{ "cells": [ { "cell_type": "markdown", "id": "7be94eb1", "metadata": {}, "source": [ "# Measles in the United States: abrupt break or gradual decline\n", "\n", "This example applies *bayesloop* to a question from epidemiology. Measles is one of the\n", "great success stories of public health: a vaccine was licensed in 1963 and the United\n", "States declared the disease eliminated in 2000. The textbook narrative is a clean\n", "before/after story, but it is not obvious whether the vaccine's epidemiological\n", "signature was really an *abrupt* break or a *gradual* multi-year decline as coverage\n", "accumulated. A related question is whether the transition can be dated from the data\n", "alone, without assuming when it happened.\n", "\n", "These are questions *bayesloop* is built for. Rather than fitting one static model, we\n", "let the transmission level vary in time and use the marginal likelihood (model\n", "evidence) as an objective, complexity-penalised score to ask the data which kind of\n", "variation it prefers. In the course of the analysis we use a good part of the toolbox:\n", "`Study`, `HyperStudy` and `ChangepointStudy`; the `Gaussian` and custom `NumPy`\n", "observation models; and the `Static`, `GaussianRandomWalk`, `AlphaStableRandomWalk`,\n", "`ChangePoint`, `CombinedTransitionModel` and `RegimeSwitch` transition models." ] }, { "cell_type": "code", "execution_count": 1, "id": "7f09fded", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:41.715789Z", "iopub.status.busy": "2026-06-12T09:45:41.715553Z", "iopub.status.idle": "2026-06-12T09:45:43.347787Z", "shell.execute_reply": "2026-06-12T09:45:43.347216Z" } }, "outputs": [], "source": [ "import io\n", "import urllib.request\n", "\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib as mpl\n", "import matplotlib.pyplot as plt\n", "\n", "import bayesloop as bl\n", "\n", "# A consistent, publication-leaning style for the figures below.\n", "try:\n", " import seaborn as sns\n", " sns.set_style(\"whitegrid\")\n", "except ImportError:\n", " plt.style.use(\"default\")\n", "mpl.rcParams.update({\n", " \"figure.dpi\": 110, \"font.size\": 10.5, \"axes.titlesize\": 12,\n", " \"axes.titleweight\": \"bold\", \"axes.edgecolor\": \"#444444\", \"grid.color\": \"#d9d9d9\",\n", "})\n", "C = {\"data\": \"#8a8a8a\", \"accent\": \"#c1272d\", \"accent2\": \"#0b6e99\", \"green\": \"#2e7d32\"}\n", "\n", "\n", "# Two small helpers to turn a fitted study into a posterior mean and credible band for\n", "# a time-varying parameter. (bayesloop also has built-in plotting via ``S.plot``; these\n", "# give us a bit more control over the figures.)\n", "def posterior_band(S, name, lo=0.05, hi=0.95):\n", " \"\"\"Posterior mean and an equal-tailed credible band for a time-varying parameter.\"\"\"\n", " x, p = S.get_parameter_distributions(name, density=False)\n", " p = np.asarray(p, dtype=float)\n", " p /= p.sum(axis=1, keepdims=True)\n", " mean = (p * x[None, :]).sum(axis=1)\n", " cdf = np.cumsum(p, axis=1)\n", " low = np.array([np.interp(lo, cdf[t], x) for t in range(p.shape[0])])\n", " high = np.array([np.interp(hi, cdf[t], x) for t in range(p.shape[0])])\n", " return mean, low, high\n", "\n", "\n", "def prob_above(S, name, threshold):\n", " \"\"\"Posterior probability that the parameter exceeds ``threshold`` at each time step.