-09T19:59:20.956308", "status": "completed" }, "showInput": false, "tags": [] }, "source": [ "Here, we define `SimpleThresholdGlobalStoppingStrategy`, which stops when we observe a point better than a provided threshold. This can be useful when there is a known optimum. For example, the Branin function has an optimum of zero. When the optimum is not known, this can still be useful from a satisficing perspective: For example, maybe we need a model to take up less than a certain amount of RAM so it doesn't crash our usual hardware, but there is no benefit to further improvements." ] }, { "cell_type": "code", "execution_count": 13, "id": "b8bd2863", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:21.087728Z", "iopub.status.busy": "2023-12-09T19:59:21.087025Z", "iopub.status.idle": "2023-12-09T19:59:21.094036Z", "shell.execute_reply": "2023-12-09T19:59:21.093454Z" }, "executionStartTime": 1683829490325, "executionStopTime": 1683829490340, "originalKey": "bbd24d6e-a873-49d6-abe3-4d832acb8a60", "papermill": { "duration": 0.052889, "end_time": "2023-12-09T19:59:21.095363", "exception": false, "start_time": "2023-12-09T19:59:21.042474", "status": "completed" }, "requestMsgId": "74b77cb7-54eb-4321-afae-942b62b90f5d", "showInput": true, "tags": [] }, "outputs": [], "source": [ "class SimpleThresholdGlobalStoppingStrategy(BaseGlobalStoppingStrategy):\n", " \"\"\"\n", " A GSS that stops when we observe a point better than `threshold`.\n", " \"\"\"\n", " def __init__(\n", " self,\n", " min_trials: int,\n", " inactive_when_pending_trials: bool = True,\n", " threshold: float = 0.1\n", " ):\n", " self.threshold = threshold\n", " super().__init__(\n", " min_trials=min_trials,\n", " inactive_when_pending_trials=inactive_when_pending_trials\n", " )\n", " \n", " def _should_stop_optimization(\n", " self, experiment: Experiment\n", " ) -> Tuple[bool, str]:\n", " \"\"\"\n", " Check if the best seen is better than `self.threshold`.\n", " \"\"\"\n", " feasible_objectives = [\n", " trial.objective_mean\n", " for trial in experiment.trials_by_status[TrialStatus.COMPLETED]\n", " if constraint_satisfaction(trial)\n", " ]\n", "\n", " # Computing the interquartile for scaling the difference\n", " if len(feasible_objectives) <= 1:\n", " message = \"There are not enough feasible arms tried yet.\"\n", " return False, message\n", " \n", " minimize = experiment.optimization_config.objective.minimize\n", " if minimize:\n", " best = np.min(feasible_objectives)\n", " stop = best < self.threshold\n", " else:\n", " best = np.max(feasible_objectives)\n", " stop = best > self.threshold\n", "\n", " comparison = \"less\" if minimize else \"greater\"\n", " if stop:\n", " message = (\n", " f\"The best objective seen is {best:.3f}, which is {comparison} \"\n", " f\"than the threshold of {self.threshold:.3f}.\"\n", " )\n", " else:\n", " message = \"\"\n", "\n", " return stop, message" ] }, { "cell_type": "code", "execution_count": 14, "id": "af37cf61", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:21.184238Z", "iopub.status.busy": "2023-12-09T19:59:21.183759Z", "iopub.status.idle": "2023-12-09T19:59:21.187432Z", "shell.execute_reply": "2023-12-09T19:59:21.186786Z" }, "executionStartTime": 1683829491609, "executionStopTime": 1683829491626, "originalKey": "f3dc5682-0539-4c85-a66a-0d3128f0cc1c", "papermill": { "duration": 0.05044, "end_time": "2023-12-09T19:59:21.188966", "exception": false, "start_time": "2023-12-09T19:59:21.138526", "status": "completed" }, "requestMsgId": "9ee9e413-be32-49fc-a7bc-8e1898d1dbf5", "showInput": true, "tags": [] }, "outputs": [], "source": [ "stopping_strategy = SimpleThresholdGlobalStoppingStrategy(min_trials=5, threshold=1.)" ] }, { "cell_type": "code", "execution_count": 15, "id": "2ee70322", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:21.280104Z", "iopub.status.busy": "2023-12-09T19:59:21.279399Z", "iopub.status.idle": "2023-12-09T19:59:21.289197Z", "shell.execute_reply": "2023-12-09T19:59:21.288540Z" }, "executionStartTime": 1683829491833, "executionStopTime": 1683829491894, "originalKey": "3d6c1ab2-c3ee-49c8-9969-45f2455bbd60", "papermill": { "duration": 0.0564, "end_time": "2023-12-09T19:59:21.291307", "exception": false, "start_time": "2023-12-09T19:59:21.234907", "status": "completed" }, "requestMsgId": "08232010-46f8-4b28-b581-454ddacdc57b", "showInput": true, "tags": [] }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "[WARNING 12-09 19:59:21] ax.service.ax_client: Random seed set to 0. Note that this setting only affects the Sobol quasi-random generator and BoTorch-powered Bayesian optimization models. For the latter models, setting random seed to the same number for two optimizations will make the generated trials similar, but not exactly the same, and over time the trials will diverge more.