{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "fe6fe538",
   "metadata": {},
   "source": [
    "# 22 — Le test de forme, et ce qu'il ne peut pas séparer\n",
    "\n",
    "Le [notebook 21](21_angles_morts.ipynb) s'est terminé sur une piste, et une seule :\n",
    "\n",
    "> Tester la **forme** de l'examen, pas seulement son existence. Sous cascade, l'examen d'un rang\n",
    "> dépend des **clics au-dessus**, ce qui laisse une signature qu'un test conditionnel pourrait\n",
    "> détecter. Ce serait le troisième contrôle de la série, après l'échangeabilité et\n",
    "> l'identifiabilité.\n",
    "\n",
    "Ce notebook le construit. Il fonctionne — puis il rencontre une limite qui n'est pas la sienne.\n",
    "\n",
    "**Ce que ce notebook établit :**\n",
    "\n",
    "* le test **fonctionne** : il ne rejette jamais sous un modèle de position ($|z| < 1$ à toutes\n",
    "  les sévérités) et rejette massivement sous cascade ($z = -97$ à $-307$) ;\n",
    "* mais un **budget de clics** — un lecteur qui cesse de cliquer une fois servi, tout en\n",
    "  continuant de parcourir le fil — produit **la même signature** : $z = -144$ sous modèle de\n",
    "  position pur. Les deux ne sont **pas séparables** dans des données de clic seules ;\n",
    "* appliqué à **Baidu-ULTR**, le test ne trouve **aucune signature de cascade** : $z = +5{,}6$ au\n",
    "  seuil le plus permissif, $+0{,}5$ au plus strict — le signe est celui du confondant, pas celui\n",
    "  de la cascade ;\n",
    "* et une **erreur de protocole**, découverte en la commettant : restreindre le journal aux\n",
    "  sessions à plusieurs clics fabrique la signature recherchée. Sur un modèle de position pur, la\n",
    "  restriction fait passer $z$ de $+0{,}7$ à $\\mathbf{-45{,}7}$. C'est un *collider*."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "79748b14",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:43:53.811341Z",
     "iopub.status.busy": "2026-08-23T13:43:53.811140Z",
     "iopub.status.idle": "2026-08-23T13:43:54.346426Z",
     "shell.execute_reply": "2026-08-23T13:43:54.346002Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60000 fils de 12 positions, catalogue de 200 contenus\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from ide.exposure import load_digest\n",
    "from ide.logs import (\n",
    "    Impressions,\n",
    "    simulate_cascade,\n",
    "    simulate_feeds,\n",
    "    upstream_dependence_from_counts,\n",
    "    upstream_dependence_test,\n",
    ")\n",
    "from ide.plotting import PALETTE, save_figure, use_project_style\n",
    "\n",
    "use_project_style()\n",
    "\n",
    "CATALOGUE, FEEDS, SLOTS = 200, 60_000, 12\n",
    "print(f\"{FEEDS} fils de {SLOTS} positions, catalogue de {CATALOGUE} contenus\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "88870070",
   "metadata": {},
   "source": [
    "## 1. La construction\n",
    "\n",
    "Les deux formes d'examen ne se distinguent pas par leur allure — une décroissance géométrique et\n",
    "une décroissance polynomiale s'ajustent aussi bien l'une que l'autre sur les premiers rangs. Elles\n",
    "se distinguent par une **indépendance conditionnelle** :\n",
    "\n",
    "* sous un modèle de **position**, l'examen du rang $R$ ne dépend que de $R$ : à contenu et rang\n",
    "  fixés, le clic est indépendant de ce qui s'est passé au-dessus ;\n",
    "* sous un modèle **à cascade**, un clic au-dessus **supprime** l'examen en dessous.\n",
    "\n",
    "Pour chaque cellule $s = (\\text{contenu}, \\text{rang})$, les $n_s$ impressions se répartissent en\n",
    "$n_{1s}$ précédées d'un clic dans le même fil et $n_{0s}$ qui ne le sont pas, pour $k_s$ clics au\n",