\"\"\"\n", " x, p = S.get_parameter_distributions(name, density=False)\n", " p = np.asarray(p, dtype=float)\n", " p /= p.sum(axis=1, keepdims=True)\n", " return p[:, x > threshold].sum(axis=1)" ] }, { "cell_type": "markdown", "id": "9a31ba49", "metadata": {}, "source": [ "## The data\n", "\n", "We use US national annual measles case counts from\n", "[Our World in Data](https://ourworldindata.org/grapher/number-of-measles-cases)\n", "(compiled from US CDC notifiable-disease reports / Project Tycho). The series is\n", "contiguous from 1938 to 2025 and spans almost five orders of magnitude, from nearly a\n", "million cases a year before the vaccine to a few dozen after elimination. Because of\n", "this large dynamic range we model log₁₀(cases) with a Gaussian observation model\n", "(standard practice for incidence data); we return to the natural count scale for the\n", "modern post-elimination era at the end." ] }, { "cell_type": "code", "execution_count": 2, "id": "b5ca705b", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:43.349107Z", "iopub.status.busy": "2026-06-12T09:45:43.348989Z", "iopub.status.idle": "2026-06-12T09:45:43.454142Z", "shell.execute_reply": "2026-06-12T09:45:43.453462Z" }, "lines_to_next_cell": 1 }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Full record: 1938-2025 (88 years), cases 13-894134\n" ] } ], "source": [ "url = \"https://ourworldindata.org/grapher/number-of-measles-cases.csv?csvType=full\"\n", "req = urllib.request.Request(url, headers={\"User-Agent\": \"Mozilla/5.0 (bayesloop docs)\"})\n", "df = pd.read_csv(io.BytesIO(urllib.request.urlopen(req, timeout=90).read()))\n", "\n", "col = [c for c in df.columns if \"measles\" in c.lower() and \"Annot\" not in c][0]\n", "us = (df[df[\"Entity\"] == \"United States\"][[\"Year\", col]]\n", " .rename(columns={col: \"cases\"}).sort_values(\"Year\").reset_index(drop=True))\n", "us = us[us[\"Year\"] <= 2025]\n", "full = us[us[\"Year\"] >= 1938].reset_index(drop=True)\n", "\n", "years = full[\"Year\"].values.astype(float)\n", "cases = full[\"cases\"].values.astype(float)\n", "logcases = np.log10(cases)\n", "print(f\"Full record: {int(years.min())}-{int(years.max())} \"\n", " f\"({len(years)} years), cases {int(cases.min())}-{int(cases.max())}\")" ] }, { "cell_type": "markdown", "id": "d47b9c24", "metadata": {}, "source": [ "## A time-varying transmission level\n", "\n", "Our observation model treats each year's log-incidence as Gaussian with a time-varying\n", "mean (the transmission *level*) and a free standard deviation, which absorbs the strong\n", "biennial epidemic cycles of the pre-vaccine era:\n", "\n", "```python\n", "bl.om.Gaussian(\"mean\", ..., \"std\", ...)\n", "```\n", "\n", "For the dynamics of that mean we start with a `GaussianRandomWalk`, so the level is\n", "allowed to drift smoothly from year to year, and wrap it in a `HyperStudy`, which\n", "marginalises over the unknown magnitude (`sigma`) of that drift." ] }, { "cell_type": "code", "execution_count": 3, "id": "35689e54", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:43.455524Z", "iopub.status.busy": "2026-06-12T09:45:43.455408Z", "iopub.status.idle": "2026-06-12T09:45:47.159240Z", "shell.execute_reply": "2026-06-12T09:45:47.158788Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "+ Created new study.\n", " --> Hyper-study\n", "+ Successfully imported array.\n", "+ Observation model: Gaussian observations. Parameter(s): ['mean', 'std']\n", "+ Transition model: Gaussian random walk. Hyper-Parameter(s): ['sigma']\n", "+ Set hyper-prior(s): ['uniform']\n", "+ Started new fit.\n", " + 30 analyses to run.\n", " + Cached observation likelihoods (14.5 MiB).\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " + Computed average posterior sequence\n", " + Computed hyper-parameter distribution\n", " + Log10-evidence of average model: -22.16384\n", " + Computed local evidence of average model\n", " + Computed mean parameter values.