{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "853d7d3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "(v6_announcement)=\n",
    "\n",
    "# PyMC 6.0 & PyTensor 3.0: ecosystem updates\n",
    "\n",
    ":::{post} May 11, 2026\n",
    ":tags: release, pytensor, numba, jax\n",
    ":category: news\n",
    ":author: PyMC Contributors\n",
    ":::\n",
    "\n",
    "PyMC has been under steady development since the early 2010s.\n",
    "To mark the new major releases of PyMC 6.0 and PyTensor 3.0,\n",
    "we want to highlight the developments across the PyMC ecosystem\n",
    "(and its close cousin, the ArviZ ecosystem) that we're most excited about.\n",
    "\n",
    ":::{image} pymc_pytensor_logos.png\n",
    ":alt: PyMC and PyTensor logos\n",
    ":width: 700px\n",
    ":::\n",
    "\n",
    "The short version, before we dig in:\n",
    "\n",
    "PyMC 6.0 is `pip install`-able with no extra system setup. The default computational\n",
    "backend is now Numba; C and JAX remain available on demand.\n",
    "\n",
    "NUTS sampling is roughly **2x faster** end-to-end, thanks to the new nutpie default sampler.\n",
    "pymc-extras adds Pathfinder and DADVI for variational inference.\n",
    "\n",
    "The new `pymc.dims` module lets you write entire models against named dimensions.\n",
    "pymc-extras covers automatic marginalization of discrete latents and a high-level\n",
    "state-space API, and PyMC-BART covers Bayesian Additive Regression Trees.\n",
    "\n",
    "`pytensor.wrap_jax` brings arbitrary JAX functions into PyMC models, and PyMC's\n",
    "growing automatic logp derivation lets you build new distributions from existing ones.\n",
    "\n",
    "ArviZ has reached 1.0 with a redesigned plotting library and `xarray.DataTree`-backed\n",
    "inference data. Kulprit brings projective variable selection, and PreliZ continues to\n",
    "grow as the prior-elicitation companion."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14f8dca1",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Performance\n",
    "\n",
    "PyMC hasn't always been a breeze to use. It's written in Python, and its backend\n",
    "(Theano, and its successor PyTensor) had accrued some weight over the years.\n",
    "\n",
    "A lot of effort was devoted to speeding up all stages of work:\n",
    "from import times, to model compilation and sampling runtime."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5cfe412b",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Numba is the new default backend\n",
    "\n",
    "Historically, PyMC compiled model functions to **C**. In 6.0 we switch the default linker\n",
    "to **Numba**. For most CPU workloads Numba beats C, while being far easier to\n",
    "maintain and extend.\n",
    "It also unlocks a few things the old C path always struggled with:\n",
    "\n",
    "- **Native advanced indexing** — the hierarchical-modelling staple that the C\n",
    "  backend has always struggled with. Numba supports (almost) all cases natively\n",
    "  and is noticeably faster.\n",
    "- **LAPACK bindings for linear algebra** — `cholesky`, `solve`, `eigh`, and\n",
    "  friends aren't bottlenecked by Python.\n",
    "- **Fast `scan`** — historically the slowest piece of the C backend, now far\n",
    "  closer to looping in a low-level language (with autodiff). A big win for\n",
    "  time-series and other recursive-structure models.\n",
    "- **Native sparse support** — useful for spatial models (ICAR), state-space\n",
    "  models, and INLA-style inference.\n",
    "\n",
    "For GPU-based sampling, the **JAX backend** remains the recommended backend.\n",
    "\n",
    "A knock-on win: **PyMC is finally `pip install`-safe.** The old C backend needed a system\n",
    "BLAS installation *and* a C compiler, both beyond pip's reach. The first-time experience on a\n",
    "clean machine used to greet you with a prominent \"PyTensor could not link to a BLAS\n",
    "installation\" warning and degraded performance (or worse if a compiler couldn't be detected).\n",
    "Conda/mamba was the only reliable path."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a73c1f06",
   "metadata": {},
   "outputs": [],
   "source": [
    "### One-liner backend selection with `backend=`\n",
    "\n",
    "You can pass `backend=` to **every** sampling and inference entry point\n",
    "(`pm.sample`, `pm.sample_posterior_predictive`, `pm.fit`, and so on).\n",
    "\n",
    "Accepted values include `\"numba\"`, `\"c\"`, and `\"jax\"`.\n",
    "\n",
    "The default backend of `pm.sample` is `\"numba\"`. You only need to pass it explicitly\n",
    "when picking something else.\n",
    "\n",
    "**Availability:** Numba and C come with PyMC, though C needs both a compiler\n",
    "and a working BLAS installation at runtime to be usable. JAX has to be installed separately."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e1692880-a776-48d6-a64b-a06d647b23e4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:37.751977Z",
     "iopub.status.busy": "2026-05-11T17:04:37.751748Z",
     "iopub.status.idle": "2026-05-11T17:04:40.957431Z",
     "shell.execute_reply": "2026-05-11T17:04:40.956518Z"
    }
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import pymc as pm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "25482473-1754-400c-82d9-5dc8fad070e8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:40.961177Z",
     "iopub.status.busy": "2026-05-11T17:04:40.960727Z",
     "iopub.status.idle": "2026-05-11T17:04:40.978624Z",
     "shell.execute_reply": "2026-05-11T17:04:40.978091Z"
    }
   },
   "outputs": [],
   "source": [
    "rng = np.random.default_rng(0)\n",
    "y_obs = rng.normal(1.0, 0.5, size=50)\n",
    "\n",
    "with pm.Model() as simple_model:\n",
    "    mu = pm.Normal(\"mu\")\n",
    "    sigma = pm.HalfNormal(\"sigma\")\n",
    "    pm.Normal(\"y\", mu=mu, sigma=sigma, observed=y_obs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a1a517da",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:40.981492Z",
     "iopub.status.busy": "2026-05-11T17:04:40.981081Z",
     "iopub.status.idle": "2026-05-11T17:04:50.565264Z",
     "shell.execute_reply": "2026-05-11T17:04:50.564595Z"
    }
   },
   "outputs": [],
   "source": [
    "with simple_model:\n",
    "    idata_numba = pm.sample(backend=\"numba\", quiet=True)\n",
    "    idata_jax = pm.sample(backend=\"jax\", quiet=True)\n",
    "    idata_c = pm.sample(backend=\"c\", quiet=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9adefe56-7e2e-492a-bc90-f0b45501e714",
   "metadata": {},
   "outputs": [],
   "source": [
    "All should arrive at the same place, if not at the same time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "99aa72ff-844c-41b3-979c-e7ab30121f2b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:50.568292Z",
     "iopub.status.busy": "2026-05-11T17:04:50.568073Z",
     "iopub.status.idle": "2026-05-11T17:04:50.624233Z",
     "shell.execute_reply": "2026-05-11T17:04:50.623665Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mu (numba)</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.067</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mu (jax)</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.066</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mu (c)</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.069</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           mean     sd eti89_lb eti89_ub\n",
       "mu (numba)  1.1  0.067     0.95      1.2\n",
       "mu (jax)    1.1  0.066     0.95      1.2\n",
       "mu (c)      1.1  0.069     0.95      1.2"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([\n",
    "    pm.stats.summary(idata_numba, var_names=[\"mu\"], kind=\"stats\").rename({\"mu\": \"mu (numba)\"}),\n",
    "    pm.stats.summary(idata_jax, var_names=[\"mu\"], kind=\"stats\").rename({\"mu\": \"mu (jax)\"}),\n",
    "    pm.stats.summary(idata_c, var_names=[\"mu\"], kind=\"stats\").rename({\"mu\": \"mu (c)\"}),\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b5c6bd6-1f7e-4d8e-9f98-9e36edb0e244",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Inference"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29c6fc0f",
   "metadata": {},
   "outputs": [],
   "source": [
    "### nutpie: the new default NUTS sampler\n",
    "\n",
    "PyMC has historically latched onto Stan's implementation of NUTS,\n",
    "borrowing its code almost line-by-line.\n",
    "It was time to contribute something back, and that something is\n",
    "[**nutpie**](https://pymc-devs.github.io/nutpie/).\n",
    "\n",
    "You can easily install pymc and nutpie together as:\n",
    "\n",
    "```bash\n",
    "pip install pymc[nutpie]\n",
    "```\n",
    "\n",
    "Once installed, nutpie becomes the **default** NUTS in PyMC.\n",
    "Combined with the Numba backend, end-to-end sampling is roughly **2x faster** than the\n",
    "old PyMC + C baseline on typical benchmarks, like the radon hierarchical model.\n",
    "Low-rank adaptation can push it to **4x**.\n",
    "For the details, see [Seyboldt, Carlson, & Carpenter (2026)](https://arxiv.org/abs/2603.18845).\n",
    "\n",
    "Three things set it apart from what we had before:\n",
    "\n",
    "- **Faster diagonal adaptation.** nutpie often gets away with ~400 tuning draws where the old default\n",
    "  conservatively used 1000, and after tuning it takes fewer leapfrog steps per draw.\n",
    "- **Low-rank mass-matrix adaptation.** For posteriors with strongly correlated parameters, the\n",
    "  diagonal mass matrix used by most HMC implementations is a poor fit. nutpie offers a low-rank\n",
    "  extension of the diagonal that can dramatically reduce the number of gradient evaluations per\n",
    "  effective draw on funnels, regressions with collinear predictors, and similar shapes.\n",
    "- **Native integration with Numba and JAX backends.** No Python overhead.\n",
    "\n",
    "An experimental **normalizing-flow adaptation** is also available for really difficult\n",
    "posteriors; see the [nutpie docs](https://pymc-devs.github.io/nutpie/nf-adapt.html).\n",
    "\n",
    "If installed, nutpie (with diagonal adaptation) is selected automatically.\n",