    "total. Sous l'hypothèse d'indépendance, les $k_s$ clics se répartissent comme un tirage sans\n",
    "remise, de moments connus exactement :\n",
    "\n",
    "$$\\mathbb{E}[a_s] = \\frac{k_s n_{1s}}{n_s}, \\qquad\n",
    "\\mathbb{V}[a_s] = \\frac{k_s (n_s - k_s)\\, n_{1s} n_{0s}}{n_s^2 (n_s - 1)}$$\n",
    "\n",
    "C'est la statistique de Mantel-Haenszel, stratifiée par cellule. Le conditionnement à la cellule\n",
    "élimine la qualité du contenu : sans lui, les fils qui contiennent un clic en haut sont aussi ceux\n",
    "dont les contenus sont meilleurs, et la comparaison ne mesurerait que cela."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4b10dbe9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:43:54.347282Z",
     "iopub.status.busy": "2026-08-23T13:43:54.347176Z",
     "iopub.status.idle": "2026-08-23T13:43:57.913836Z",
     "shell.execute_reply": "2026-08-23T13:43:57.913438Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Contrôle négatif (position) et contrôle positif (cascade)\n",
      "\n",
      "                        modèle          z          p   cellules   verdict\n",
      "----------------------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             position, η = 0.0      -0.33      0.742       2197   compatible position\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             position, η = 0.5      -0.37      0.711       2150   compatible position\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             position, η = 1.0       0.08      0.934       1882   compatible position\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             position, η = 2.0       0.92      0.355        952   compatible position\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              cascade, γ = 1.0    -307.42          0       2032   REJETTE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             cascade, γ = 0.95    -248.43          0       1925   REJETTE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             cascade, γ = 0.85    -178.71          0       1660   REJETTE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              cascade, γ = 0.6     -96.70          0       1010   REJETTE\n"
     ]
    }
   ],
   "source": [
    "print(\"Contrôle négatif (position) et contrôle positif (cascade)\\n\")\n",
    "print(f\"{'modèle':>30s} {'z':>10s} {'p':>10s} {'cellules':>10s}   verdict\")\n",
    "print(\"-\" * 76)\n",
    "controls = {}\n",
    "for severity in (0.0, 0.5, 1.0, 2.0):\n",
    "    feeds = simulate_feeds([SLOTS] * FEEDS, severity=severity, catalogue=CATALOGUE,\n",
    "                           rng=np.random.default_rng(5))\n",
    "    verdict = upstream_dependence_test(feeds)\n",
    "    controls[f\"position η={severity}\"] = verdict\n",
    "    print(f\"{f'position, η = {severity}':>30s} {verdict.deviation:10.2f} \"\n",
    "          f\"{verdict.p_value:10.3g} {verdict.cells_used:10d}   \"\n",
    "          f\"{'compatible position' if verdict.position_like else 'REJETTE'}\")\n",
    "\n",
    "for continuation in (1.0, 0.95, 0.85, 0.6):\n",
    "    feeds = simulate_cascade([SLOTS] * FEEDS, continuation=continuation, catalogue=CATALOGUE,\n",
    "                             rng=np.random.default_rng(5))\n",
    "    verdict = upstream_dependence_test(feeds)\n",
    "    controls[f\"cascade γ={continuation}\"] = verdict\n",
    "    print(f\"{f'cascade, γ = {continuation}':>30s} {verdict.deviation:10.2f} \"\n",
    "          f\"{verdict.p_value:10.3g} {verdict.cells_used:10d}   \"\n",
    "          f\"{'compatible position' if verdict.position_like else 'REJETTE'}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5d6f8de0",