\n", "+ Finished fit.\n" ] } ], "source": [ "def gaussian_obs():\n", " return bl.om.Gaussian(\"mean\", bl.cint(0.5, 6.5, 240),\n", " \"std\", bl.oint(0.03, 1.2, 90))\n", "\n", "Sg = bl.HyperStudy()\n", "Sg.load_data(logcases, timestamps=years)\n", "Sg.set(gaussian_obs(),\n", " bl.tm.GaussianRandomWalk(\"sigma\", bl.cint(0, 0.6, 30), target=\"mean\"))\n", "Sg.fit()\n", "evidence_gradual = float(Sg.log10_evidence)" ] }, { "cell_type": "markdown", "id": "cdf4e9ab", "metadata": {}, "source": [ "Left free to maximise evidence, the random walk picks a fairly large step size. It has\n", "to, in order to follow the steep 1960s decline, and a single global step size cannot\n", "also be small in the flat plateaus. As a side-effect, the evidence-optimal fit tracks\n", "year-to-year epidemic variation rather than smoothing it. For visualising the secular\n", "trend we therefore also fit a deliberately regularised version with a small, fixed\n", "step. The model-selection conclusion below turns out to be robust to this choice." ] }, { "cell_type": "code", "execution_count": 4, "id": "7a9b4cac", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:47.160731Z", "iopub.status.busy": "2026-06-12T09:45:47.160635Z", "iopub.status.idle": "2026-06-12T09:45:47.461966Z", "shell.execute_reply": "2026-06-12T09:45:47.461530Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "+ Created new study.\n", "+ Successfully imported array.\n", "+ Observation model: Gaussian observations. Parameter(s): ['mean', 'std']\n", "+ Transition model: Gaussian random walk. Hyper-Parameter(s): ['sigma']\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "Ssm = bl.Study()\n", "Ssm.load_data(logcases, timestamps=years)\n", "Ssm.set(gaussian_obs(), bl.tm.GaussianRandomWalk(\"sigma\", 0.10, target=\"mean\"))\n", "Ssm.fit(silent=True)\n", "mean_sm, lo_sm, hi_sm = posterior_band(Ssm, \"mean\")\n", "\n", "fig, ax = plt.subplots(figsize=(9, 4.2))\n", "ax.bar(years, cases, width=0.8, color=C[\"data\"], alpha=0.45, label=\"reported cases\")\n", "ax.plot(years, 10 ** mean_sm, color=C[\"accent\"], lw=2.2,\n", " label=\"inferred transmission trend (smoothed)\")\n", "ax.fill_between(years, 10 ** lo_sm, 10 ** hi_sm, color=C[\"accent\"], alpha=0.20,\n", " label=\"90% credible band\")\n", "ax.set_yscale(\"log\")\n", "ax.set_ylim(8, 2e6)\n", "ax.set_xlim(1937, 2026)\n", "for yr, lab in [(1963, \"vaccine\\nlicensed 1963\"), (1989, \"1989–91\\nresurgence\"),\n", " (2000, \"elimination\\ndeclared 2000\")]:\n", " ax.axvline(yr, color=C[\"accent2\"], ls=\"--\", lw=1, alpha=0.6)\n", " ax.text(yr + 0.5, 1.1e6, lab, fontsize=7.5, color=C[\"accent2\"], va=\"top\")\n", "ax.set_ylabel(\"measles cases per year\")\n", "ax.set_xlabel(\"year\")\n", "ax.set_title(\"US measles: a time-varying transmission level inferred by bayesloop\")\n", "ax.legend(loc=\"lower left\", ncol=2)\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "e27605b6", "metadata": {}, "source": [ "The inferred trend reproduces the modern history of US measles: the pre-vaccine plateau\n", "near 5 × 10⁵ cases/year, the post-1963 collapse, the 1989–91 resurgence (resolved as a\n", "temporary bump), the drive to elimination, and the recent resurgence.