    "You can manually switch between implementations using the `nuts_sampler` argument of `pm.sample`.\n",
    "Other options include `\"pymc\"`, `\"numpyro\"`, and `\"blackjax\"`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2624281e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:50.627021Z",
     "iopub.status.busy": "2026-05-11T17:04:50.626767Z",
     "iopub.status.idle": "2026-05-11T17:04:52.189495Z",
     "shell.execute_reply": "2026-05-11T17:04:52.188947Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "NUTS[nutpie]: [mu, sigma]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "36ef39456c0f43c89ac945c1eae0e0d1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with simple_model:\n",
    "    idata_nutpie = pm.sample(nuts_sampler=\"nutpie\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4a217674-4492-477c-b1a9-65bb5d98d69a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:52.209922Z",
     "iopub.status.busy": "2026-05-11T17:04:52.209747Z",
     "iopub.status.idle": "2026-05-11T17:04:52.264043Z",
     "shell.execute_reply": "2026-05-11T17:04:52.263507Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mu</th>\n",
       "      <td>1.06</td>\n",
       "      <td>0.069</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.2</td>\n",
       "      <td>3573</td>\n",
       "      <td>2733</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0012</td>\n",
       "      <td>0.00084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sigma</th>\n",
       "      <td>0.471</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.4</td>\n",
       "      <td>0.55</td>\n",
       "      <td>3791</td>\n",
       "      <td>2793</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00084</td>\n",
       "      <td>0.00067</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        mean     sd eti89_lb eti89_ub  ess_bulk  ess_tail r_hat mcse_mean  \\\n",
       "mu      1.06  0.069     0.95      1.2      3573      2733  1.00    0.0012   \n",
       "sigma  0.471   0.05      0.4     0.55      3791      2793  1.00   0.00084   \n",
       "\n",
       "       mcse_sd  \n",
       "mu     0.00084  \n",
       "sigma  0.00067  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pm.stats.summary(idata_nutpie)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c363a1f",
   "metadata": {},
   "outputs": [],
   "source": [
    "A classic nasty posterior: a simple linear regression where the covariate sits far from zero,\n",
    "making the intercept and slope strongly anti-correlated in the likelihood."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "01857569",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:52.266710Z",
     "iopub.status.busy": "2026-05-11T17:04:52.266359Z",
     "iopub.status.idle": "2026-05-11T17:04:52.278938Z",
     "shell.execute_reply": "2026-05-11T17:04:52.278426Z"
    }
   },
   "outputs": [],
   "source": [
    "rng_corr = np.random.default_rng(1)\n",
    "N = 200\n",
    "x_far = rng_corr.normal(1000.0, 1.0, size=N)\n",
    "y_far = 2.0 + 3.0 * x_far + rng_corr.normal(0.0, 1.0, size=N)\n",
    "\n",
    "with pm.Model() as regression_far:\n",
    "    a = pm.Normal(\"a\", 0.0, 10)\n",
    "    b = pm.Normal(\"b\", 0.0, 10.0)\n",
    "    sigma = pm.HalfNormal(\"sigma\", 1.0)\n",
    "    pm.Normal(\"y\", mu=a + b * x_far, sigma=sigma, observed=y_far)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "89f3a535-3def-4e17-b8eb-963989197bec",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:52.281383Z",
     "iopub.status.busy": "2026-05-11T17:04:52.281059Z",
     "iopub.status.idle": "2026-05-11T17:04:54.663700Z",
     "shell.execute_reply": "2026-05-11T17:04:54.663059Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>-1</td>\n",
       "      <td>10.7</td>\n",
       "      <td>-17</td>\n",
       "      <td>17</td>\n",
       "      <td>305</td>\n",
       "      <td>371</td>\n",
       "      <td>1.01</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>3.002</td>\n",
       "      <td>0.0107</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>305</td>\n",
       "      <td>369</td>\n",
       "      <td>1.01</td>\n",
       "      <td>0.00061</td>\n",
       "      <td>0.00041</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean      sd eti89_lb eti89_ub  ess_bulk  ess_tail r_hat mcse_mean  \\\n",
       "a     -1    10.7      -17       17       305       371  1.01      0.61   \n",
       "b  3.002  0.0107        3        3       305       369  1.01   0.00061   \n",
       "\n",
       "   mcse_sd  \n",
       "a     0.41  \n",
       "b  0.00041  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with regression_far:\n",
    "    idata_diag = pm.sample(nuts={\"adaptation\": \"diag\"}, quiet=True)\n",
    "\n",
    "pm.stats.summary(idata_diag, var_names=[\"a\", \"b\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bbe0fd82-920e-4c67-9a79-04a486d03f2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "The low rank adaptation fares much better than the default diagonal mass matrix adaptation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "baae6bc9-554d-4bc6-912d-4216cb7df389",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:54.667838Z",
     "iopub.status.busy": "2026-05-11T17:04:54.667266Z",
     "iopub.status.idle": "2026-05-11T17:04:56.420487Z",
     "shell.execute_reply": "2026-05-11T17:04:56.419974Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>-0.6</td>\n",
       "      <td>9.8</td>\n",
       "      <td>-16</td>\n",
       "      <td>15</td>\n",
       "      <td>6487</td>\n",
       "      <td>3271</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.084</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>3.0025</td>\n",
       "      <td>0.0098</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>6482</td>\n",
       "      <td>3142</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00012</td>\n",
       "      <td>8.4e-05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     mean      sd eti89_lb eti89_ub  ess_bulk  ess_tail r_hat mcse_mean  \\\n",
       "a    -0.6     9.8      -16       15      6487      3271  1.00      0.12   \n",
       "b  3.0025  0.0098        3        3      6482      3142  1.00   0.00012   \n",
       "\n",
       "   mcse_sd  \n",
       "a    0.084  \n",
       "b  8.4e-05  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with regression_far:\n",
    "    idata_lr = pm.sample(nuts={\"adaptation\": \"low_rank\"}, quiet=True)\n",
    "\n",
    "pm.stats.summary(idata_lr, var_names=[\"a\", \"b\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "337036a2",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Variational inference (pymc-extras): Pathfinder & DADVI\n",
    "\n",
    "pymc-extras includes two modern Variational Inference solutions.\n",
    "\n",
    "**Pathfinder** Pathfinder is a parallel quasi-Newton variational approximation:\n",
    "it runs L-BFGS from multiple starting points and combines the resulting trajectories using Pareto-smoothed importance sampling.\n",
    "It's fast, parallelizable, and useful both as a NUTS warm-start and as a standalone approximation. See [Zhang et al. (2022)](https://arxiv.org/abs/2108.03782)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b7b8af99",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:56.423383Z",
     "iopub.status.busy": "2026-05-11T17:04:56.422971Z",
     "iopub.status.idle": "2026-05-11T17:04:59.691304Z",
     "shell.execute_reply": "2026-05-11T17:04:59.690737Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c76f1946733b418e9596e36588d01258",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pymc_extras import fit_pathfinder\n",
    "\n",
    "with simple_model:\n",
    "    idata_pf = fit_pathfinder(display_summary=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7fdadabf-8f9d-4b92-847a-0df054118139",
   "metadata": {},
   "outputs": [],
   "source": [
    "**DADVI** — *Deterministic* ADVI.\n",
    "Instead of stochastic optimization of the ELBO, it draws a fixed Monte Carlo sample upfront\n",
    "and hands the resulting deterministic objective to a second-order optimizer.\n",
    "Convergence is unambiguous (no more \"is the ELBO done bouncing around?\"),\n",
    "it works directly with off-the-shelf optimizers like L-BFGS, and linear-response covariance corrections\n",
    "fix the variance underestimation that mean-field ADVI is famous for. See [Giordano, Ingram & Broderick (2024)](https://jmlr.org/papers/volume25/23-1015/23-1015.pdf)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e4d5fea0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:04:59.728691Z",
     "iopub.status.busy": "2026-05-11T17:04:59.728486Z",
     "iopub.status.idle": "2026-05-11T17:05:09.466850Z",
     "shell.execute_reply": "2026-05-11T17:05:09.466108Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e01e3874a243404aa99a0d7fefcb75d0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pymc_extras.inference import fit_dadvi\n",
    "\n",
    "with simple_model:\n",
    "    idata_dadvi = fit_dadvi()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8a598630-2361-4686-92d8-bbc2672d31da",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:09.571802Z",
     "iopub.status.busy": "2026-05-11T17:05:09.571565Z",
     "iopub.status.idle": "2026-05-11T17:05:09.610534Z",
     "shell.execute_reply": "2026-05-11T17:05:09.610011Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>mu (pathfinder)</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.069</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mu (dadvi)</th>\n",
       "      <td>1.1</td>\n",
       "      <td>0.087</td>\n",
       "      <td>0.94</td>\n",
       "      <td>1.2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                mean     sd eti89_lb eti89_ub\n",
       "mu (pathfinder)  1.1  0.069     0.95      1.2\n",
       "mu (dadvi)       1.1  0.087     0.94      1.2"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat([\n",
    "    pm.stats.summary(idata_pf, var_names=[\"mu\"], kind=\"stats\").rename({\"mu\": \"mu (pathfinder)\"}),\n",