   "metadata": {},
   "source": [
    "Le test sépare donc parfaitement les deux modèles **quand ils sont purs**. C'est la condition\n",
    "minimale, et elle est remplie.\n",
    "\n",
    "## 2. Ce qu'il ne peut pas séparer\n",
    "\n",
    "Un lecteur qui ne cherche qu'une chose cesse de cliquer une fois servi — **tout en continuant de\n",
    "parcourir le fil**. Son examen est celui d'un modèle de position ; seule la production de clics\n",
    "s'arrête. C'est un **budget de clics**, et c'est un comportement parfaitement ordinaire.\n",
    "\n",
    "La question décide du test : ce budget produit-il la même signature qu'une cascade ?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "49444150",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:43:57.914817Z",
     "iopub.status.busy": "2026-08-23T13:43:57.914743Z",
     "iopub.status.idle": "2026-08-23T13:43:59.530823Z",
     "shell.execute_reply": "2026-08-23T13:43:59.530405Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                          modèle          z   clics/fil   verdict\n",
      "--------------------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       position, budget illimité       0.74       0.990   compatible position\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            position, budget = 3      -2.15       0.979   REJETTE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            position, budget = 2     -21.10       0.917   REJETTE\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            position, budget = 1    -144.24       0.661   REJETTE\n",
      "cascade, γ = 0,85 (pour mémoire)    -178.71\n"
     ]
    }
   ],
   "source": [
    "def budgeted_position_model(budget, rng, severity=1.0, base=0.25, dispersion=0.8):\n",
    "    # Examen intact — il ne dépend que du rang. Seuls les `budget` premiers clics sont conservés.\n",
    "    quality = base * rng.lognormal(0.0, dispersion, size=CATALOGUE)\n",
    "    items = rng.integers(0, CATALOGUE, size=FEEDS * SLOTS).reshape(FEEDS, SLOTS)\n",
    "    ranks = np.tile(np.arange(1, SLOTS + 1), (FEEDS, 1))\n",
    "    drawn = rng.random((FEEDS, SLOTS)) < np.clip(\n",
    "        quality[items] * ranks.astype(float) ** (-severity), 0.0, 1.0)\n",
    "    kept = drawn & (np.cumsum(drawn, axis=1) <= budget)\n",
    "    return Impressions(\n",
    "        items=items.ravel().astype(np.int64), ranks=ranks.ravel().astype(np.int64),\n",
    "        clicks=kept.ravel().astype(float),\n",
    "        feeds=np.repeat(np.arange(FEEDS), SLOTS).astype(np.int64),\n",
    "        feed_lengths=np.full(FEEDS, SLOTS, dtype=np.int64))\n",
    "\n",
    "\n",
    "print(f\"{'modèle':>32s} {'z':>10s} {'clics/fil':>11s}   verdict\")\n",
    "print(\"-\" * 74)\n",
    "budget_scan = {}\n",
    "for budget in (99, 3, 2, 1):\n",
    "    feeds = budgeted_position_model(budget, np.random.default_rng(5))\n",
    "    verdict = upstream_dependence_test(feeds)\n",
    "    per_feed = feeds.clicks.reshape(FEEDS, SLOTS).sum(axis=1).mean()\n",
    "    budget_scan[budget] = (verdict, per_feed)\n",
    "    label = \"position, budget illimité\" if budget == 99 else f\"position, budget = {budget}\"\n",
    "    print(f\"{label:>32s} {verdict.deviation:10.2f} {per_feed:11.3f}   \"\n",
    "          f\"{'compatible position' if verdict.position_like else 'REJETTE'}\")\n",
    "\n",
    "reference = controls[\"cascade γ=0.85\"]\n",
    "print(f\"{'cascade, γ = 0,85 (pour mémoire)':>32s} {reference.deviation:10.2f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "489bd8be",
   "metadata": {},
   "source": [
    "### Une limite d'identification, pas un défaut du test\n",
