{ "cells": [ { "cell_type": "markdown", "id": "310640b7", "metadata": {}, "source": [ "# Online study\n", "\n", "`OnlineStudy` performs sequential inference: new observations are incorporated one at a time with `step`, and the current posterior becomes the prior for the next update. This is useful when data arrive from a stream and the complete time series is not available upfront.\n", "\n", "The example below reuses the coal mining disaster data, but feeds the observations to the model one after another. We compare a static rate model against a dynamic rate model while storing the full history for plotting." ] }, { "cell_type": "code", "execution_count": 1, "id": "b0c18ae0", "metadata": { "execution": { "iopub.execute_input": "2026-04-27T19:17:27.448597Z", "iopub.status.busy": "2026-04-27T19:17:27.448149Z", "iopub.status.idle": "2026-04-27T19:17:28.441917Z", "shell.execute_reply": "2026-04-27T19:17:28.441404Z" } }, "outputs": [], "source": [ "%matplotlib inline\n", "import matplotlib.pyplot as plt\n", "plt.style.use('seaborn-v0_8-whitegrid') # plot styling\n", "\n", "import bayesloop as bl\n", "\n", "source = bl.Study(silent=True)\n", "source.load_example_data(silent=True)\n", "data = source.raw_data\n", "years = source.raw_timestamps" ] }, { "cell_type": "code", "execution_count": 2, "id": "48527ac0", "metadata": { "execution": { "iopub.execute_input": "2026-04-27T19:17:28.443284Z", "iopub.status.busy": "2026-04-27T19:17:28.443159Z", "iopub.status.idle": "2026-04-27T19:17:28.468680Z", "shell.execute_reply": "2026-04-27T19:17:28.468299Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "+ Added transition model: static (no hyper-parameters)\n", "+ Added transition model: dynamic (2 combination(s) of the following hyper-parameters: ['sigma'])\n", "+ Start model fit\n", " + Set prior (function): jeffreys. Values have been re-normalized.\n" ] } ], "source": [ "S = bl.OnlineStudy(store_history=True, silent=True)\n", "\n", "L = bl.om.Poisson('accident_rate', bl.oint(0, 6, 300))\n", "S.set_observation_model(L, silent=True)\n", "\n", "S.add_transition_model('static', bl.tm.Static())\n", "S.add_transition_model('dynamic', bl.tm.GaussianRandomWalk('sigma', [0.1, 0.3], target='accident_rate'))\n", "S.set_transition_model_prior([0.5, 0.5], silent=True)\n", "\n", "for value in data:\n", " S.step(value)" ] }, { "cell_type": "markdown", "id": "6cb464f1", "metadata": {}, "source": [ "The stored history can be queried just like in a retrospective study. Here we plot the posterior mean of the accident rate and the accumulated probability of the dynamic transition model." ] }, { "cell_type": "code", "execution_count": 3, "id": "dafd423d", "metadata": { "execution": { "iopub.execute_input": "2026-04-27T19:17:28.469723Z", "iopub.status.busy": "2026-04-27T19:17:28.469648Z", "iopub.status.idle": "2026-04-27T19:17:28.603318Z", "shell.execute_reply": "2026-04-27T19:17:28.602848Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "rate_mean = S.get_parameter_mean_values('accident_rate')\n", "dynamic_probability = S.get_transition_model_probabilities('dynamic')\n", "\n", "plt.figure(figsize=(8, 5))\n", "\n", "plt.subplot(211)\n", "plt.bar(years, data, align='center', facecolor='r', alpha=.45)\n", "plt.plot(years, rate_mean, color='k', lw=2)\n", "plt.xlim([1851, 1962])\n", "plt.ylabel('accident rate')\n", "\n", "plt.subplot(212)\n", "plt.plot(years, dynamic_probability, color='g', lw=2)\n", "plt.xlim([1851, 1962])\n", "plt.ylim([0, 1])\n", "plt.xlabel('year')\n", "plt.ylabel('p(dynamic)')\n", "\n", "plt.tight_layout();" ] }, { "cell_type": "markdown", "id": "5fd0d3c3", "metadata": {}, "source": [ "For a larger OnlineStudy use case with multiple transition models and hyperparameter grids, see the [stock market fluctuations example](../examples/stockmarketfluctuations.ipynb)." ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "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": "3.12.4" } }, "nbformat": 4, "nbformat_minor": 5 }