    "    pm.stats.summary(idata_dadvi, var_names=[\"mu\"], kind=\"stats\").rename({\"mu\": \"mu (dadvi)\"}),\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19739d9a-8ee6-4578-9241-742207ea008e",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Modelling\n",
    "\n",
    "A brief tour of four modelling additions: named-dimension models with `pymc.dims`,\n",
    "automatic marginalization of discrete latents, the state-space API, and PyMC-BART."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c11afe36",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Named dims with `pymc.dims`\n",
    "\n",
    "PyMC's `dims=` argument has existed since 2020 to label the axes of model variables.\n",
    "The new `pymc.dims` module (backed by PyTensor's XTensor) lets you write **entire models**\n",
    "against named dimensions, with automatic broadcasting, no more `[:, None, :]` gymnastics,\n",
    "and no more guessing which axis is which.\n",
    "\n",
    "A compact example (adapted from the [pymc.dims core\n",
    "notebook](https://www.pymc.io/projects/docs/en/latest/learn/core_notebooks/dims_module.html))."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "a2274d4a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:09.613387Z",
     "iopub.status.busy": "2026-05-11T17:05:09.613048Z",
     "iopub.status.idle": "2026-05-11T17:05:09.674243Z",
     "shell.execute_reply": "2026-05-11T17:05:09.673669Z"
    }
   },
   "outputs": [],
   "source": [
    "import pymc.dims as pmd\n",
    "\n",
    "coords = {\n",
    "    \"participant\": range(5),\n",
    "    \"trial\": range(20),\n",
    "    \"item\": range(3),\n",
    "}\n",
    "observed_response_np = np.ones((5, 20), dtype=int)\n",
    "\n",
    "with pm.Model(coords=coords) as dmodel:\n",
    "    observed_response = pmd.Data(\n",
    "        \"observed_response\",\n",
    "        observed_response_np,\n",
    "        dims=(\"participant\", \"trial\"),\n",
    "    )\n",
    "\n",
    "    participant_preference = pmd.ZeroSumNormal(\n",
    "        \"participant_preference\",\n",
    "        core_dims=\"item\",\n",
    "        dims=(\"participant\", \"item\"),\n",
    "    )\n",
    "    time_effects = pmd.Normal(\"time_effects\", dims=(\"item\", \"trial\"))\n",
    "\n",
    "    trial_preference = pmd.Deterministic(\n",
    "        \"trial_preference\",\n",
    "        participant_preference + time_effects,\n",
    "    )\n",
    "\n",
    "    response = pmd.Categorical(\n",
    "        \"response\",\n",
    "        p=pmd.math.softmax(trial_preference, dim=\"item\"),\n",
    "        core_dims=\"item\",\n",
    "        observed=observed_response,\n",
    "    )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d470235f",
   "metadata": {},
   "outputs": [],
   "source": [
    "Note the absence of `None`-indexing to create new axes, right-aligning conventions or numerical `axis` arguments.\n",
    "\n",
    "Also new: the `Model.table()` method renders a summary of every variable together\n",
    "with its expression and dimensions, a much more compact view than the good old `Model.to_graphviz()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6d6ab7aa",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:09.677105Z",
     "iopub.status.busy": "2026-05-11T17:05:09.676565Z",
     "iopub.status.idle": "2026-05-11T17:05:09.753982Z",
     "shell.execute_reply": "2026-05-11T17:05:09.753380Z"
    },
    "lines_to_next_cell": 2
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">                 Variable </span> <span style=\"font-weight: bold\">Expression                              </span> <span style=\"font-weight: bold\">Dimensions                           </span>\n",
       "─────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
       "      observed_response<span style=\"font-weight: bold\"> =</span>  Data                                     participant[5] × trial[20]           \n",
       "                                                                                                         \n",
       " participant_preference<span style=\"font-weight: bold\"> ~</span>  ZeroSumNormal(&lt;constant&gt;, &lt;constant&gt;)    participant[5] × item[3]             \n",
       "           time_effects<span style=\"font-weight: bold\"> ~</span>  Normal(0, 1)                             item[3] × trial[20]                  \n",
       "                                                                    <span style=\"font-style: italic\">Parameter count = 75</span>                 \n",
       "                                                                                                         \n",
       "       trial_preference<span style=\"font-weight: bold\"> =</span>  f(time_effects, participant_preference)  participant[5] × item[3] × trial[20] \n",
       "                                                                                                         \n",
       "               response<span style=\"font-weight: bold\"> ~</span>  Categorical(f(trial_preference))         participant[5] × trial[20]           \n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m \u001b[0m\u001b[1m                Variable\u001b[0m\u001b[1m \u001b[0m \u001b[1mExpression                             \u001b[0m\u001b[1m \u001b[0m \u001b[1mDimensions                          \u001b[0m\u001b[1m \u001b[0m\n",
       "─────────────────────────────────────────────────────────────────────────────────────────────────────────\n",
       "      observed_response\u001b[1m =\u001b[0m  Data                                     participant[5] × trial[20]           \n",
       "                                                                                                         \n",
       " participant_preference\u001b[1m ~\u001b[0m  ZeroSumNormal(<constant>, <constant>)    participant[5] × item[3]             \n",
       "           time_effects\u001b[1m ~\u001b[0m  Normal(0, 1)                             item[3] × trial[20]                  \n",
       "                                                                    \u001b[3mParameter count = 75\u001b[0m                 \n",
       "                                                                                                         \n",
       "       trial_preference\u001b[1m =\u001b[0m  f(time_effects, participant_preference)  participant[5] × item[3] × trial[20] \n",
       "                                                                                                         \n",
       "               response\u001b[1m ~\u001b[0m  Categorical(f(trial_preference))         participant[5] × trial[20]           \n"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dmodel.table()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3d76f7e1",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Automatic marginalization (pymc-extras)\n",
    "\n",
    "Discrete latent variables are a pain to sample with gradient-based MCMC.\n",
    "Traditionally you'd have to sum them out by hand and rewrite your model as a marginal likelihood.\n",
    "`pymc_extras.marginalize` automates that process."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "c690e2a8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:09.756946Z",
     "iopub.status.busy": "2026-05-11T17:05:09.756583Z",
     "iopub.status.idle": "2026-05-11T17:05:09.781430Z",
     "shell.execute_reply": "2026-05-11T17:05:09.780811Z"
    }
   },
   "outputs": [],
   "source": [
    "from pymc_extras import marginalize, recover_marginals\n",
    "\n",
    "rng_mix = np.random.default_rng(2)\n",
    "y_mix = rng_mix.normal(loc=np.tile([-2, 2], reps=50))\n",
    "mix_coords = {\"component\": [\"left\", \"right\"], \"obs\": range(len(y_mix))}\n",
    "\n",
    "with pm.Model(coords=mix_coords) as mix_model:\n",
    "    w = pm.Dirichlet(\"w\", a=np.ones(2))\n",
    "    mu = pm.Normal(\"mu\", mu=[-3.0, 3.0], dims=\"component\")\n",
    "    sigma = pm.HalfNormal(\"sigma\", dims=\"component\")\n",
    "    z = pm.Categorical(\"z\", p=w, dims=\"obs\")\n",
    "    pm.Normal(\"y\", mu=mu[z], sigma=sigma[z], observed=y_mix, dims=\"obs\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "35b59278-e07d-4f84-b60e-62eca6798083",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:09.784114Z",
     "iopub.status.busy": "2026-05-11T17:05:09.783912Z",
     "iopub.status.idle": "2026-05-11T17:05:17.459647Z",
     "shell.execute_reply": "2026-05-11T17:05:17.458924Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>w[0]</th>\n",
       "      <td>0.496</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.41</td>\n",
       "      <td>0.58</td>\n",
       "      <td>2474</td>\n",
       "      <td>2371</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0011</td>\n",
       "      <td>0.00081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>w[1]</th>\n",
       "      <td>0.504</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.42</td>\n",
       "      <td>0.59</td>\n",
       "      <td>2474</td>\n",
       "      <td>2371</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0011</td>\n",
       "      <td>0.00081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mu[left]</th>\n",
       "      <td>-1.91</td>\n",
       "      <td>0.162</td>\n",
       "      <td>-2.2</td>\n",
       "      <td>-1.6</td>\n",
       "      <td>2266</td>\n",
       "      <td>2577</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0034</td>\n",
       "      <td>0.0029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mu[right]</th>\n",
       "      <td>1.874</td>\n",
       "      <td>0.179</td>\n",
       "      <td>1.6</td>\n",
       "      <td>2.1</td>\n",
       "      <td>1973</td>\n",
       "      <td>1993</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0041</td>\n",
       "      <td>0.0035</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            mean     sd eti89_lb eti89_ub  ess_bulk  ess_tail r_hat mcse_mean  \\\n",
       "w[0]       0.496  0.055     0.41     0.58      2474      2371  1.00    0.0011   \n",
       "w[1]       0.504  0.055     0.42     0.59      2474      2371  1.00    0.0011   \n",
       "mu[left]   -1.91  0.162     -2.2     -1.6      2266      2577  1.00    0.0034   \n",
