{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "68c73488",
   "metadata": {},
   "source": [
    "# 16 — L'exploration réellement enregistrée dans MIND\n",
    "\n",
    "Le [notebook 15](15_rang_adverse_et_severite.ipynb) a montré comment estimer la sévérité $\\eta$\n",
    "du biais de position au lieu de la poser, et ce que coûte de la poser de travers : jusqu'à\n",
    "**+179 %** sur le chiffre publié. Il a aussi montré à quelle condition cette estimation existe —\n",
    "que la plateforme n'ait pas toujours classé les mêmes contenus aux mêmes places.\n",
    "\n",
    "Cette condition est une propriété du **jeu de données**, pas de la méthode. D'où le préalable\n",
    "inscrit à la feuille de route : mesurer l'exploration réelle avant d'évaluer quoi que ce soit.\n",
    "Ce notebook la mesure sur [MIND](https://msnews.github.io/) (*Microsoft News Dataset*, Wu et al.,\n",
    "ACL 2020), le jeu de référence de la recommandation d'actualité, celui sur lequel l'évaluation\n",
    "de l'[ADE](../docs/ade.md) était prévue.\n",
    "\n",
    "**Ce que ce notebook établit :**\n",
    "\n",
    "* l'ordre enregistré dans MIND est **indiscernable d'un mélange** — $z = +0{,}12$, $p = 0{,}91$,\n",
    "  répliqué sur le second découpage — dans un jeu où le test détecterait $\\eta = 0{,}02$ à douze\n",
    "  écarts-types ;\n",
    "* la courbe la plus naturelle à tracer, le taux de clic par position, y décroît pourtant de\n",
    "  0,108 à 0,038 et donne $\\hat\\eta = 0{,}39$ : un **artefact de composition**, les positions\n",
    "  élevées n'existant que dans les fils longs ;\n",
    "* `estimate_position_bias` accepte ce jeu sans broncher et renvoie **trois sévérités\n",
    "  incompatibles** selon un seuil de nuisance, dont une négative, toutes assorties d'une erreur\n",
    "  type inférieure à 0,005 — le contrôle d'identifiabilité du notebook 15 est donc nécessaire et\n",
    "  **non suffisant** ;\n",
    "* le mélange ne débiaise pas les clics : il **détruit la variable** qui permettrait de les\n",
    "  corriger, et ramène l'analyste à l'estimation naïve dont le notebook 14 a mesuré l'erreur.\n",
    "\n",
    "Le jeu brut n'est pas versionné — licence de recherche Microsoft, 135 Mo. Ce notebook lit le\n",
    "**condensé** `data/mind_digest.npz`, qui rend à l'identique tous les chiffres ci-dessous et se\n",
    "reconstruit par `scripts/fetch_mind.py` puis `scripts/build_mind_digest.py`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "48ec0bc3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:44.942792Z",
     "iopub.status.busy": "2026-08-23T13:40:44.942481Z",
     "iopub.status.idle": "2026-08-23T13:40:45.671870Z",
     "shell.execute_reply": "2026-08-23T13:40:45.671365Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train   156965 fils    5843444 contenus servis  taux de clic 0.0404\n",
      "       longueur de fil : médiane 24, moyenne 37.2, maximum 299\n",
      "dev      73152 fils    2740998 contenus servis  taux de clic 0.0406\n",
      "       longueur de fil : médiane 23, moyenne 37.5, maximum 295\n",
      "\n",
      "source (SHA-256) : a424547c8fa17c9ea4879c2110221c2b2f2709f4f7615ede0b0d8e4765158658\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from ide.mind import (\n",
    "    click_rate_by_rank,\n",
    "    detectable_severity,\n",
    "    exchangeability_test,\n",
    "    load_digest,\n",
    "    naive_severity_fit,\n",
    "    simulate_feeds,\n",
    ")\n",
    "from ide.offpolicy import (\n",
    "    estimate_position_bias,\n",
    "    naive,\n",
    "    rank_propensities,\n",
    "    simulate_logged_feedback,\n",
    "    simulate_ranked_feedback,\n",
    "    snips,\n",
    "    value_under_policy,\n",
    ")\n",