    "\n",
    "Un budget de **un** produit $z = -144$ ; une cascade véritable produit $z = -179$. Les deux sont\n",
    "massivement rejetés et **rien dans les clics ne les distingue** — c'est attendu, puisque dans les\n",
    "deux cas le journal montre exactement la même chose : après un clic, plus rien.\n",
    "\n",
    "La différence entre les deux est pourtant **considérable pour ce qui nous occupe** :\n",
    "\n",
    "| | Après un clic, le contenu en dessous… | Conséquence pour un réordonnancement |\n",
    "|---|---|---|\n",
    "| **cascade** | n'est pas examiné | le remonter change tout |\n",
    "| **budget de clics** | est examiné mais ne sera pas cliqué | le remonter ne change rien au clic |\n",
    "\n",
    "Autrement dit, la question à laquelle ce test devait répondre — *quelle exposition attribuer aux\n",
    "rangs profonds ?* — reste ouverte, et elle ne peut pas être tranchée par des clics.\n",
    "\n",
    "!!! tip \"Ce qui la trancherait\"\n",
    "    Une mesure de l'**examen** plutôt que du clic : temps d'affichage, profondeur de défilement,\n",
    "    contenu sorti de l'écran. Baidu-ULTR publie précisément ces colonnes — `displayed_time`,\n",
    "    `slipoff_count_after_click`, `serp_height` — que ce dépôt n'a jamais lues. C'est la piste la\n",
    "    plus directe, et elle ne demande aucune donnée nouvelle.\n",
    "\n",
    "## 3. Sur données réelles\n",
    "\n",
    "Le test s'applique à Baidu-ULTR, seul journal du dépôt qui enregistre à la fois l'ordre servi et\n",
    "l'identité des documents."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "eba9831c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:43:59.531656Z",
     "iopub.status.busy": "2026-08-23T13:43:59.531580Z",
     "iopub.status.idle": "2026-08-23T13:43:59.592336Z",
     "shell.execute_reply": "2026-08-23T13:43:59.591955Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baidu-ULTR : 64200 sessions, 524164 documents servis\n",
      "clics par session : moyenne 0.688, 53.4 % sans clic, 13.2 % à plus d'un clic\n",
      "\n",
      "  seuil         z          p   cellules   verdict\n",
      "----------------------------------------------------------\n",
      "      2     5.648   1.62e-08        780   REJETTE\n",
      "      5     1.202      0.229        261   compatible position\n",
      "     10     0.535      0.593        151   compatible position\n",
      "     20     0.609      0.543        108   compatible position\n"
     ]
    }
   ],
   "source": [
    "# Le journal brut de Baidu-ULTR n'est pas versionné (0,9 Go, licence CC BY-NC). Le condensé\n",
    "# porte les quatre comptes par cellule dont le test a besoin — impressions, impressions\n",
    "# précédées d'un clic, clics, clics précédés — et rend le même chiffre à la décimale près.\n",
    "counts = load_digest().upstream_counts(\"baidu\")\n",
    "sessions = load_digest().impressions(\"baidu\")\n",
    "per_session = np.bincount(sessions.feeds, weights=sessions.clicks,\n",
    "                          minlength=sessions.feed_count)\n",
    "\n",
    "print(f\"Baidu-ULTR : {sessions.feed_count} sessions, {sessions.served} documents servis\")\n",
    "print(f\"clics par session : moyenne {per_session.mean():.3f}, \"\n",
    "      f\"{100 * (per_session == 0).mean():.1f} % sans clic, \"\n",
    "      f\"{100 * (per_session > 1).mean():.1f} % à plus d'un clic\\n\")\n",
    "\n",
    "print(f\"{'seuil':>7s} {'z':>9s} {'p':>10s} {'cellules':>10s}   verdict\")\n",
    "print(\"-\" * 58)\n",
    "baidu_scan = {}\n",
    "for minimum in (2, 5, 10, 20):\n",
    "    verdict = upstream_dependence_from_counts(*counts, minimum_impressions=minimum)\n",