       "mu[right]  1.874  0.179      1.6      2.1      1973      1993  1.00    0.0041   \n",
       "\n",
       "           mcse_sd  \n",
       "w[0]       0.00081  \n",
       "w[1]       0.00081  \n",
       "mu[left]    0.0029  \n",
       "mu[right]   0.0035  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with marginalize(mix_model, [\"z\"]) as marginal_model:\n",
    "    idata_mix = pm.sample(quiet=True)\n",
    "\n",
    "pm.stats.summary(idata_mix, var_names=[\"w\", \"mu\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a46d1539",
   "metadata": {},
   "outputs": [],
   "source": [
    "After sampling the marginalized model, you can still **recover draws for the discrete\n",
    "latents** with `recover_marginals`. The component assignments for each observation come back\n",
    "as proper posterior estimates."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "833aa1f9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:17.463106Z",
     "iopub.status.busy": "2026-05-11T17:05:17.462498Z",
     "iopub.status.idle": "2026-05-11T17:05:18.820752Z",
     "shell.execute_reply": "2026-05-11T17:05:18.820135Z"
    }
   },
   "outputs": [],
   "source": [
    "with marginal_model:\n",
    "    idata_mix = recover_marginals(idata_mix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "2dc2dd9e-1c06-4a4c-b871-d16d70afbec5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:18.823671Z",
     "iopub.status.busy": "2026-05-11T17:05:18.823446Z",
     "iopub.status.idle": "2026-05-11T17:05:18.850708Z",
     "shell.execute_reply": "2026-05-11T17:05:18.850098Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>z[0]</th>\n",
       "      <td>0.0047</td>\n",
       "      <td>0.069</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>z[1]</th>\n",
       "      <td>1</td>\n",
       "      <td>0.057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>z[2]</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>z[3]</th>\n",
       "      <td>0.22</td>\n",
       "      <td>0.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>z[4]</th>\n",
       "      <td>0.38</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>z[5]</th>\n",
       "      <td>1</td>\n",
       "      <td>0.016</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        mean     sd\n",
       "z[0]  0.0047  0.069\n",
       "z[1]       1  0.057\n",
       "z[2]       0      0\n",
       "z[3]    0.22   0.41\n",
       "z[4]    0.38   0.48\n",
       "z[5]       1  0.016"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pm.stats.summary(idata_mix.isel(obs=slice(None, 6)), var_names=[\"z\"], kind=\"stats\")[[\"mean\", \"sd\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c8a6030",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Statespace (pymc-extras)\n",
    "\n",
    "`pymc_extras.statespace` offers Bayesian statespace models with an increasingly rich high-level API:\n",
    "`BayesianETS`, `BayesianSARIMAX`, `BayesianVARMAX`, plus a `structural` package for\n",
    "composing custom models out of level, trend, seasonal, and cyclical components.\n",
    "\n",
    "Here we fit a Holt–Winters ETS model (level + trend + additive quarterly seasonality) to a\n",
    "synthetic series generated from the same innovation-form process. Missing entries are\n",
    "imputed by the Kalman filter, and an 8-quarter forecast comes for free."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "5a1bd4ed",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:18.853651Z",
     "iopub.status.busy": "2026-05-11T17:05:18.853249Z",
     "iopub.status.idle": "2026-05-11T17:05:18.902222Z",
     "shell.execute_reply": "2026-05-11T17:05:18.901626Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-style: italic\">                       Model Requirements                       </span>\n",
       "                                                                \n",
       " <span style=\"font-weight: bold\"> Variable         </span> <span style=\"font-weight: bold\"> Shape </span> <span style=\"font-weight: bold\"> Constraints   </span> <span style=\"font-weight: bold\">        Dimensions </span> \n",
       " ────────────────────────────────────────────────────────────── \n",
       "  initial_level      <span style=\"font-weight: bold\">()</span>                                   <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "  initial_trend      <span style=\"font-weight: bold\">()</span>                                   <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "  initial_seasonal   <span style=\"font-weight: bold\">(</span><span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">4</span>,<span style=\"font-weight: bold\">)</span>                    <span style=\"font-weight: bold\">(</span><span style=\"color: #008000; text-decoration-color: #008000\">'seasonal_lag'</span>,<span style=\"font-weight: bold\">)</span>  \n",
       "  alpha              <span style=\"font-weight: bold\">()</span>      <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">0</span> &lt; alpha &lt; <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1</span>                <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "  beta               <span style=\"font-weight: bold\">()</span>      <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">0</span> &lt; beta &lt; <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1</span>                 <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "  gamma              <span style=\"font-weight: bold\">()</span>      <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">0</span> &lt; gamma &lt; <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1</span>                <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "  sigma_state        <span style=\"font-weight: bold\">()</span>      Positive                     <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "  sigma_obs          <span style=\"font-weight: bold\">()</span>      Positive                     <span style=\"color: #800080; text-decoration-color: #800080; font-style: italic\">None</span>  \n",
       "                                                                \n",
       "<span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f; font-style: italic\"> These parameters should be assigned priors inside a PyMC model </span>\n",
       "<span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f; font-style: italic\">    block before calling the build_statespace_graph method.     </span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[3m                       Model Requirements                       \u001b[0m\n",
       "                                                                \n",
       " \u001b[1m \u001b[0m\u001b[1mVariable        \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mShape\u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mConstraints  \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1m       Dimensions\u001b[0m\u001b[1m \u001b[0m \n",
       " ────────────────────────────────────────────────────────────── \n",
       "  initial_level      \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m                                   \u001b[3;35mNone\u001b[0m  \n",
       "  initial_trend      \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m                                   \u001b[3;35mNone\u001b[0m  \n",
       "  initial_seasonal   \u001b[1m(\u001b[0m\u001b[1;36m4\u001b[0m,\u001b[1m)\u001b[0m                    \u001b[1m(\u001b[0m\u001b[32m'seasonal_lag'\u001b[0m,\u001b[1m)\u001b[0m  \n",
       "  alpha              \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m      \u001b[1;36m0\u001b[0m < alpha < \u001b[1;36m1\u001b[0m                \u001b[3;35mNone\u001b[0m  \n",
       "  beta               \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m      \u001b[1;36m0\u001b[0m < beta < \u001b[1;36m1\u001b[0m                 \u001b[3;35mNone\u001b[0m  \n",
       "  gamma              \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m      \u001b[1;36m0\u001b[0m < gamma < \u001b[1;36m1\u001b[0m                \u001b[3;35mNone\u001b[0m  \n",
       "  sigma_state        \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m      Positive                     \u001b[3;35mNone\u001b[0m  \n",
       "  sigma_obs          \u001b[1m(\u001b[0m\u001b[1m)\u001b[0m      Positive                     \u001b[3;35mNone\u001b[0m  \n",
       "                                                                \n",
       "\u001b[2;3m These parameters should be assigned priors inside a PyMC model \u001b[0m\n",
       "\u001b[2;3m    block before calling the build_statespace_graph method.     \u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pymc_extras.statespace.models import BayesianETS\n",
    "\n",
    "rng = np.random.default_rng(3)\n",
    "T, m = 60, 4\n",
    "sigma_obs_true = 0.8\n",
    "level, trend = 10.0, 0.05\n",
    "seasonal = 2.0 * np.sin(2 * np.pi * np.arange(m) / m)\n",
    "y_ts = np.array([\n",
    "    level + trend * i + seasonal[i % m] + rng.normal(0, sigma_obs_true)\n",
    "    for i in range(T)\n",
    "])\n",
    "\n",
    "dates = pd.date_range(\"2000-Q1\", periods=T, freq=\"QS\")\n",
    "y_df = pd.DataFrame({\"y\": y_ts}, index=dates)\n",
    "y_df.loc[\"2007-Q1\":\"2008-Q4\", \"y\"] = np.nan  # Kalman filter will impute\n",
    "\n",
    "ets = BayesianETS(\n",
    "    order=(\"A\", \"A\", \"A\"),\n",
    "    endog_names=[\"y\"],\n",
    "    seasonal_periods=m,\n",
    "    measurement_error=True,\n",
    "    stationary_initialization=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cb4e2cb3-0113-4525-9031-c7499601ac50",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:18.905986Z",
     "iopub.status.busy": "2026-05-11T17:05:18.905557Z",
     "iopub.status.idle": "2026-05-11T17:05:49.154923Z",
     "shell.execute_reply": "2026-05-11T17:05:49.154334Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ImputationWarning: Provided data contains missing values and will be automatically imputed as hidden states during Kalman filtering.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "NUTS[nutpie]: [initial_level, initial_trend, initial_seasonal, alpha, beta, gamma, sigma_obs]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5b61d0a703154515976e679e134de692",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with pm.Model(coords=ets.coords) as ets_model:\n",
