Add script: merged PR count per author aggregated by week over the past year

_doc/practice/years/2026/github_stat_pr.ipynb

@@ -0,0 +1,326 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Nombre de PR fusionnées par personne agrégées par semaine\n",
+    "\n",
+    "Ce notebook récupère, via l'API GitHub, le nombre de *pull requests* (PR) fusionnées\n",
+    "pour **un ou plusieurs dépôts**, les regroupe par auteur et par semaine sur l'année écoulée,\n",
+    "puis affiche le résultat sous forme de graphique.\n",
+    "\n",
+    "**Dépendances :** `requests`, `pandas`, `matplotlib`.\n",
+    "\n",
+    "**Token GitHub :** l'API GitHub limite les appels non authentifiés à 60 requêtes par heure.\n",
+    "Pour lever cette limite, définissez la variable d'environnement `GITHUB_TOKEN`\n",
+    "avec un *Personal Access Token* (PAT) GitHub :\n",
+    "\n",
+    "```bash\n",
+    "export GITHUB_TOKEN=ghp_xxxxxxxxxxxxxxxxxxxx\n",
+    "```\n",
+    "\n",
+    "Sans token, le notebook fonctionne mais peut être limité sur de grands dépôts."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "import datetime\n",
+    "import requests\n",
+    "import pandas as pd\n",
+    "import matplotlib.pyplot as plt\n",
+    "import matplotlib.dates as mdates"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Paramètres\n",
+    "\n",
+    "Modifiez `REPOS` pour lister les dépôts à analyser sous la forme\n",
+    "`[(owner, repo), ...]`. Vous pouvez ajouter autant de dépôts que vous le souhaitez."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "REPOS = [\n",
+    "    (\"sdpython\", \"teachpyx\"),\n",
+    "    # (\"sdpython\", \"onnx-extended\"),  # ajoutez d'autres dépôts ici\n",
+    "]\n",
+    "\n",
+    "# Jeton d'authentification GitHub (optionnel mais recommandé)\n",
+    "GITHUB_TOKEN = os.environ.get(\"GITHUB_TOKEN\", \"\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Récupération des PR fusionnées via l'API GitHub\n",
+    "\n",
+    "L'API REST GitHub expose le point d'accès `/repos/{owner}/{repo}/pulls`\n",
+    "avec `state=closed`. On filtre ensuite les PR dont le champ `merged_at` est renseigné\n",
+    "et dont la date de fusion est dans les 12 derniers mois.\n",
+    "\n",
+    "La pagination est gérée via le paramètre `page`.\n",
+    "La boucle principale itère sur chaque dépôt listé dans `REPOS`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def fetch_merged_prs(owner: str, repo: str, token: str = \"\") -> list[dict]:\n",
+    "    \"\"\"Récupère toutes les PR fusionnées au cours de l'année écoulée pour un dépôt.\n",
+    "\n",
+    "    :param owner: propriétaire du dépôt GitHub\n",
+    "    :param repo: nom du dépôt GitHub\n",
+    "    :param token: jeton d'authentification GitHub (optionnel)\n",
+    "    :return: liste de dictionnaires avec les champs ``author``, ``merged_at``, ``repo``\n",
+    "    \"\"\"\n",
+    "    headers = {\"Accept\": \"application/vnd.github+json\"}\n",
+    "    if token:\n",
+    "        headers[\"Authorization\"] = f\"Bearer {token}\"\n",
+    "\n",
+    "    since = datetime.datetime.now(datetime.timezone.utc) - datetime.timedelta(days=365)\n",
+    "\n",
+    "    results = []\n",
+    "    page = 1\n",
+    "    per_page = 100\n",
+    "\n",
+    "    while True:\n",
+    "        url = (\n",
+    "            f\"https://api.github.com/repos/{owner}/{repo}/pulls\"\n",
+    "            f\"?state=closed&per_page={per_page}&page={page}&sort=updated&direction=desc\"\n",
+    "        )\n",
+    "        response = requests.get(url, headers=headers, timeout=30)\n",
+    "        try:\n",
+    "            response.raise_for_status()\n",
+    "        except requests.HTTPError as exc:\n",
+    "            status = exc.response.status_code\n",
+    "            if status == 401:\n",
+    "                raise RuntimeError(\n",
+    "                    \"Authentification refusée (401). Vérifiez votre GITHUB_TOKEN.\"\n",
+    "                ) from exc\n",
+    "            if status == 403:\n",
+    "                raise RuntimeError(\n",
+    "                    \"Accès refusé (403). Vous avez peut-être atteint la limite de l'API \"\n",
+    "                    \"GitHub (60 requêtes/h sans token). Définissez GITHUB_TOKEN.\"\n",
+    "                ) from exc\n",
+    "            if status == 404:\n",
+    "                raise RuntimeError(\n",
+    "                    f\"Dépôt introuvable (404) : {owner}/{repo}. Vérifiez OWNER et REPO.\"\n",
+    "                ) from exc\n",
+    "            raise\n",
+    "        prs = response.json()\n",
+    "\n",
+    "        if not prs:\n",
+    "            break\n",
+    "\n",
+    "        stop = False\n",
+    "        for pr in prs:\n",
+    "            merged_at = pr.get(\"merged_at\")\n",
+    "            if not merged_at:\n",
+    "                continue\n",
+    "            merged_dt = datetime.datetime.fromisoformat(merged_at.replace(\"Z\", \"+00:00\"))\n",
+    "            if merged_dt < since:\n",
+    "                stop = True\n",
+    "                break\n",
+    "            author = (pr.get(\"user\") or {}).get(\"login\", \"unknown\")\n",
