Update Getting Started for PyMC v4.x (#2442) (#2444)

* Replaced outdated PyMC3 with PyMCV4.x workflow.

* Using sigma instead of sd to prevent DeprecationWarning.

* Removed az.from_pymc3() and integrated prior, posterior, and posterior predictive sampling directly using pm.sample() and .extend().

Signed-off-by: PiyushPanwarFST <[email protected]>

Author: PiyushPanwarFST

doc/source/getting_started/CreatingInferenceData.ipynb

@@ -4138,12 +4138,12 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## From PyMC3"
+    "## From PyMC"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "metadata": {
     "ExecuteTime": {
      "end_time": "2020-06-19T06:23:42.715900Z",
@@ -4152,7 +4152,7 @@
    },
    "outputs": [],
    "source": [
-    "import pymc3 as pm\n",
+    "import pymc as pm\n",
     "\n",
     "draws = 500\n",
     "chains = 2\n",
@@ -4166,7 +4166,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": null,
    "metadata": {
     "ExecuteTime": {
      "end_time": "2020-06-05T06:47:16.996019Z",
@@ -10202,24 +10202,19 @@
    ],
    "source": [
     "with pm.Model() as model:\n",
-    "    mu = pm.Normal(\"mu\", mu=0, sd=5)\n",
+    "    mu = pm.Normal(\"mu\", mu=0, sigma=5)\n",
     "    tau = pm.HalfCauchy(\"tau\", beta=5)\n",
-    "    theta_tilde = pm.Normal(\"theta_tilde\", mu=0, sd=1, shape=eight_school_data[\"J\"])\n",
+    "    theta_tilde = pm.Normal(\"theta_tilde\", mu=0, sigma=1, shape=eight_school_data[\"J\"])\n",
+    "    \n",
     "    theta = pm.Deterministic(\"theta\", mu + tau * theta_tilde)\n",
-    "    pm.Normal(\"obs\", mu=theta, sd=eight_school_data[\"sigma\"], observed=eight_school_data[\"y\"])\n",
+    "    \n",
+    "    pm.Normal(\"obs\", mu=theta, sigma=eight_school_data[\"sigma\"], observed=eight_school_data[\"y\"])\n",
     "\n",
-    "    trace = pm.sample(draws, chains=chains)\n",
-    "    prior = pm.sample_prior_predictive()\n",
-    "    posterior_predictive = pm.sample_posterior_predictive(trace)\n",
+    "    idata = pm.sample(draws=draws, chains=chains)\n",
+    "    idata.extend(pm.sample_prior_predictive())\n",
+    "    idata.extend(pm.sample_posterior_predictive(idata))\n",
     "\n",
-    "    pm_data = az.from_pymc3(\n",
-    "        trace=trace,\n",
-    "        prior=prior,\n",
-    "        posterior_predictive=posterior_predictive,\n",
-    "        coords={\"school\": np.arange(eight_school_data[\"J\"])},\n",
-    "        dims={\"theta\": [\"school\"], \"theta_tilde\": [\"school\"]},\n",
-    "    )\n",
-    "pm_data"
+    "idata"
    ]
   },
   {