    "    pm.Normal(\"initial_level\", mu=10, sigma=5)\n",
    "    pm.Normal(\"initial_trend\", mu=0, sigma=1)\n",
    "    pm.ZeroSumNormal(\"initial_seasonal\", sigma=2, dims=\"seasonal_lag\")\n",
    "\n",
    "    pm.Beta(\"alpha\", 1, 1)\n",
    "    pm.Beta(\"beta\", 1, 1)\n",
    "    pm.Beta(\"gamma\", 1, 1)\n",
    "\n",
    "    pm.Deterministic(\"sigma_state\", pm.math.constant(np.array(0.0)))\n",
    "    pm.HalfNormal(\"sigma_obs\", 1)\n",
    "\n",
    "    ets.build_statespace_graph(y_df)\n",
    "\n",
    "    idata_ets = pm.sample(backend=\"jax\", progressbar=\"combined\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eeb930da",
   "metadata": {},
   "outputs": [],
   "source": [
    "An 8-quarter forecast falls out of `ets.forecast`, and `ets.sample_conditional_posterior`\n",
    "fills in the imputed gap inside the training window. Structural decomposition and scenario\n",
    "analyses are similarly built in."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "4961b995",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:05:49.526949Z",
     "iopub.status.busy": "2026-05-11T17:05:49.526757Z",
     "iopub.status.idle": "2026-05-11T17:06:45.626533Z",
     "shell.execute_reply": "2026-05-11T17:06:45.625924Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ImputationWarning: Provided data contains missing values and will be automatically imputed as hidden states during Kalman filtering.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Sampling: [filtered_posterior, filtered_posterior_observed, predicted_posterior, predicted_posterior_observed, smoothed_posterior, smoothed_posterior_observed]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "ImputationWarning: Provided data contains missing values and will be automatically imputed as hidden states during Kalman filtering.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Sampling: [forecast_combined]\n"
     ]
    }
   ],
   "source": [
    "smoothed = ets.sample_conditional_posterior(idata_ets, progressbar=False)\n",
    "forecast = ets.forecast(idata_ets, start=y_df.index[-1], periods=8, progressbar=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3c86e336",
   "metadata": {},
   "outputs": [],
   "source": [
    "Plot the smoothed in-sample fit (orange) and the 8-quarter forecast (blue),\n",
    "with 89% credible bands."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "194adadd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:06:45.632867Z",
     "iopub.status.busy": "2026-05-11T17:06:45.632546Z",
     "iopub.status.idle": "2026-05-11T17:06:45.894784Z",
     "shell.execute_reply": "2026-05-11T17:06:45.894161Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sm = smoothed[\"smoothed_posterior_observed\"].sel(observed_state=\"y\")\n",
    "fc = forecast[\"forecast_observed\"].sel(observed_state=\"y\")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(12, 4))\n",
    "ax.plot(y_df.index, y_df[\"y\"], \"k.\", label=\"observed\")\n",
    "ax.plot(y_df.index, sm.mean((\"chain\", \"draw\")), \"C1\", label=\"smoothed (in-sample)\")\n",
    "ax.fill_between(\n",
    "    y_df.index,\n",
    "    sm.quantile(0.055, (\"chain\", \"draw\")),\n",
    "    sm.quantile(0.945, (\"chain\", \"draw\")),\n",
    "    color=\"C1\",\n",
    "    alpha=0.2,\n",
    ")\n",
    "ax.plot(fc.time, fc.mean((\"chain\", \"draw\")), \"C0\", label=\"forecast\")\n",
    "ax.fill_between(\n",
    "    fc.time,\n",
    "    fc.quantile(0.055, (\"chain\", \"draw\")),\n",
    "    fc.quantile(0.945, (\"chain\", \"draw\")),\n",
    "    color=\"C0\",\n",
    "    alpha=0.2,\n",
    ")\n",
    "ax.legend(loc=\"lower right\", ncol=3, frameon=False);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f201c68",
   "metadata": {},
   "outputs": [],
   "source": [
    "### PyMC-BART\n",
    "\n",
    "[PyMC-BART](https://www.pymc.io/projects/bart/) brings **Bayesian Additive Regression\n",
    "Trees** to PyMC: a non-parametric prior on functions, expressed as an ensemble of\n",
    "decision trees.\n",
    "\n",
    "Newer versions add plotting utilities (partial dependence, variable importance)\n",
    "and direct variable-importance integration with **Kulprit** (see next section).\n",
    "BART can then guide downstream variable selection on a parametric follow-up model.\n",
    "An experimental Rust tree sampler is also available."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "0dd08c4a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:06:45.897536Z",
     "iopub.status.busy": "2026-05-11T17:06:45.897062Z",
     "iopub.status.idle": "2026-05-11T17:07:41.052157Z",
     "shell.execute_reply": "2026-05-11T17:07:41.051343Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Multiprocess sampling (4 chains in 4 jobs)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "CompoundStep\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      ">PGBART: [mu]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      ">NUTS: [sigma]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "83379bfcc43749db94468c89878a08f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Sampling 4 chains for 1_000 tune and 1_000 draw iterations (4_000 + 4_000 draws total) took 53 seconds.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The rhat statistic is larger than 1.01 for some parameters. This indicates problems during sampling. See https://arxiv.org/abs/1903.08008 for details\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import pymc_bart as pmb\n",
    "\n",
    "rng_bart = np.random.default_rng(4)\n",
    "n = 200\n",
    "X_bart = pd.DataFrame(\n",
    "    rng_bart.normal(size=(n, 4)),\n",
    "    columns=[\"x0\", \"x1\", \"x2\", \"x3\"],\n",
    ")\n",
    "# Only X_0 and X_2 actually drive the response\n",
    "y_bart = np.sin(1.5 * X_bart[\"x0\"]) + 0.7 * X_bart[\"x2\"] ** 2 + rng_bart.normal(0, 0.3, n)\n",
    "\n",
    "with pm.Model() as bart_model:\n",
    "    mu = pmb.BART(\"mu\", X_bart, y_bart, m=50)\n",
    "    sigma = pm.HalfNormal(\"sigma\", 1.0)\n",
    "    pm.Normal(\"y\", mu=mu, sigma=sigma, observed=y_bart)\n",
    "    idata_bart = pm.sample(progressbar=\"combined\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b46f68ff-6ec9-4f12-876d-0973fe1bada7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:42.012568Z",
     "iopub.status.busy": "2026-05-11T17:07:42.012362Z",
     "iopub.status.idle": "2026-05-11T17:07:45.157100Z",
     "shell.execute_reply": "2026-05-11T17:07:45.156356Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Variable importance — columns 0 and 2 should account for all cumulative R2.\n",
    "vi = pmb.compute_variable_importance(idata_bart, bartrv=mu, X=X_bart)\n",
    "pmb.plot_variable_importance(vi);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4cbd9c93-d6b4-413f-9d7b-6244d0ea8f49",
   "metadata": {},
   "outputs": [],
   "source": [
    "## More building blocks\n",
    "\n",
    "Two tools for when PyMC's built-ins don't quite fit: `wrap_jax` plugs arbitrary\n",
    "JAX functions into a model, and automatic logp derivation builds new\n",
    "distributions from existing ones."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59a4f51b-28d5-4380-8ce0-9b8444616880",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Drop-in JAX with `pytensor.wrap_jax`\n",
    "\n",
    "`pytensor.wrap_jax` lets you call any JAX-jittable function directly inside a PyMC model.\n",
    "That means the whole JAX ecosystem (Equinox neural networks, Diffrax ODE solvers, Optax optimizers)\n",
    "composes directly with your priors and likelihoods. Train a Bayesian neural network.\n",
    "Put a prior on the parameters of an ODE. Mix and match."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a057e5a9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:45.159963Z",
     "iopub.status.busy": "2026-05-11T17:07:45.159758Z",
     "iopub.status.idle": "2026-05-11T17:07:48.875693Z",
     "shell.execute_reply": "2026-05-11T17:07:48.874966Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "NUTS[nutpie]: [a, b]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2b355d4a2c974b1894cd89e6e98d91bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>0.13</td>\n",
       "      <td>0.152</td>\n",
       "      <td>-0.11</td>\n",
       "      <td>0.38</td>\n",
       "      <td>2229</td>\n",
       "      <td>2124</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0032</td>\n",
       "      <td>0.0025</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>0.88</td>\n",
       "      <td>0.261</td>\n",
       "      <td>0.51</td>\n",
       "      <td>1.3</td>\n",
       "      <td>2159</td>\n",
       "      <td>2099</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.0058</td>\n",
       "      <td>0.0046</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean     sd eti89_lb eti89_ub  ess_bulk  ess_tail r_hat mcse_mean mcse_sd\n",
       "b  0.13  0.152    -0.11     0.38      2229      2124  1.00    0.0032  0.0025\n",
       "a  0.88  0.261     0.51      1.3      2159      2099  1.00    0.0058  0.0046"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import jax.numpy as jnp\n",
    "from pytensor import wrap_jax\n",
    "\n",
    "@wrap_jax\n",
    "def forward(x, a, b):\n",
    "    return jnp.tanh(a * x + b)\n",
    "\n",
    "rng = np.random.default_rng(5)\n",
    "x_wj = rng.normal(size=100)\n",
    "y_wj = np.tanh(0.8 * x_wj + 0.1) + rng.normal(0, 0.1, size=100)\n",
    "\n",
    "with pm.Model() as wrap_jax_model:\n",
    "    a = pm.Normal(\"a\")\n",
    "    b = pm.Normal(\"b\")\n",
    "    mu_wj = forward(x_wj, a, b)\n",
    "    pm.Normal(\"y\", mu=mu_wj, observed=y_wj)\n",
    "\n",
    "    idata_wj = pm.sample(backend=\"jax\", progressbar=\"combined\")\n",
    "\n",
    "pm.stats.summary(idata_wj)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "43378e1c-425f-4960-8354-4a5820ac9b4b",
   "metadata": {},
   "outputs": [],
   "source": [
    "Sampling still runs through nutpie. `backend=\"jax\"` just tells PyMC to compile the model through the JAX path,\n",
