From d944a1d48d3032491ceef0f5978fac1189f92cc5 Mon Sep 17 00:00:00 2001
From: Elizaveta Semenova <elizaveta.p.semenova@gmail.com>
Date: Wed, 20 Nov 2024 21:14:20 +0000
Subject: [PATCH] custom distribution with

---
 09_intro_to_Numpyro.ipynb | 33 +++++++++------------------------
 1 file changed, 9 insertions(+), 24 deletions(-)

diff --git a/09_intro_to_Numpyro.ipynb b/09_intro_to_Numpyro.ipynb
index eb64f5e..6f00b22 100644
--- a/09_intro_to_Numpyro.ipynb
+++ b/09_intro_to_Numpyro.ipynb
@@ -392,16 +392,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 76,
    "metadata": {},
    "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "sample: 100%|██████████| 1500/1500 [00:01<00:00, 1422.24it/s, 7 steps of size 7.31e-01. acc. prob=0.89]"
-     ]
-    },
     {
      "name": "stdout",
      "output_type": "stream",
@@ -413,13 +406,6 @@
       "\n",
       "Number of divergences: 0\n"
      ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "\n"
-     ]
     }
    ],
    "source": [
@@ -442,7 +428,7 @@
     "\n",
     "\n",
     "nuts_kernel = NUTS(model)\n",
-    "mcmc = MCMC(nuts_kernel, num_warmup=500, num_samples=1000)\n",
+    "mcmc = MCMC(nuts_kernel, num_warmup=500, num_samples=1000, progress_bar=False)\n",
     "mcmc.run(jax.random.PRNGKey(0), data = data)\n",
     "\n",
     "mcmc.print_summary()"
@@ -456,14 +442,13 @@
     "\n",
     "The typical elements that we will need to write are model in Numpyro are as follows:\n",
     "\n",
-    "- parameters sampled with  <font color='green'>`numpyro.sample`</font>\n",
-    "- parameters sampled from any of the available distributsions using, e.g.  <font color='green'>`dist.Beta(alpha, beta)`</font> \n",
-    "- likelihood constructed by adding `obs=...` to the sampling statement: <font color='green'>`numpyro.sample('obs', dist.Binomial(n, p), obs=h)`</font>\n",
-    "- the sampling algorithm which we would like to use. NUTS is a good default oprtion: <font color='green'>`kernel = NUTS(model)` </font>,\n",
-    "- number of warm-up steps, number of iterations, number of chains, e.g. <font color='green'>`MCMC(kernel, num_warmup=1000, num_samples=2000, num_chains=4)`</font>,\n",
-    "- using <font color='green'>`Predictive` </font> class we can generate predictions.\n",
-    "\n",
-    "\n"
+    "- sample parameters with  <font color='green'>`numpyro.sample`</font>,\n",
+    "- sample parameters from any of the built-in distributsions using, e.g.  <font color='green'>`dist.Beta(alpha, beta)`</font>,\n",
+    "- specify likelihood by adding `obs=...` to the sampling statement: <font color='green'>`numpyro.sample('obs', dist.Binomial(n, p), obs=h)`</font>,\n",
+    "- specify a sampling algorithm. NUTS is a good default option: <font color='green'>`kernel = NUTS(model)` </font>,\n",
+    "- specify number of warm-up steps, number of iterations, number of chains, e.g. <font color='green'>`MCMC(kernel, num_warmup=1000, num_samples=2000, num_chains=4)`</font>,\n",
+    "- use <font color='green'>`Predictive` </font> class we can generate predictions,\n",
+    "- use <font color='green'>`numpyro.factor` </font> to implement custom distributions."
    ]
   },
   {