    "from ide.plotting import PALETTE, save_figure, use_project_style\n",
    "\n",
    "use_project_style()\n",
    "\n",
    "digest = load_digest()\n",
    "train = digest.impressions(\"train\")\n",
    "dev = digest.impressions(\"dev\")\n",
    "\n",
    "for split, impressions in ((\"train\", train), (\"dev\", dev)):\n",
    "    lengths = impressions.feed_lengths\n",
    "    print(f\"{split:6s} {impressions.feed_count:7d} fils  {impressions.served:9d} contenus servis\"\n",
    "          f\"  taux de clic {impressions.click_rate:.4f}\")\n",
    "    print(f\"       longueur de fil : médiane {np.median(lengths):.0f}, moyenne \"\n",
    "          f\"{lengths.mean():.1f}, maximum {lengths.max()}\")\n",
    "print(f\"\\nsource (SHA-256) : {digest.sources['train']}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b17c5dd",
   "metadata": {},
   "source": [
    "## 1. La condition d'identifiabilité est satisfaite — abondamment\n",
    "\n",
    "Le notebook 15 avait posé le contrôle à faire avant d'estimer $\\eta$ : combien de contenus ont\n",
    "été servis à **plusieurs rangs distincts** ? C'est la seule variation qui identifie la sévérité,\n",
    "et une plateforme parfaitement déterministe n'en produit aucune.\n",
    "\n",
    "Sur MIND, la réponse est sans ambiguïté."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "1fc1c687",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:45.672753Z",
     "iopub.status.busy": "2026-08-23T13:40:45.672640Z",
     "iopub.status.idle": "2026-08-23T13:40:45.959917Z",
     "shell.execute_reply": "2026-08-23T13:40:45.959421Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "contenus distincts                       : 20288\n",
      "contenus au-dessus du seuil d'impressions: 2950\n",
      "contenus vus à plusieurs rangs           : 2655\n",
      "rangs distincts par contenu (médiane)    : 16\n",
      "rang maximal observé                     : 299\n",
      "\n",
      "Un contenu médian retenu a été servi à seize positions différentes.\n",
      "À ce compte, l'exploration paraît idéale — c'est ce que ce notebook va défaire.\n"
     ]
    }
   ],
   "source": [
    "items, ranks, clicks = digest.rows(\"train\")\n",
    "coverage = digest.coverage(\"train\")\n",
    "print(f\"contenus distincts                       : {coverage.items}\")\n",
    "print(f\"contenus au-dessus du seuil d'impressions: {coverage.items_above_threshold}\")\n",
    "print(f\"contenus vus à plusieurs rangs           : {coverage.items_with_variation}\")\n",
    "print(f\"rangs distincts par contenu (médiane)    : {coverage.median_distinct_ranks:.0f}\")\n",
    "print(f\"rang maximal observé                     : {coverage.maximum_rank}\")\n",
    "print(\"\\nUn contenu médian retenu a été servi à seize positions différentes.\")\n",
    "print(\"À ce compte, l'exploration paraît idéale — c'est ce que ce notebook va défaire.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b86ddf41",
   "metadata": {},
   "source": [
    "## 2. La courbe qu'on trace naturellement\n",
    "\n",
    "Taux de clic par position, tous fils confondus. C'est le premier graphique que produit quiconque\n",
    "cherche un biais de position, et il a exactement l'allure attendue."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5b6a96bc",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:45.960749Z",
     "iopub.status.busy": "2026-08-23T13:40:45.960669Z",
     "iopub.status.idle": "2026-08-23T13:40:46.129255Z",
     "shell.execute_reply": "2026-08-23T13:40:46.128789Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " rang  taux de clic\n",
      "--------------------\n",
      "    1       0.10825\n",
      "    2       0.10778\n",
      "    3       0.08091\n",
      "    5       0.06580\n",