    "so the wrapped function stays native end-to-end."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c7084dc-d8a6-4585-a35c-e106b7f3b6c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Automatic probability & `CustomDist`\n",
    "\n",
    "Most PPLs let you define models by chaining random variables and derive the log-density automatically.\n",
    "What's less known is that PyMC can also derive log-densities for transformations of existing distributions:\n",
    "exponentiating, taking maxima, slicing. All without writing the math yourself.\n",
    "\n",
    "Here we build a log-Student-T as exp of a Student-T, and ask PyMC for its log-density at 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "d7d2fe13",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:48.950093Z",
     "iopub.status.busy": "2026-05-11T17:07:48.949914Z",
     "iopub.status.idle": "2026-05-11T17:07:49.004648Z",
     "shell.execute_reply": "2026-05-11T17:07:49.004026Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "logp(LogStudentT, 0.5): -0.5713\n"
     ]
    }
   ],
   "source": [
    "log_student_t = pm.math.exp(pm.StudentT.dist(nu=4, mu=0, sigma=1))\n",
    "\n",
    "print(f\"logp(LogStudentT, 0.5): {pm.logp(log_student_t, 0.5).eval():.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24eb3b10-931d-46c7-ae71-4af71faf56de",
   "metadata": {
    "lines_to_next_cell": 2
   },
   "outputs": [],
   "source": [
    "The way to wire this into a model is with `pm.CustomDist`, which turns any `dist`-building\n",
    "function into a proper PyMC distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "3df42ad9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:49.007468Z",
     "iopub.status.busy": "2026-05-11T17:07:49.007210Z",
     "iopub.status.idle": "2026-05-11T17:07:49.219681Z",
     "shell.execute_reply": "2026-05-11T17:07:49.218992Z"
    },
    "lines_to_next_cell": 2
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "logp(y): -0.5713\n"
     ]
    }
   ],
   "source": [
    "def log_student_t_dist(nu, mu, sigma, size):\n",
    "    return pm.math.exp(pm.StudentT.dist(nu=nu, mu=mu, sigma=sigma, size=size))\n",
    "\n",
    "\n",
    "with pm.Model() as m:\n",
    "    mu = pm.Normal(\"mu\")\n",
    "    sigma = pm.HalfNormal(\"sigma\")\n",
    "    y = pm.CustomDist(\"y\", 4.0, mu, sigma, dist=log_student_t_dist, observed=0.5)\n",
    "\n",
    "print(f\"logp(y): {m.point_logps(round_vals=4)['y']}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b084d348-a3c4-46f1-9dcc-b7f2e179cb1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Bayesian workflow"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "722e7327-d071-40e8-bc6f-b5fc709c58c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "### ArviZ 1.0\n",
    "\n",
    "PyMC 6.0 is compatible with the new **ArviZ 1.0**. Most of the changes are\n",
    "invisible, the object returned by `pm.sample` still resembles the good old\n",
    "`InferenceData`, and the usual `az.plot_*` and `az.summary` calls keep working.\n",
    "For a full rundown, see the [migration guide](https://python.arviz.org/en/stable/user_guide/migration_guide.html).",
    "\n",
    "A few changes worth knowing:\n",
    "\n",
    "- **InferenceData → xarray.DataTree.** Arbitrary nesting is now allowed and any\n",
    "  xarray-supported I/O format works. Existing netCDF/Zarr files keep loading.\n",
    "- **New plotting library.** arviz-plots supports three backends (Matplotlib,\n",
    "  Bokeh, Plotly) and returns a `PlotCollection` built on grammar-of-graphics\n",
    "  ideas, with cleaner faceting and aesthetic mappings.\n",
    "- **New plots.** The suite of built-in plots has grown notably around\n",
    "  posterior-predictive checks. Browse them in the\n",
    "  [example gallery](https://python.arviz.org/projects/plots/en/latest/gallery/index.html).\n",
    "- **New default credible interval** has changed from 0.94 to 0.89, to remind\n",
    "  everyone that no single number can rule us all. The default interval kind is\n",
    "  now equal-tailed (`\"eti\"`) rather than highest-density (`\"hdi\"`), and WAIC has\n",
    "  been removed in favor of PSIS-LOO-CV.\n",
    "\n",
    "Oh and `plot_trace` is now `plot_trace_dist`..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "d2ef5e2f-5223-4d27-9f8d-c50151933c31",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:49.222537Z",
     "iopub.status.busy": "2026-05-11T17:07:49.222241Z",
     "iopub.status.idle": "2026-05-11T17:07:49.892276Z",
     "shell.execute_reply": "2026-05-11T17:07:49.891545Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1280x401.804 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import arviz as az\n",
    "\n",
    "az.plot_trace_dist(idata_numba);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "063606f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "... although you may want to try `plot_rank_dist` instead."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "56a633e3-cdd3-4612-970f-ea7d590b712c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:49.896320Z",
     "iopub.status.busy": "2026-05-11T17:07:49.895829Z",
     "iopub.status.idle": "2026-05-11T17:07:53.875124Z",
     "shell.execute_reply": "2026-05-11T17:07:53.873069Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1280x401.804 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "az.plot_rank_dist(idata_numba);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "560e5ef4-8561-4ae2-ad1c-acd240e239cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "### PreliZ — prior elicitation\n",
    "\n",
    "[PreliZ](https://preliz.readthedocs.io/) helps with one of the trickiest parts of the Bayesian workflow: prior choice.\n",
    "\n",
    "There's a new [Distributions Gallery](https://preliz.readthedocs.io/en/latest/gallery_content.html) to help get familiar with every family PreliZ supports.\n",
    "\n",
    "Recent PreliZ work integrates PyMC distributions and pymc-extras `Prior` objects\n",
    "directly: `maxent`, matching, and plotting all accept them. `from_pymc` /\n",
    "`from_prior` convert them explicitly.\n",
    "\n",
    "`maxent` is one of the package's most useful tools: it picks the maximum-entropy distribution matching a target probability interval, a principled way to translate vague prior knowledge into a concrete prior. It also accepts any summary statistic (mean, variance, mode, …) as a fixed constraint:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "06d6a32a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:53.879446Z",
     "iopub.status.busy": "2026-05-11T17:07:53.878775Z",
     "iopub.status.idle": "2026-05-11T17:07:55.564169Z",
     "shell.execute_reply": "2026-05-11T17:07:55.563408Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import preliz as pz\n",
    "\n",
    "# \"Find a Gamma with mass 0.9 between 1 and 10 — and pin the mean to 4.\"\n",
    "(prior, _fig) = pz.maxent(pz.Gamma(), lower=1, upper=10, mass=0.90, fixed_stat=(\"mean\", 4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "9168a044-b38a-4d1a-aa6a-9b6bc3baff8a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:55.568151Z",
     "iopub.status.busy": "2026-05-11T17:07:55.567613Z",
     "iopub.status.idle": "2026-05-11T17:07:56.995114Z",
     "shell.execute_reply": "2026-05-11T17:07:56.994273Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>sd</th>\n",
       "      <th>eti89_lb</th>\n",
       "      <th>eti89_ub</th>\n",
       "      <th>ess_bulk</th>\n",
       "      <th>ess_tail</th>\n",
       "      <th>r_hat</th>\n",
       "      <th>mcse_mean</th>\n",
       "      <th>mcse_sd</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>x</th>\n",
       "      <td>4</td>\n",
       "      <td>2.6</td>\n",
       "      <td>0.86</td>\n",
       "      <td>9</td>\n",
       "      <td>1196</td>\n",
       "      <td>1209</td>\n",
       "      <td>1.01</td>\n",
       "      <td>0.067</td>\n",
       "      <td>0.066</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  mean   sd eti89_lb eti89_ub  ess_bulk  ess_tail r_hat mcse_mean mcse_sd\n",
       "x    4  2.6     0.86        9      1196      1209  1.01     0.067   0.066"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with pm.Model() as m:\n",
    "    x = prior.to_pymc(\"x\")\n",
    "    idata = pm.sample(quiet=True)\n",
    "pm.stats.summary(idata)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff9836c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Kulprit — variable selection\n",
    "\n",
    "[Kulprit](https://kulprit.readthedocs.io/) tackles a problem that shows up in almost\n",
    "every real study: you fit your best model with all the predictors you measured, but\n",
    "for interpretation or deployment you want a smaller one. Which predictors do you\n",
    "actually need?\n",
    "\n",
    "Given a **Bambi** model with all candidates, Kulprit projects the full posterior onto\n",
    "smaller submodels using [projective predictive inference](https://kulprit.readthedocs.io/en/latest/examples/Overview.html).\n",
    "The projection keeps the uncertainty calibration of the original, and scales better\n",
    "than LOO-based comparison when the candidate count is large."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "1aa306ff",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-05-11T17:07:56.999142Z",
     "iopub.status.busy": "2026-05-11T17:07:56.998565Z",
     "iopub.status.idle": "2026-05-11T17:08:07.863090Z",
     "shell.execute_reply": "2026-05-11T17:08:07.862170Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selected: ['Intercept', 'x0', 'x2']\n"
     ]
    }
   ],
   "source": [
    "import bambi\n",
    "import kulprit\n",
    "\n",
    "# 4 candidate predictors; only x0 and x2 actually drive the response.\n",
    "rng_k = np.random.default_rng(6)\n",
    "n = 200\n",
    "df_k = pd.DataFrame(rng_k.normal(size=(n, 4)), columns=[\"x0\", \"x1\", \"x2\", \"x3\"])\n",
    "df_k[\"y\"] = 0.5 + 1.5 * df_k[\"x0\"] + 0.8 * df_k[\"x2\"] + rng_k.normal(0, 0.3, n)\n",
    "\n",
    "full_model = bambi.Model(\"y ~ x0 + x1 + x2 + x3\", df_k)\n",
    "idata_full = full_model.fit(quiet=True)\n",