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.service.utils.instantiation: Created search space: SearchSpace(parameters=[RangeParameter(name='x1', parameter_type=FLOAT, range=[-5.0, 10.0]), RangeParameter(name='x2', parameter_type=FLOAT, range=[0.0, 15.0])], parameter_constraints=[]).\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.core.experiment: The is_test flag has been set to True. This flag is meant purely for development and integration testing purposes. If you are running a live experiment, please set this flag to False\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.modelbridge.dispatch_utils: Using Models.BOTORCH_MODULAR since there are more ordered parameters than there are categories for the unordered categorical parameters.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.modelbridge.dispatch_utils: Calculating the number of remaining initialization trials based on num_initialization_trials=None max_initialization_trials=None num_tunable_parameters=2 num_trials=None use_batch_trials=False\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.modelbridge.dispatch_utils: calculated num_initialization_trials=5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.modelbridge.dispatch_utils: num_completed_initialization_trials=0 num_remaining_initialization_trials=5\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.modelbridge.dispatch_utils: `verbose`, `disable_progbar`, and `jit_compile` are not yet supported when using `choose_generation_strategy` with ModularBoTorchModel, dropping these arguments.\n" ] }, { "name": "stderr", "output_type": "stream", "text": [ "[INFO 12-09 19:59:21] ax.modelbridge.dispatch_utils: Using Bayesian Optimization generation strategy: GenerationStrategy(name='Sobol+BoTorch', steps=[Sobol for 5 trials, BoTorch for subsequent trials]). Iterations after 5 will take longer to generate due to model-fitting.\n" ] } ], "source": [ "ax_client_custom_gss = AxClient(\n", " global_stopping_strategy=stopping_strategy,\n", " random_seed=0,\n", " verbose_logging=False,\n", ")\n", "\n", "ax_client_custom_gss.create_experiment(\n", " name=\"branin_test_experiment\",\n", " parameters=params,\n", " objectives={\"branin\": ObjectiveProperties(minimize=True)},\n", " is_test=True,\n", ")" ] }, { "cell_type": "code", "execution_count": 16, "id": "13a7e791", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:21.380762Z", "iopub.status.busy": "2023-12-09T19:59:21.380073Z", "iopub.status.idle": "2023-12-09T19:59:23.000434Z", "shell.execute_reply": "2023-12-09T19:59:22.999740Z" }, "executionStartTime": 1683829492064, "executionStopTime": 1683829495338, "originalKey": "a306cb15-364f-4e91-b569-9067843a7578", "papermill": { "duration": 1.667026, "end_time": "2023-12-09T19:59:23.001994", "exception": false, "start_time": "2023-12-09T19:59:21.334968", "status": "completed" }, "requestMsgId": "81121dac-3a2a-4dde-b866-44e448e73ad5", "showInput": true, "tags": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "The best objective seen is 0.593, which is less than the threshold of 1.000.