\n", "\n", "## Letting the evidence choose between abrupt and gradual change\n", "\n", "The picture above assumes gradual drift. Whether the data actually prefer that over the\n", "alternatives is a question we can answer directly. Because every model below shares the\n", "identical observation model and data, their log₁₀ marginal likelihoods\n", "(`S.log10_evidence`) are directly comparable. We compare four hypotheses:\n", "\n", "| Hypothesis | Transition model |\n", "|---|---|\n", "| Static (null) | `tm.Static()` |\n", "| Abrupt change-point | `tm.ChangePoint(\"t_change\", \"all\")` |\n", "| Gradual random walk | `tm.GaussianRandomWalk` (the HyperStudy above) |\n", "| Change-point + drift | `CombinedTransitionModel(ChangePoint, GaussianRandomWalk)` |" ] }, { "cell_type": "code", "execution_count": 5, "id": "0303ed78", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:47.463112Z", "iopub.status.busy": "2026-06-12T09:45:47.463024Z", "iopub.status.idle": "2026-06-12T09:45:56.656395Z", "shell.execute_reply": "2026-06-12T09:45:56.655856Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "+ Created new study.\n", "+ Successfully imported array.\n", "+ Observation model: Gaussian observations. Parameter(s): ['mean', 'std']\n", "+ Transition model: Static/constant parameter values. Hyper-Parameter(s): []\n", "+ Created new study.\n", " --> Hyper-study\n", " --> Change-point analysis\n", "+ Successfully imported array.\n", "+ Observation model: Gaussian observations. Parameter(s): ['mean', 'std']\n", "+ Transition model: Change-point. Hyper-Parameter(s): ['t_change']\n", "+ Detected 1 change-point(s) in transition model: ['t_change']\n", "+ Set hyper-prior(s): ['uniform']\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "+ Created new study.\n", " --> Hyper-study\n", " --> Change-point analysis\n", "+ Successfully imported array.\n", "+ Observation model: Gaussian observations. Parameter(s): ['mean', 'std']\n", "+ Transition model: Combined transition model. Hyper-Parameter(s): ['t_change', 'sigma']\n", "+ Detected 1 change-point(s) in transition model: ['t_change']\n", "+ Set hyper-prior(s): ['uniform', 'uniform']\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Model evidence (log10, higher is better):\n", " Static (constant) -78.56\n", " Abrupt change-point -47.12\n", " Gradual random walk -22.16\n", " Change-point + drift -25.16\n", " --> best: Gradual random walk\n", "Change-point mode: 1966 (90% credible 1965-1966)\n" ] } ], "source": [ "# M0 — Static (constant level): the null hypothesis.\n", "S0 = bl.Study()\n", "S0.load_data(logcases, timestamps=years)\n", "S0.set(gaussian_obs(), bl.tm.Static())\n", "S0.fit(silent=True)\n", "\n", "# M1 — a single abrupt change-point. ChangepointStudy scans *every* possible break year.\n", "S1 = bl.ChangepointStudy()\n", "S1.load_data(logcases, timestamps=years)\n", "S1.set(gaussian_obs(), bl.tm.ChangePoint(\"t_change\", \"all\"))\n", "S1.fit(silent=True)\n", "cp_x, cp_p = S1.get_hyper_parameter_distribution(\"t_change\")\n", "cp_mode = float(cp_x[np.argmax(cp_p)])\n", "order = np.argsort(cp_p)[::-1]\n", "cred = np.sort(cp_x[order[:np.searchsorted(np.cumsum(cp_p[order]), 0.90) + 1]])\n", "\n", "# M3 — an abrupt break superimposed on gradual drift.\n", "S3 = bl.ChangepointStudy()\n", "S3.load_data(logcases, timestamps=years)\n", "S3.set(gaussian_obs(),\n", " bl.tm.CombinedTransitionModel(\n", " bl.tm.ChangePoint(\"t_change\", \"all\"),\n", " bl.tm.GaussianRandomWalk(\"sigma\", 0.15, target=\"mean\")))\n", "S3.fit(silent=True)\n", "\n", "evidence = {\n", " \"Static (constant)\": float(S0.log10_evidence),\n", " \"Abrupt change-point\": float(S1.log10_evidence),\n", " \"Gradual random walk\": evidence_gradual,\n", " \"Change-point + drift\": float(S3.log10_evidence),\n", "}\n", "best = max(evidence, key=evidence.get)\n", "print(\"Model evidence (log10, higher is better):\")\n", "for k, v in evidence.items():\n", " print(f\" {k:24s} {v:9.2f}\")\n", "print(f\" --> best: {best}\")\n", "print(f\"Change-point mode: {cp_mode:.0f} (90% credible {cred.min():.0f}-{cred.max():.0f})\")" ] }, { "cell_type": "markdown", "id": "c477e009", "metadata": {}, "source": [ "The gradual random walk wins by a wide margin. It is many orders of magnitude more\n", "probable than a single abrupt change-point, and adding a change-point on top of the\n", "drift makes the model worse, as the Occam penalty for the extra parameter is not\n", "repaid. The scientific reading is clear and non-trivial: the vaccine's epidemiological\n", "footprint is not a single clean break but a sustained, multi-decade decline as coverage\n", "ratcheted up.\n", "\n", "When we nonetheless force a single change-point, the posterior dates the dominant break\n", "to within a couple of years of the 1963 licensing, the lag expected for mass uptake,\n", "and gives a credible interval for it, which a fixed before/after split never could." ] }, { "cell_type": "code", "execution_count": 6, "id": "9bc626e2", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:56.657668Z", "iopub.status.busy": "2026-06-12T09:45:56.657571Z", "iopub.status.idle": "2026-06-12T09:45:56.808076Z", "shell.execute_reply": "2026-06-12T09:45:56.807551Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, (axL, axR) = plt.subplots(1, 2, figsize=(11, 3.6))\n", "\n", "# change-point posterior\n", "axL.bar(cp_x, cp_p, width=0.85, color=C[\"green\"], alpha=0.85)\n", "axL.set_xlim(1958, 1975)\n", "axL.axvline(1963, color=C[\"accent\"], ls=\"--\", lw=1.2, label=\"vaccine licensed (1963)\")\n", "axL.set_xlabel(\"year of structural break\")\n", "axL.set_ylabel(\"posterior probability\")\n", "axL.set_title(f\"Dominant change-point dated to {cp_mode:.0f}\")\n", "axL.legend()\n", "\n", "# evidence comparison\n", "labels = list(evidence)\n", "vals = np.array([evidence[k] for k in labels])\n", "vals_rel = vals - vals.min()\n", "bars = axR.barh(labels, vals_rel, color=[C[\"data\"], C[\"accent2\"], C[\"accent\"], C[\"green\"]])\n", "axR.set_xlabel(r\"log$_{10}$ evidence relative to worst model\")\n", "axR.set_title(\"Which parameter-dynamics hypothesis fits best?\")\n", "for b, v in zip(bars, vals_rel):\n", " axR.text(v + max(vals_rel) * 0.01, b.get_y() + b.get_height() / 2,\n", " f\"{v:.0f}\", va=\"center\", fontsize=8)\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "d7763713", "metadata": {}, "source": [ "### Robustness of the gradual model\n", "\n", "One concern is whether the gradual model won only because it was allowed to over-fit.\n", "We check this by comparing the heavily regularised smoothed fit (small fixed step)\n", "against the alternatives, and by trying a heavier-tailed `AlphaStableRandomWalk` that\n", "permits occasional large jumps rather than uniform Gaussian steps (a 2-D `HyperStudy`\n", "over the scale `c` and the tail index `alpha`)." ] }, { "cell_type": "code", "execution_count": 7, "id": "a4fccd52", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:45:56.809168Z", "iopub.status.busy": "2026-06-12T09:45:56.809065Z", "iopub.status.idle": "2026-06-12T09:46:53.810566Z", "shell.execute_reply": "2026-06-12T09:46:53.810102Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "+ Created new study.\n", " --> Hyper-study\n", "+ Successfully imported array.