    "pm.compute_log_likelihood(idata_full, model=full_model.backend.model, progressbar=False)\n",
    "\n",
    "ppi = kulprit.ProjectionPredictive(full_model, idata_full)\n",
    "ppi.project()\n",
    "print(f\"Selected: {ppi.select()}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "713efbf9-dbc1-4161-bac6-eb869436fc33",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Thanks\n",
    "\n",
    "Almost none of what we covered above came from a single person. Since the v5 and v2\n",
    "releases of PyMC and PyTensor, each project has seen contributions from around 150\n",
    "different individuals: bug fixes, documentation, new features, entire new packages.\n",
    "The surrounding libraries (ArviZ, nutpie, PyMC-BART, pymc-extras, PreliZ, Kulprit,\n",
    "Bambi) are maintained by a similarly broad and overlapping community.\n",
    "\n",
    "Whether you'd like to contribute or just try things out, the [PyMC Discourse](https://discourse.pymc.io/)\n",
    "is the easiest place to ask questions, share feedback, and discuss ideas. The GitHub repos\n",
    "are open for bug reports and pull requests. See you there.\n",
    "\n",
    ":::{image} star_rating.png\n",
    ":alt: star rating\n",
    ":width: 700px\n",
    ":::\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4fa78c53-cd3d-4dad-81f3-a1a3e9766745",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "jupytext": {
   "formats": "py:percent,ipynb"
  },
  "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": "3.13.9"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "180f2b026d6e4a89a579d3cf0ab7af07": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "2.0.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "2.0.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border_bottom": null,
       "border_left": null,
       "border_right": null,
       "border_top": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "2b355d4a2c974b1894cd89e6e98d91bf": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/output",
       "_model_module_version": "1.0.0",
       "_model_name": "OutputModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/output",
       "_view_module_version": "1.0.0",
       "_view_name": "OutputView",
       "layout": "IPY_MODEL_c24f0304bbba4bada5c7de28614d6bcb",
       "msg_id": "",
       "outputs": [
        {
         "data": {
          "text/html": "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">                                                                                                                   \n <span style=\"font-weight: bold\"> Progress                       </span> <span style=\"font-weight: bold\"> Draw           </span> <span style=\"font-weight: bold\"> Speed                         </span> <span style=\"font-weight: bold\"> Elapsed       </span> <span style=\"font-weight: bold\"> Remaining     </span> \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━</span>   5600             1806.58 draws/s                 0:00:03         0:00:00        \n                                                                                                                   \n</pre>\n",
          "text/plain": "                                                                                                                   \n \u001b[1m \u001b[0m\u001b[1mProgress                      \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mDraw          \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mSpeed                        \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mElapsed      \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mRemaining    \u001b[0m\u001b[1m \u001b[0m \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m   5600             1806.58 draws/s                 0:00:03         0:00:00        \n                                                                                                                   \n"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ],
       "tabbable": null,
       "tooltip": null
      }
     },
     "36ef39456c0f43c89ac945c1eae0e0d1": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/output",
       "_model_module_version": "1.0.0",
       "_model_name": "OutputModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/output",
       "_view_module_version": "1.0.0",
       "_view_name": "OutputView",
       "layout": "IPY_MODEL_180f2b026d6e4a89a579d3cf0ab7af07",
       "msg_id": "",
       "outputs": [
        {
         "data": {
          "text/html": "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">                                                                                                                   \n <span style=\"font-weight: bold\">                      </span> <span style=\"font-weight: bold\">           </span> <span style=\"font-weight: bold\">           </span> <span style=\"font-weight: bold\">           </span> <span style=\"font-weight: bold\"> Grad      </span> <span style=\"font-weight: bold\">                    </span> <span style=\"font-weight: bold\">          </span> <span style=\"font-weight: bold\">          </span> \n <span style=\"font-weight: bold\"> Progress             </span> <span style=\"font-weight: bold\"> Draw      </span> <span style=\"font-weight: bold\"> Divergen… </span> <span style=\"font-weight: bold\"> Step size </span> <span style=\"font-weight: bold\"> evals     </span> <span style=\"font-weight: bold\"> Speed              </span> <span style=\"font-weight: bold\"> Elapsed  </span> <span style=\"font-weight: bold\"> Remaini… </span> \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━</span>   1400        0           1.206       3           0.00 draws/s         0:00:00    0:00:00   \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━</span>   1400        0           1.122       3           0.00 draws/s         0:00:00    0:00:00   \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━</span>   1400        0           1.192       3           0.00 draws/s         0:00:00    0:00:00   \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━</span>   1400        0           1.053       1           0.00 draws/s         0:00:00    0:00:00   \n                                                                                                                   \n</pre>\n",
          "text/plain": "                                                                                                                   \n \u001b[1m                      \u001b[0m \u001b[1m           \u001b[0m \u001b[1m           \u001b[0m \u001b[1m           \u001b[0m \u001b[1m \u001b[0m\u001b[1mGrad     \u001b[0m\u001b[1m \u001b[0m \u001b[1m                    \u001b[0m \u001b[1m          \u001b[0m \u001b[1m          \u001b[0m \n \u001b[1m \u001b[0m\u001b[1mProgress            \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mDraw     \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mDivergen…\u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mStep size\u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mevals    \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mSpeed             \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mElapsed \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mRemaini…\u001b[0m\u001b[1m \u001b[0m \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━\u001b[0m   1400        0           1.206       3           0.00 draws/s         0:00:00    0:00:00   \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━\u001b[0m   1400        0           1.122       3           0.00 draws/s         0:00:00    0:00:00   \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━\u001b[0m   1400        0           1.192       3           0.00 draws/s         0:00:00    0:00:00   \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━\u001b[0m   1400        0           1.053       1           0.00 draws/s         0:00:00    0:00:00   \n                                                                                                                   \n"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ],
       "tabbable": null,
       "tooltip": null
      }
     },
     "480d15930d6e4b82bf861b6e158975c3": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "2.0.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "2.0.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border_bottom": null,
       "border_left": null,
       "border_right": null,
       "border_top": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "59b36e4f6ebb4cdbb95d0a6a21d13dbe": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "2.0.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "2.0.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border_bottom": null,
       "border_left": null,
       "border_right": null,
       "border_top": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "5b61d0a703154515976e679e134de692": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/output",
       "_model_module_version": "1.0.0",
       "_model_name": "OutputModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/output",
       "_view_module_version": "1.0.0",
       "_view_name": "OutputView",
       "layout": "IPY_MODEL_480d15930d6e4b82bf861b6e158975c3",
       "msg_id": "",
       "outputs": [
        {
         "data": {
          "text/html": "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">                                                                                                                   \n <span style=\"font-weight: bold\"> Progress                       </span> <span style=\"font-weight: bold\"> Draw           </span> <span style=\"font-weight: bold\"> Speed                         </span> <span style=\"font-weight: bold\"> Elapsed       </span> <span style=\"font-weight: bold\"> Remaining     </span> \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━</span>   5600             313.20 draws/s                  0:00:17         0:00:00        \n                                                                                                                   \n</pre>\n",