    "    baidu_scan[minimum] = verdict\n",
    "    print(f\"{minimum:7d} {verdict.deviation:9.3f} {verdict.p_value:10.3g} \"\n",
    "          f\"{verdict.cells_used:10d}   \"\n",
    "          f\"{'compatible position' if verdict.position_like else 'REJETTE'}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef69a363",
   "metadata": {},
   "source": [
    "### Lecture\n",
    "\n",
    "**Aucune signature de cascade.** L'écart réduit est **positif** à tous les seuils — $+5{,}6$ au\n",
    "plus permissif, $+0{,}5$ à $+0{,}6$ aux plus stricts — alors qu'une cascade le rendrait\n",
    "*négatif*. Le signe compte\n",
    "autant que l'ampleur, et il est ici celui du **confondant** annoncé : un lecteur plus enclin à\n",
    "cliquer clique davantage partout, donc plus haut *et* plus bas.\n",
    "\n",
    "Trois lectures possibles, et le dépôt ne peut pas trancher entre elles :\n",
    "\n",
    "1. l'examen sur Baidu-ULTR est effectivement proche d'un modèle de position ;\n",
    "2. il y a une cascade, mais le confondant d'hétérogénéité la masque — le test est conservateur,\n",
    "   il l'a toujours dit ;\n",
    "3. la couverture est trop mince : 780 cellules au seuil 2, 108 au seuil 20, parce qu'une même\n",
    "   URL réapparaît rarement dans un journal de recherche. C'est la même limite que celle qui\n",
    "   pesait déjà sur l'estimation de $\\hat\\eta$.\n",
    "\n",
    "Le rejet au seuil 2 ($p = 1{,}6 \\times 10^{-8}$) mérite d'être noté pour ce qu'il est : un rejet\n",
    "**du bon côté pour le confondant**, obtenu sur les cellules les moins observées. Il n'indique pas\n",
    "une cascade — il indiquerait plutôt que l'hétérogénéité des lecteurs est réelle et mesurable.\n",
    "\n",
    "**Ce que cela ne fait pas**, et qu'il serait tentant de croire : cela ne valide pas la loi de\n",
    "puissance. Le [notebook 21](21_angles_morts.ipynb) a montré qu'elle se trompe d'un facteur 4 110\n",
    "au douzième rang **si** l'examen est une cascade. Ne pas détecter de cascade n'établit pas son\n",
    "absence, surtout avec 151 cellules.\n",
    "\n",
    "## 4. L'erreur que j'ai commise, et le contrôle qui l'a rattrapée\n",
    "\n",
    "Le tableau ci-dessus donne des sessions dont 53 % n'ont aucun clic. L'idée vient naturellement de\n",
    "restreindre le test aux sessions à **plusieurs clics** — là où un budget de un est exclu par\n",
    "construction, donc là où cascade et budget devraient enfin se séparer.\n",
    "\n",
    "Appliquée à Baidu-ULTR, cette restriction donne $z = -8{,}2$ ($p = 2 \\times 10^{-16}$) : une\n",
    "signature de cascade nette, sur données réelles. J'allais la publier.\n",
    "\n",
    "La restriction demande le journal complet — elle ne se calcule pas depuis des comptes agrégés,\n",
    "ce qui aurait dû être un premier avertissement. Le chiffre ci-dessous est donc repris de la\n",
    "mesure sur le fichier brut, et c'est ce qui suit qui compte."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d787937f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:43:59.593166Z",
     "iopub.status.busy": "2026-08-23T13:43:59.593094Z",
     "iopub.status.idle": "2026-08-23T13:44:01.055476Z",
     "shell.execute_reply": "2026-08-23T13:44:01.055038Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Baidu-ULTR restreint aux sessions à ≥ 2 clics : z = -8,228, p = 1,9e-16, 100 cellules\n",
      "(mesuré sur le journal brut ; voir la note ci-dessus)\n",
      "\n",
      "Le même protocole, appliqué à des journaux dont on connaît la vérité :\n",
      "\n",
      "                          modèle   z (tout)   z (≥ 2 clics)\n",