+    "            results.append({\"author\": author, \"merged_at\": merged_dt, \"repo\": f\"{owner}/{repo}\"})\n",
+    "\n",
+    "        if stop:\n",
+    "            break\n",
+    "\n",
+    "        page += 1\n",
+    "\n",
+    "    return results\n",
+    "\n",
+    "\n",
+    "merged_prs = []\n",
+    "for owner, repo in REPOS:\n",
+    "    prs = fetch_merged_prs(owner, repo, GITHUB_TOKEN)\n",
+    "    print(f\"  {owner}/{repo} : {len(prs)} PR(s) fusionnée(s)\")\n",
+    "    merged_prs.extend(prs)\n",
+    "\n",
+    "print(f\"Total : {len(merged_prs)} PR(s) fusionnée(s) sur l'ensemble des dépôts.\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Agrégation par auteur et par semaine"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = pd.DataFrame(merged_prs)\n",
+    "\n",
+    "if df.empty:\n",
+    "    print(\"Aucune donnée à afficher.\")\n",
+    "else:\n",
+    "    # Tronque la date au lundi de la semaine\n",
+    "    df[\"week\"] = df[\"merged_at\"].dt.to_period(\"W\").dt.start_time\n",
+    "\n",
+    "    weekly = (\n",
+    "        df.groupby([\"repo\", \"author\", \"week\"])\n",
+    "        .size()\n",
+    "        .reset_index(name=\"pr_count\")\n",
+    "    )\n",
+    "    print(weekly.head(10))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Tableau croisé (auteur × semaine, agrégé sur tous les dépôts)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "if not df.empty:\n",
+    "    # Agrégation sur tous les dépôts\n",
+    "    pivot = weekly.pivot_table(\n",
+    "        index=\"author\", columns=\"week\", values=\"pr_count\", aggfunc=\"sum\", fill_value=0\n",
+    "    )\n",
+    "    # Tri par nombre total de PR décroissant\n",
+    "    pivot = pivot.loc[pivot.sum(axis=1).sort_values(ascending=False).index]\n",
+    "    pivot"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Tableau croisé par dépôt"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "if not df.empty and len(REPOS) > 1:\n",
+    "    for repo_name, grp in weekly.groupby(\"repo\"):\n",
+    "        pvt = grp.pivot_table(\n",
+    "            index=\"author\", columns=\"week\", values=\"pr_count\", aggfunc=\"sum\", fill_value=0\n",
+    "        )\n",
+    "        pvt = pvt.loc[pvt.sum(axis=1).sort_values(ascending=False).index]\n",
+    "        print(f\"\\n=== {repo_name} ===\")\n",
+    "        display(pvt)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Visualisation : nombre de PR fusionnées par semaine (empilé par auteur)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "if not df.empty:\n",
+    "    fig, ax = plt.subplots(figsize=(14, 5))\n",
+    "\n",
+    "    stacked_height = None\n",
+    "    weeks = pivot.columns  # DatetimeIndex\n",
+    "    week_nums = mdates.date2num(weeks.to_pydatetime())\n",
+    "\n",
+    "    for author in pivot.index:\n",
+    "        values = pivot.loc[author].values\n",
+    "        if stacked_height is None:\n",
+    "            ax.bar(week_nums, values, width=5, label=author)\n",
+    "            stacked_height = values.copy()\n",
+    "        else:\n",
+    "            ax.bar(week_nums, values, width=5, bottom=stacked_height, label=author)\n",
+    "            stacked_height += values\n",
+    "\n",
+    "    ax.xaxis.set_major_formatter(mdates.DateFormatter(\"%Y-%m-%d\"))\n",
+    "    ax.xaxis.set_major_locator(mdates.WeekdayLocator(byweekday=mdates.MO, interval=4))\n",
+    "    plt.xticks(rotation=45, ha=\"right\")\n",
+    "    ax.set_xlabel(\"Semaine\")\n",
+    "    ax.set_ylabel(\"Nombre de PR fusionnées\")\n",
+    "    repos_label = \", \".join(f\"{o}/{r}\" for o, r in REPOS)\n    ax.set_title(f\"PR fusionnées par semaine — {repos_label}\")\n",
+    "    ax.legend(loc=\"upper left\", bbox_to_anchor=(1, 1), title=\"Auteur\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Visualisation : carte de chaleur (heatmap auteur × semaine)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "if not df.empty:\n",
+    "    fig, ax = plt.subplots(figsize=(14, max(3, len(pivot) * 0.5)))\n",
+    "\n",
+    "    im = ax.imshow(pivot.values, aspect=\"auto\", cmap=\"YlOrRd\")\n",
+    "    plt.colorbar(im, ax=ax, label=\"Nombre de PR\")\n",
+    "\n",
+    "    ax.set_yticks(range(len(pivot.index)))\n",
+    "    ax.set_yticklabels(pivot.index)\n",
+    "\n",
+    "    # Affiche une étiquette de semaine sur 4\n",
+    "    step = max(1, len(pivot.columns) // 12)\n",
+    "    ax.set_xticks(range(0, len(pivot.columns), step))\n",
+    "    ax.set_xticklabels(\n",
+    "        [str(d)[:10] for d in pivot.columns[::step]], rotation=45, ha=\"right\"\n",
+    "    )\n",
+    "\n",
+    "    repos_label = \", \".join(f\"{o}/{r}\" for o, r in REPOS)\n    ax.set_title(f\"Heatmap des PR fusionnées — {repos_label}\")\n",
+    "    ax.set_xlabel(\"Semaine\")\n",
+    "    ax.set_ylabel(\"Auteur\")\n",
+    "    plt.tight_layout()\n",
+    "    plt.show()"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.12.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

_doc/practice/years/2026/index.rst

@@ -8,3 +8,4 @@
     :caption: machine learning
 
     parcoursup_2026
+    github_stat_pr