          "text/plain": "                                                                                                                   \n \u001b[1m \u001b[0m\u001b[1mProgress                      \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mDraw          \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mSpeed                        \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mElapsed      \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mRemaining    \u001b[0m\u001b[1m \u001b[0m \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m   5600             313.20 draws/s                  0:00:17         0:00:00        \n                                                                                                                   \n"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ],
       "tabbable": null,
       "tooltip": null
      }
     },
     "6c3507a0a4a74adf90c7a00358340039": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "2.0.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "2.0.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border_bottom": null,
       "border_left": null,
       "border_right": null,
       "border_top": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "7ea77f1ce5b74d3c89ef166bcf4c45c3": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "2.0.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "2.0.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border_bottom": null,
       "border_left": null,
       "border_right": null,
       "border_top": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "83379bfcc43749db94468c89878a08f3": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/output",
       "_model_module_version": "1.0.0",
       "_model_name": "OutputModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/output",
       "_view_module_version": "1.0.0",
       "_view_name": "OutputView",
       "layout": "IPY_MODEL_59b36e4f6ebb4cdbb95d0a6a21d13dbe",
       "msg_id": "",
       "outputs": [
        {
         "data": {
          "text/html": "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">                                                                                                                   \n <span style=\"font-weight: bold\"> Progress                       </span> <span style=\"font-weight: bold\"> Draw           </span> <span style=\"font-weight: bold\"> Speed                         </span> <span style=\"font-weight: bold\"> Elapsed       </span> <span style=\"font-weight: bold\"> Remaining     </span> \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  <span style=\"color: #1f77b4; text-decoration-color: #1f77b4\">━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━</span>   7996             156.90 draws/s                  0:00:49         0:00:00        \n                                                                                                                   \n</pre>\n",
          "text/plain": "                                                                                                                   \n \u001b[1m \u001b[0m\u001b[1mProgress                      \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mDraw          \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mSpeed                        \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mElapsed      \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mRemaining    \u001b[0m\u001b[1m \u001b[0m \n ───────────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  \u001b[38;2;31;119;180m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m   7996             156.90 draws/s                  0:00:49         0:00:00        \n                                                                                                                   \n"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ],
       "tabbable": null,
       "tooltip": null
      }
     },
     "c24f0304bbba4bada5c7de28614d6bcb": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "2.0.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "2.0.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border_bottom": null,
       "border_left": null,
       "border_right": null,
       "border_top": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "c76f1946733b418e9596e36588d01258": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/output",
       "_model_module_version": "1.0.0",
       "_model_name": "OutputModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/output",
       "_view_module_version": "1.0.0",
       "_view_name": "OutputView",
       "layout": "IPY_MODEL_7ea77f1ce5b74d3c89ef166bcf4c45c3",
       "msg_id": "",
       "outputs": [
        {
         "data": {
          "text/html": "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">                                                                                                              \n <span style=\"font-weight: bold\"> Path    </span> <span style=\"font-weight: bold\"> Status     </span> <span style=\"font-weight: bold\"> Steps  </span> <span style=\"font-weight: bold\"> Steps/s  </span> <span style=\"font-weight: bold\"> Best step </span> <span style=\"font-weight: bold\"> Best ELBO    </span> <span style=\"font-weight: bold\"> Cur ELBO     </span> <span style=\"font-weight: bold\"> Step size  </span> <span style=\"font-weight: bold\"> Elapsed  </span> \n ──────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  Path 1    ok           7        275.8/s    4           -37.436        —              —            0:00:00   \n  Path 2    ok           15       163.3/s    14          -37.449        —              —            0:00:00   \n  Path 3    ok           9        114.2/s    6           -37.459        —              —            0:00:01   \n  Path 4    ok           7        204.1/s    6           -37.398        —              —            0:00:01   \n                                                                                                              \n</pre>\n",
          "text/plain": "                                                                                                              \n \u001b[1m \u001b[0m\u001b[1mPath   \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mStatus    \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mSteps \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mSteps/s \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mBest step\u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mBest ELBO   \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mCur ELBO    \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mStep size \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mElapsed \u001b[0m\u001b[1m \u001b[0m \n ──────────────────────────────────────────────────────────────────────────────────────────────────────────── \n  Path 1    ok           7        275.8/s    4           -37.436        —              —            0:00:00   \n  Path 2    ok           15       163.3/s    14          -37.449        —              —            0:00:00   \n  Path 3    ok           9        114.2/s    6           -37.459        —              —            0:00:01   \n  Path 4    ok           7        204.1/s    6           -37.398        —              —            0:00:01   \n                                                                                                              \n"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ],
       "tabbable": null,
       "tooltip": null
      }
     },
     "e01e3874a243404aa99a0d7fefcb75d0": {
      "model_module": "@jupyter-widgets/output",
      "model_module_version": "1.0.0",
      "model_name": "OutputModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/output",
       "_model_module_version": "1.0.0",
       "_model_name": "OutputModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/output",
       "_view_module_version": "1.0.0",
       "_view_name": "OutputView",
       "layout": "IPY_MODEL_6c3507a0a4a74adf90c7a00358340039",
       "msg_id": "",
       "outputs": [
        {
         "data": {
          "text/html": "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">                                                                                                    \n <span style=\"font-weight: bold\"> Minimizing                        </span> <span style=\"font-weight: bold\"> Elapsed </span> <span style=\"font-weight: bold\"> Iteration </span> <span style=\"font-weight: bold\"> Objective   </span> <span style=\"font-weight: bold\"> ||grad||   </span> <span style=\"font-weight: bold\"> gᵀHg/|g|²   </span> \n ────────────────────────────────────────────────────────────────────────────────────────────────── \n  <span style=\"color: #729c1f; text-decoration-color: #729c1f\">━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━</span>   <span style=\"color: #808000; text-decoration-color: #808000\">0:00:00</span>   <span style=\"color: #008000; text-decoration-color: #008000\">16/16    </span>   39.66329292   0.00000000   +8.6046e+01  \n                                                                                                    \n</pre>\n",
          "text/plain": "                                                                                                    \n \u001b[1m \u001b[0m\u001b[1mMinimizing                       \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mElapsed\u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mIteration\u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mObjective  \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1m||grad||  \u001b[0m\u001b[1m \u001b[0m \u001b[1m \u001b[0m\u001b[1mgᵀHg/|g|²  \u001b[0m\u001b[1m \u001b[0m \n ────────────────────────────────────────────────────────────────────────────────────────────────── \n  \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m   \u001b[33m0:00:00\u001b[0m   \u001b[32m16/16    \u001b[0m   39.66329292   0.00000000   +8.6046e+01  \n                                                                                                    \n"
         },
         "metadata": {},
         "output_type": "display_data"
        }
       ],
       "tabbable": null,
       "tooltip": null
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}