      "------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       position, budget illimité       0.74          -45.72\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "            position, budget = 2     -21.10          -52.49\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               cascade, γ = 0,85    -178.71  aucune cellule\n"
     ]
    }
   ],
   "source": [
    "def restrict_to_multiclick(impressions, minimum_clicks=2):\n",
    "    per_feed = np.bincount(impressions.feeds, weights=impressions.clicks,\n",
    "                           minlength=impressions.feed_count)\n",
    "    keep = np.isin(impressions.feeds, np.flatnonzero(per_feed >= minimum_clicks))\n",
    "    return Impressions(items=impressions.items[keep], ranks=impressions.ranks[keep],\n",
    "                       clicks=impressions.clicks[keep], feeds=impressions.feeds[keep],\n",
    "                       feed_lengths=impressions.feed_lengths)\n",
    "\n",
    "\n",
    "print(\"Baidu-ULTR restreint aux sessions à ≥ 2 clics : z = -8,228, p = 1,9e-16, 100 cellules\")\n",
    "print(\"(mesuré sur le journal brut ; voir la note ci-dessus)\\n\")\n",
    "\n",
    "print(\"Le même protocole, appliqué à des journaux dont on connaît la vérité :\\n\")\n",
    "print(f\"{'modèle':>32s} {'z (tout)':>10s} {'z (≥ 2 clics)':>15s}\")\n",
    "print(\"-\" * 60)\n",
    "collider = {}\n",
    "for label, feeds in (\n",
    "    (\"position, budget illimité\", budgeted_position_model(99, np.random.default_rng(5))),\n",
    "    (\"position, budget = 2\", budgeted_position_model(2, np.random.default_rng(5))),\n",
    "    (\"cascade, γ = 0,85\", simulate_cascade([SLOTS] * FEEDS, continuation=0.85,\n",
    "                                           catalogue=CATALOGUE, rng=np.random.default_rng(5))),\n",
    "):\n",
    "    full = upstream_dependence_test(feeds, minimum_impressions=2)\n",
    "    limited = upstream_dependence_test(restrict_to_multiclick(feeds), minimum_impressions=2)\n",
    "    collider[label] = (full.deviation, limited.deviation, limited.cells_used)\n",
    "    shown = f\"{limited.deviation:15.2f}\" if limited.cells_used else f\"{'aucune cellule':>15s}\"\n",
    "    print(f\"{label:>32s} {full.deviation:10.2f} {shown}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f69b2c6f",
   "metadata": {},
   "source": [
    "!!! failure \"La restriction fabrique la signature qu'elle cherche\"\n",
    "    Sur un journal simulé sous **modèle de position pur, sans aucune cascade et sans budget**, la\n",
    "    restriction aux sessions à deux clics ou plus fait passer l'écart réduit de $+0{,}74$ à\n",
    "    $\\mathbf{-45{,}7}$.\n",
    "\n",
    "    La raison est classique et porte un nom : le nombre de clics d'un fil est un **collider** de\n",
    "    ses clics individuels. Conditionner dessus induit une dépendance négative entre eux — c'est\n",
    "    le paradoxe de Berkson. Le $-8{,}2$ obtenu sur Baidu-ULTR ne mesure donc pas une cascade : il\n",
    "    mesure la sélection que je venais d'opérer.\n",
    "\n",
    "C'est la troisième fois dans ce dépôt qu'un protocole apparemment raisonnable fabrique son propre\n",
    "résultat, et la troisième fois qu'un **contrôle sur données simulées** le rattrape avant\n",
    "publication. La règle qui s'en dégage vaut d'être écrite : *tout protocole appliqué à des données\n",
    "réelles doit d'abord être appliqué à des données dont on connaît la réponse.*\n",
    "\n",
    "La mise en garde est désormais dans la documentation de la fonction, à l'endroit où quelqu'un\n",
    "aura l'idée de le refaire.\n",
    "\n",
    "## 5. La figure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ab375d47",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:44:01.056298Z",