\n" ] } ], "source": [ "for i in range(25):\n", " try:\n", " parameters, trial_index = ax_client_custom_gss.get_next_trial()\n", " except OptimizationShouldStop as exc:\n", " print(exc.message)\n", " break\n", " ax_client_custom_gss.complete_trial(\n", " trial_index=trial_index, raw_data=evaluate(parameters)\n", " )" ] }, { "cell_type": "code", "execution_count": 17, "id": "95372186", "metadata": { "customInput": null, "customOutput": null, "execution": { "iopub.execute_input": "2023-12-09T19:59:23.092431Z", "iopub.status.busy": "2023-12-09T19:59:23.091769Z", "iopub.status.idle": "2023-12-09T19:59:23.135893Z", "shell.execute_reply": "2023-12-09T19:59:23.135199Z" }, "executionStartTime": 1683829495351, "executionStopTime": 1683829495740, "originalKey": "3cb59624-d9bb-4b7a-9f57-7cb968dce889", "papermill": { "duration": 0.091511, "end_time": "2023-12-09T19:59:23.137612", "exception": false, "start_time": "2023-12-09T19:59:23.046101", "status": "completed" }, "requestMsgId": "4dd4ed93-07ab-4dd1-92a9-f003f405ccbc", "showInput": true, "tags": [] }, "outputs": [ { "data": { "application/vnd.plotly.v1+json": { "config": { "linkText": "Export to plot.ly", "plotlyServerURL": "https://plot.ly", "showLink": false }, "data": [ { "hoverinfo": "none", "legendgroup": "", "line": { "width": 0 }, "mode": "lines", "showlegend": false, "type": "scatter", "x": [ 1, 2, 3, 4, 5, 6, 7, 8, 9 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811 ] }, { "fill": "tonexty", "fillcolor": "rgba(128,177,211,0.3)", "legendgroup": "objective value", "line": { "color": "rgba(128,177,211,1)" }, "mode": "lines", "name": "objective value", "text": [ "
Parameterization:
x1: 2.126607894897461
x2: 8.887859880924225", "
Parameterization:
x1: 3.681450095027685
x2: 0.5568291060626507", "
Parameterization:
x1: 9.260048242285848
x2: 12.935160705819726", "
Parameterization:
x1: -3.1931319646537304
x2: 3.921633088029921", "
Parameterization:
x1: -1.7758215544745326
x2: 14.612732883542776", "
Parameterization:
x1: 6.232787563351421
x2: 0.0", "
Parameterization:
x1: 3.828656443042684
x2: 3.054375210082606", "
Parameterization:
x1: 2.5135536215481267
x2: 2.036740458858706", "
Parameterization:
x1: 3.275043554719373
x2: 1.8422413342357475" ], "type": "scatter", "x": [ 1, 2, 3, 4, 5, 6, 7, 8, 9 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811 ] }, { "fill": "tonexty", "fillcolor": "rgba(128,177,211,0.3)", "hoverinfo": "none", "legendgroup": "", "line": { "width": 0 }, "mode": "lines", "showlegend": false, "type": "scatter", "x": [ 1, 2, 3, 4, 5, 6, 7, 8, 9 ], "y": [ 37.28956500488602, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 3.545194409651578, 2.8369856015690296, 0.5928152446626811 ] }, { "line": { "color": "rgba(141,211,199,1)", "dash": "dash" }, "mode": "lines", "name": "model change", "type": "scatter", "x": [ 5, 5 ], "y": [ 0.5928152446626811, 37.28956500488602 ] } ], "layout": { "showlegend": true, "template": { "data": { "bar": [ { "error_x": { "color": "#2a3f5f" }, "error_y": { "color": "#2a3f5f" }, "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "bar" } ], "barpolar": [ { "marker": { "line": { "color": "#E5ECF6", "width": 0.5 }, "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "barpolar" } ], "carpet": [ { "aaxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "baxis": { "endlinecolor": "#2a3f5f", "gridcolor": "white", "linecolor": "white", "minorgridcolor": "white", "startlinecolor": "#2a3f5f" }, "type": "carpet" } ], "choropleth": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "choropleth" } ], "contour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "contour" } ], "contourcarpet": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "contourcarpet" } ], "heatmap": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "heatmap" } ], "heatmapgl": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "heatmapgl" } ], "histogram": [ { "marker": { "pattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 } }, "type": "histogram" } ], "histogram2d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "histogram2d" } ], "histogram2dcontour": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "histogram2dcontour" } ], "mesh3d": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "type": "mesh3d" } ], "parcoords": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "parcoords" } ], "pie": [ { "automargin": true, "type": "pie" } ], "scatter": [ { "fillpattern": { "fillmode": "overlay", "size": 10, "solidity": 0.2 }, "type": "scatter" } ], "scatter3d": [ { "line": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatter3d" } ], "scattercarpet": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattercarpet" } ], "scattergeo": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergeo" } ], "scattergl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattergl" } ], "scattermapbox": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scattermapbox" } ], "scatterpolar": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolar" } ], "scatterpolargl": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterpolargl" } ], "scatterternary": [ { "marker": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "type": "scatterternary" } ], "surface": [ { "colorbar": { "outlinewidth": 0, "ticks": "" }, "colorscale": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "type": "surface" } ], "table": [ { "cells": { "fill": { "color": "#EBF0F8" }, "line": { "color": "white" } }, "header": { "fill": { "color": "#C8D4E3" }, "line": { "color": "white" } }, "type": "table" } ] }, "layout": { "annotationdefaults": { "arrowcolor": "#2a3f5f", "arrowhead": 0, "arrowwidth": 1 }, "autotypenumbers": "strict", "coloraxis": { "colorbar": { "outlinewidth": 0, "ticks": "" } }, "colorscale": { "diverging": [ [ 0, "#8e0152" ], [ 0.1, "#c51b7d" ], [ 0.2, "#de77ae" ], [ 0.3, "#f1b6da" ], [ 0.4, "#fde0ef" ], [ 0.5, "#f7f7f7" ], [ 0.6, "#e6f5d0" ], [ 0.7, "#b8e186" ], [ 0.8, "#7fbc41" ], [ 0.9, "#4d9221" ], [ 1, "#276419" ] ], "sequential": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ], "sequentialminus": [ [ 0.0, "#0d0887" ], [ 0.1111111111111111, "#46039f" ], [ 0.2222222222222222, "#7201a8" ], [ 0.3333333333333333, "#9c179e" ], [ 0.4444444444444444, "#bd3786" ], [ 0.5555555555555556, "#d8576b" ], [ 0.6666666666666666, "#ed7953" ], [ 0.7777777777777778, "#fb9f3a" ], [ 0.8888888888888888, "#fdca26" ], [ 1.0, "#f0f921" ] ] }, "colorway": [ "#636efa", "#EF553B", "#00cc96", "#ab63fa", "#FFA15A", "#19d3f3", "#FF6692", "#B6E880", "#FF97FF", "#FECB52" ], "font": { "color": "#2a3f5f" }, "geo": { "bgcolor": "white", "lakecolor": "white", "landcolor": "#E5ECF6", "showlakes": true, "showland": true, "subunitcolor": "white" }, "hoverlabel": { "align": "left" }, "hovermode": "closest", "mapbox": { "style": "light" }, "paper_bgcolor": "white", "plot_bgcolor": "#E5ECF6", "polar": { "angularaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "radialaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "scene": { "xaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "yaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" }, "zaxis": { "backgroundcolor": "#E5ECF6", "gridcolor": "white", "gridwidth": 2, "linecolor": "white", "showbackground": true, "ticks": "", "zerolinecolor": "white" } }, "shapedefaults": { "line": { "color": "#2a3f5f" } }, "ternary": { "aaxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "baxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" }, "bgcolor": "#E5ECF6", "caxis": { "gridcolor": "white", "linecolor": "white", "ticks": "" } }, "title": { "x": 0.05 }, "xaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 }, "yaxis": { "automargin": true, "gridcolor": "white", "linecolor": "white", "ticks": "", "title": { "standoff": 15 }, "zerolinecolor": "white", "zerolinewidth": 2 } } }, "title": { "text": "Model performance vs. # of iterations" }, "xaxis": { "title": { "text": "Iteration" } }, "yaxis": { "title": { "text": "Branin" } } } }, "text/html": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "render(ax_client_custom_gss.get_optimization_trace())" ] }, { "cell_type": "code", "execution_count": null, "id": "4ca80c6e", "metadata": { "customInput": null, "originalKey": "5f4eaa42-a8cb-42b2-b8b4-b2fa53398270", "papermill": { "duration": 0.04841, "end_time": "2023-12-09T19:59:23.234353", "exception": false, "start_time": "2023-12-09T19:59:23.185943", "status": "completed" }, "showInput": true, "tags": [] }, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.18" }, "papermill": { "default_parameters": {}, "duration": 23.993232, "end_time": "2023-12-09T19:59:24.190617", "environment_variables": {}, "exception": null, "input_path": "/tmp/tmp.bwZKabbZig/Ax-main/tutorials/gss.ipynb", "output_path": "/tmp/tmp.bwZKabbZig/Ax-main/tutorials/gss.ipynb", "parameters": {}, "start_time": "2023-12-09T19:59:00.197385", "version": "2.5.0" } }, "nbformat": 4, "nbformat_minor": 5 }