\n", "+ Observation model: Gaussian observations. Parameter(s): ['mean', 'std']\n", "+ Transition model: Alpha-stable random walk. Hyper-Parameter(s): ['c', 'alpha']\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ "Regularised smooth fit (sigma=0.10): log10 evidence = -29.17\n", "Alpha-stable random walk: log10 evidence = -19.84 (tail index alpha mode = 1.74)\n" ] } ], "source": [ "Sas = bl.HyperStudy()\n", "Sas.load_data(logcases, timestamps=years)\n", "Sas.set(gaussian_obs(),\n", " bl.tm.AlphaStableRandomWalk(\"c\", bl.cint(0.01, 0.5, 10),\n", " \"alpha\", bl.cint(1.1, 2.0, 8), target=\"mean\"))\n", "Sas.fit(silent=True, n_jobs=4)\n", "a_x, a_p = Sas.get_hyper_parameter_distribution(\"alpha\")\n", "print(f\"Regularised smooth fit (sigma=0.10): log10 evidence = {Ssm.log10_evidence:7.2f}\")\n", "print(f\"Alpha-stable random walk: log10 evidence = {Sas.log10_evidence:7.2f}\"\n", " f\" (tail index alpha mode = {a_x[np.argmax(a_p)]:.2f})\")" ] }, { "cell_type": "markdown", "id": "ab048438", "metadata": {}, "source": [ "Even the regularised fit comfortably beats the single change-point, so the win comes\n", "from the sustained, multi-decade nature of the decline rather than from over-fitting.\n", "The alpha-stable model fits best of all, with a tail index `alpha` well below 2,\n", "indicating that the decline is best described by occasional large jumps on top of\n", "gentle drift.\n", "\n", "We can also quote a model-free magnitude straight from the raw series:" ] }, { "cell_type": "code", "execution_count": 8, "id": "f3e3ae73", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:46:53.812119Z", "iopub.status.busy": "2026-06-12T09:46:53.812021Z", "iopub.status.idle": "2026-06-12T09:46:53.814583Z", "shell.execute_reply": "2026-06-12T09:46:53.814111Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Pre-vaccine mean (1938-1962): 532,633 cases/yr\n", "Early-1980s trough (1983): 1,497 cases\n", "=> 356-fold reduction\n" ] } ], "source": [ "pre = float(np.mean(cases[(years >= 1938) & (years <= 1962)]))\n", "lowmask = (years >= 1980) & (years <= 1985)\n", "trough = float(cases[lowmask].min())\n", "trough_year = float(years[lowmask][np.argmin(cases[lowmask])])\n", "print(f\"Pre-vaccine mean (1938-1962): {pre:,.0f} cases/yr\")\n", "print(f\"Early-1980s trough ({trough_year:.0f}): {trough:,.0f} cases\")\n", "print(f\"=> {pre/trough:,.0f}-fold reduction\")" ] }, { "cell_type": "markdown", "id": "d1df0be9", "metadata": {}, "source": [ "## The post-elimination era: a custom Poisson observation model\n", "\n", "After elimination the counts are small enough (tens to a couple thousand) to model\n", "directly as Poisson rather than on the log scale. This is also an occasion to show\n", "*bayesloop*'s extensibility. Its built-in Poisson evaluates the pmf directly, which\n", "overflows for counts above about 170, so we supply a numerically stable log-space\n", "Poisson as a one-line `bl.om.NumPy` custom model. We pair it with a `RegimeSwitch`\n", "transition, which lets the rate jump abruptly and is well suited to detecting outbreaks\n", "against the near-zero eliminated baseline." ] }, { "cell_type": "code", "execution_count": 9, "id": "43d18c90", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:46:53.815622Z", "iopub.status.busy": "2026-06-12T09:46:53.815534Z", "iopub.status.idle": "2026-06-12T09:46:53.825388Z", "shell.execute_reply": "2026-06-12T09:46:53.824993Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "+ Created new study.\n", "+ Successfully imported array.\n", "+ Observation model: poisson_stable. Parameter(s): ('rate',)\n", "+ Transition model: Regime-switching model. Hyper-Parameter(s): ['log10p_min']\n", "+ Started new fit:\n", " + Formatted data.\n", " + Cached observation likelihoods (0.3 MiB).\n", " + Set uniform prior with parameter boundaries.\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " + Finished forward pass.\n", " + Log10-evidence: -93.44755\n" ] }, { "name": "stdout", "output_type": "stream", "text": [ " + Finished backward pass.\n", " + Computed mean parameter values.