     "iopub.status.busy": "2026-08-23T13:44:01.056221Z",
     "iopub.status.idle": "2026-08-23T13:44:01.489889Z",
     "shell.execute_reply": "2026-08-23T13:44:01.489509Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1488x1032 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure, axes = plt.subplots(2, 2, figsize=(12.4, 8.6))\n",
    "\n",
    "# (a) contrôles négatif et positif\n",
    "ax = axes[0, 0]\n",
    "labels = [f\"position\\nη = {s}\" for s in (0.0, 0.5, 1.0, 2.0)]\n",
    "labels += [f\"cascade\\nγ = {c}\" for c in (1.0, 0.95, 0.85, 0.6)]\n",
    "values = [controls[f\"position η={s}\"].deviation for s in (0.0, 0.5, 1.0, 2.0)]\n",
    "values += [controls[f\"cascade γ={c}\"].deviation for c in (1.0, 0.95, 0.85, 0.6)]\n",
    "colours = [PALETTE[\"order\"]] * 4 + [PALETTE[\"field\"]] * 4\n",
    "ax.bar(np.arange(len(values)), values, color=colours, alpha=0.85, width=0.65)\n",
    "ax.axhline(-1.96, color=PALETTE[\"neutral\"], linestyle=\":\", linewidth=1.2)\n",
    "ax.axhline(0.0, color=PALETTE[\"neutral\"], linewidth=0.8)\n",
    "ax.set_yscale(\"symlog\", linthresh=10)\n",
    "ax.set_xticks(np.arange(len(values)))\n",
    "ax.set_xticklabels(labels, fontsize=7)\n",
    "ax.set_ylabel(\"écart réduit $z$ (symlog)\")\n",
    "ax.set_title(\"(a) le test sépare les deux modèles purs\")\n",
    "\n",
    "# (b) le budget imite la cascade\n",
    "ax = axes[0, 1]\n",
    "budgets = [99, 3, 2, 1]\n",
    "positions = np.arange(len(budgets) + 1)\n",
    "values = [budget_scan[b][0].deviation for b in budgets] + [reference.deviation]\n",
    "names = [\"illimité\", \"3\", \"2\", \"1\", \"cascade\\nγ = 0,85\"]\n",
    "ax.bar(positions, values, color=[PALETTE[\"order\"]] * 4 + [PALETTE[\"field\"]], alpha=0.85,\n",
    "       width=0.6)\n",
    "ax.axhline(-1.96, color=PALETTE[\"neutral\"], linestyle=\":\", linewidth=1.2)\n",
    "ax.set_yscale(\"symlog\", linthresh=10)\n",
    "ax.set_xticks(positions)\n",
    "ax.set_xticklabels(names, fontsize=8)\n",
    "ax.set_xlabel(\"budget de clics d'un modèle de POSITION\")\n",
    "ax.set_ylabel(\"écart réduit $z$ (symlog)\")\n",
    "ax.annotate(\"indiscernables\\ndans des clics seuls\", xy=(3.5, -155), xytext=(1.9, -22),\n",
    "            fontsize=9, color=PALETTE[\"disorder\"], ha=\"center\",\n",
    "            arrowprops={\"arrowstyle\": \"-[, widthB=2.6, lengthB=0.6\",\n",
    "                        \"color\": PALETTE[\"disorder\"], \"linewidth\": 1.2})\n",
    "ax.set_title(\"(b) un budget de clics imite une cascade\")\n",
    "\n",
    "# (c) Baidu-ULTR\n",
    "ax = axes[1, 0]\n",
    "thresholds = list(baidu_scan)\n",
    "deviations = [baidu_scan[t].deviation for t in thresholds]\n",
    "cells = [baidu_scan[t].cells_used for t in thresholds]\n",
    "ax.plot(np.arange(len(thresholds)), deviations, \"o-\", color=PALETTE[\"remedy\"], markersize=6,\n",
    "        label=\"Baidu-ULTR\")\n",
    "ax.axhspan(-1.96, 1.96, color=PALETTE[\"neutral\"], alpha=0.12)\n",
    "ax.axhline(0.0, color=PALETTE[\"neutral\"], linewidth=0.8)\n",
    "ax.text(0.05, 1.0, \"zone de non-rejet\", fontsize=8, color=PALETTE[\"neutral\"])\n",
    "ax.annotate(\"une cascade\\npousserait ici\", xy=(1.6, -3.0), xytext=(1.15, -5.6), fontsize=8,\n",
    "            color=PALETTE[\"field\"],\n",
    "            arrowprops={\"arrowstyle\": \"->\", \"color\": PALETTE[\"field\"], \"linewidth\": 1.0})\n",
    "for index, (threshold, count) in enumerate(zip(thresholds, cells, strict=True)):\n",
    "    ax.annotate(f\"{count} cellules\", xy=(index, deviations[index]), xytext=(0, 9),\n",
    "                textcoords=\"offset points\", fontsize=7.5, ha=\"center\",\n",
    "                color=PALETTE[\"neutral\"])\n",
    "ax.set_xticks(np.arange(len(thresholds)))\n",
    "ax.set_xticklabels([str(t) for t in thresholds])\n",
    "ax.set_xlabel(\"seuil d'impressions par cellule\")\n",
    "ax.set_ylabel(\"écart réduit $z$\")\n",
    "ax.set_ylim(-6.5, 8.5)\n",
    "ax.set_title(\"(c) sur Baidu-ULTR, aucune signature de cascade\")\n",
    "\n",
    "# (d) le collider\n",
    "ax = axes[1, 1]\n",
    "names = list(collider)\n",
    "full_values = [collider[name][0] for name in names]\n",
    "limited_values = [collider[name][1] if collider[name][2] else np.nan for name in names]\n",
    "positions = np.arange(len(names))\n",
    "ax.bar(positions - 0.19, full_values, width=0.36, color=PALETTE[\"order\"], alpha=0.85,\n",
    "       label=\"journal entier\")\n",
    "ax.bar(positions + 0.19, limited_values, width=0.36, color=PALETTE[\"disorder\"], alpha=0.85,\n",
    "       label=\"restreint aux fils à ≥ 2 clics\")\n",
    "for index, name in enumerate(names):\n",
    "    if not collider[name][2]:\n",
    "        ax.text(index + 0.19, -1.0, \"aucune\\ncellule\", fontsize=7, ha=\"center\", va=\"top\",\n",
    "                color=PALETTE[\"neutral\"])\n",
    "ax.axhline(-1.96, color=PALETTE[\"neutral\"], linestyle=\":\", linewidth=1.2)\n",
    "ax.set_yscale(\"symlog\", linthresh=10)\n",
    "ax.set_xticks(positions)\n",
    "ax.set_xticklabels([name.replace(\", \", \"\\n\") for name in names], fontsize=7.5)\n",
    "ax.set_ylabel(\"écart réduit $z$ (symlog)\")\n",
    "ax.set_title(\"(d) la restriction fabrique la signature qu'elle cherche\")\n",
    "ax.legend(loc=\"lower left\", fontsize=8)\n",
    "\n",
    "save_figure(figure, \"fig22_test_de_forme\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "06b8f334",
   "metadata": {},
   "source": [
    "## 6. Ce que ce troisième contrôle apporte\n",
    "\n",
    "**Il existe, et il fonctionne.** Le dépôt dispose désormais de trois contrôles à passer sur un\n",
    "journal, dans cet ordre :\n",
    "\n",
    "| Contrôle | Question | Verdict sur MIND | Verdict sur Baidu-ULTR |\n",
    "|---|---|---|---|\n",
    "| **échangeabilité** | l'ordre dit-il quelque chose ? | non ($z = +0{,}12$) | oui ($z = -206$) |\n",
    "| **identifiabilité** | y a-t-il de quoi estimer $\\eta$ ? | oui, artificiellement | oui, 55 documents |\n",
    "| **forme** | l'examen dépend-il de l'amont ? | inapplicable | non détecté ($z = +0{,}5$ à $+0{,}6$) |\n",
    "\n",
    "**Mais il ne tranche pas ce qu'on voulait trancher.** Un budget de clics imite une cascade, et\n",
    "rien dans les clics ne les sépare. La question — *quelle exposition attribuer aux rangs\n",
    "profonds ?* — reste ouverte, et elle demande une mesure de l'examen, non du clic.\n",
    "\n",
    "**Et il a coûté une erreur, rattrapée.** Restreindre aux sessions à plusieurs clics semblait la\n",
    "façon évidente de séparer budget et cascade. C'est un collider, et le contrôle sur données\n",
    "simulées l'a montré avant publication.\n",
    "\n",
    "## Piste ouverte\n",
    "\n",
    "Baidu-ULTR publie `displayed_time`, `serp_height` et `slipoff_count_after_click` — des mesures de\n",
    "l'**examen** et non du clic. Ce dépôt n'a jamais lu ces colonnes. Elles trancheraient précisément\n",
    "ce que les clics ne peuvent pas : un contenu situé sous un clic a-t-il été affiché assez longtemps\n",
    "pour être vu ?\n",
    "\n",
    "C'est la suite directe, elle ne demande aucune donnée nouvelle, et elle est la seule voie connue\n",
    "pour séparer les deux modèles."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}