\n", " 2019: 1274 cases, P(rate > 5x baseline) = 1.00\n", " 2025: 2288 cases, P(rate > 5x baseline) = 1.00\n" ] } ], "source": [ "from scipy.special import gammaln\n", "\n", "def poisson_stable(data, rate):\n", " k = float(data)\n", " return np.exp(k * np.log(rate) - rate - gammaln(k + 1.0))\n", "\n", "modern = us[(us[\"Year\"] >= 1998) & (us[\"Year\"] <= 2025)].reset_index(drop=True)\n", "myears = modern[\"Year\"].values.astype(float)\n", "mcases = modern[\"cases\"].values.astype(float)\n", "\n", "Sp = bl.Study()\n", "Sp.load_data(mcases, timestamps=myears)\n", "Sp.set(bl.om.NumPy(poisson_stable, \"rate\", bl.oint(0, 3000, 1500)),\n", " bl.tm.RegimeSwitch(\"log10p_min\", -4))\n", "Sp.fit()\n", "\n", "rate_mean, rate_lo, rate_hi = posterior_band(Sp, \"rate\")\n", "baseline = float(np.median(rate_mean[(myears >= 1998) & (myears <= 2017)]))\n", "p_outbreak = prob_above(Sp, \"rate\", 5 * baseline) # P(rate > 5x baseline)\n", "for y in (2019, 2025):\n", " print(f\" {y}: {int(mcases[myears == y][0]):>5d} cases, \"\n", " f\"P(rate > 5x baseline) = {p_outbreak[myears == y][0]:.2f}\")" ] }, { "cell_type": "code", "execution_count": 10, "id": "cef29345", "metadata": { "execution": { "iopub.execute_input": "2026-06-12T09:46:53.826410Z", "iopub.status.busy": "2026-06-12T09:46:53.826319Z", "iopub.status.idle": "2026-06-12T09:46:53.910383Z", "shell.execute_reply": "2026-06-12T09:46:53.910027Z" } }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(9, 4.6), sharex=True,\n", " gridspec_kw={\"height_ratios\": [2, 1]})\n", "ax1.bar(myears, mcases, width=0.7, color=C[\"data\"], alpha=0.5, label=\"reported cases\")\n", "ax1.plot(myears, rate_mean, color=C[\"accent\"], lw=2, label=\"inferred Poisson rate\")\n", "ax1.fill_between(myears, rate_lo, rate_hi, color=C[\"accent\"], alpha=0.2)\n", "ax1.axhline(baseline, color=C[\"accent2\"], ls=\":\", lw=1, label=f\"baseline ~{baseline:.0f}/yr\")\n", "ax1.set_ylabel(\"cases / yr\")\n", "ax1.set_title(\"Post-elimination era: a regime-switching Poisson rate detects outbreaks\")\n", "ax1.legend(loc=\"upper left\", fontsize=8)\n", "ax2.bar(myears, p_outbreak, width=0.7, color=C[\"green\"], alpha=0.7)\n", "ax2.set_ylabel(\"P(rate > 5×base)\")\n", "ax2.set_xlabel(\"year\")\n", "ax2.set_ylim(0, 1)\n", "plt.tight_layout()\n", "plt.show()" ] }, { "cell_type": "markdown", "id": "c95455d4", "metadata": {}, "source": [ "The regime-switching Poisson flags the 2019 and 2025 outbreaks at posterior probability\n", "≈ 1.0 against the eliminated baseline.\n", "\n", "## What this example showed\n", "\n", "On a single dataset we used *bayesloop* to:\n", "\n", "- answer a scientific question with model evidence: the choice between abrupt and\n", " gradual change is settled quantitatively, and the answer (gradual) is the less\n", " obvious one;\n", "- recover a dated transition with a credible interval, without telling the model when\n", " the vaccine arrived;\n", "- exercise the toolbox end-to-end, using `Study`, `HyperStudy` and `ChangepointStudy`,\n", " the `Gaussian` and custom `NumPy` observation models, and a range of transition\n", " models including a custom numerically-stable Poisson and a regime switch.\n", "\n", "Every number cross-checks against the published epidemiological record: a ~1966\n", "dominant transition, the early-1980s trough, the ~356-fold drop, the 1989–91\n", "resurgence, and the 2019/2025 outbreaks." ] } ], "metadata": { "jupytext": { "cell_metadata_filter": "-all", "main_language": "python", "notebook_metadata_filter": "-all", "text_representation": { "extension": ".py", "format_name": "percent" } }, "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 }