      "   10       0.05038\n",
      "   15       0.04269\n",
      "   20       0.03797\n",
      "\n",
      "ajustement log-log sur les rangs 1-20 : η = 0.388\n",
      "Une décroissance de 0,108 à 0,038, régulière, sur 5,8 millions de contenus servis.\n"
     ]
    }
   ],
   "source": [
    "aggregate = click_rate_by_rank(train, maximum_rank=20)\n",
    "severity_aggregate = naive_severity_fit(train, maximum_rank=20)\n",
    "\n",
    "print(f\"{'rang':>5s} {'taux de clic':>13s}\")\n",
    "print(\"-\" * 20)\n",
    "for position in (1, 2, 3, 5, 10, 15, 20):\n",
    "    print(f\"{position:5d} {aggregate[position - 1]:13.5f}\")\n",
    "print(f\"\\najustement log-log sur les rangs 1-20 : η = {severity_aggregate:.3f}\")\n",
    "print(\"Une décroissance de 0,108 à 0,038, régulière, sur 5,8 millions de contenus servis.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0ebfbcbe",
   "metadata": {},
   "source": [
    "### Ce que cette courbe mesure en réalité\n",
    "\n",
    "Les fils de MIND n'ont pas tous la même longueur : la médiane est de 24 contenus, la moyenne de\n",
    "37, le maximum de 299. Or **la position 20 n'existe que dans les fils d'au moins 20 contenus**,\n",
    "et le taux de clic *par contenu servi* y est mécaniquement plus faible — un lecteur qui clique\n",
    "une fois dans un fil de 100 contenus produit un taux de 0,01, le même clic dans un fil de 4\n",
    "contenus en produit 0,25.\n",
    "\n",
    "La courbe agrégée mélange donc deux choses : l'effet du rang, et la composition du mélange de\n",
    "longueurs. Il suffit de tenir la longueur fixée pour les séparer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4fa20c07",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:46.130085Z",
     "iopub.status.busy": "2026-08-23T13:40:46.130000Z",
     "iopub.status.idle": "2026-08-23T13:40:46.255985Z",
     "shell.execute_reply": "2026-08-23T13:40:46.255525Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " longueur    fils   taux de clic aux positions 1 à 10\n",
      "------------------------------------------------------------------------------\n",
      "        6    3053   0.173 0.179 0.180 0.175 0.179 0.184\n",
      "       10    4279   0.125 0.120 0.125 0.120 0.118 0.118 0.122 0.125 0.112 0.116\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       15    2184   0.079 0.082 0.073 0.077 0.085 0.083 0.082 0.080 0.089 0.073\n",
      "       20    2187   0.055 0.069 0.069 0.060 0.059 0.067 0.070 0.062 0.064 0.068\n",
      "       30    2412   0.053 0.049 0.048 0.053 0.050 0.053 0.055 0.046 0.056 0.051\n",
      "       50    1108   0.034 0.039 0.044 0.038 0.041 0.043 0.026 0.042 0.039 0.041\n",
      "\n",
      " longueur  η apparent\n",
      "----------------------\n",
      "        6      -0.022\n",
      "       10       0.025\n",
      "       15      -0.025\n",
      "       20      -0.033\n",
      "       30      -0.002\n",
      "       50       0.037\n",
      "\n",
      "À longueur fixée, la courbe est plate — et sa pente change de signe d'une longueur à\n",
      "l'autre. Les 0,39 de la courbe agrégée ne mesuraient que le mélange des longueurs.\n"
     ]
    }
   ],
   "source": [
    "print(f\"{'longueur':>9s} {'fils':>7s}   taux de clic aux positions 1 à 10\")\n",
    "print(\"-\" * 78)\n",
    "fixed = {}\n",
    "for length in (6, 10, 15, 20, 30, 50):\n",
    "    count = int((train.feed_lengths == length).sum())\n",
    "    rates = click_rate_by_rank(train, maximum_rank=min(length, 10), feed_length=length)\n",
    "    fixed[length] = click_rate_by_rank(train, maximum_rank=length, feed_length=length)\n",
    "    row = \" \".join(f\"{value:.3f}\" for value in rates)\n",
    "    print(f\"{length:9d} {count:7d}   {row}\")\n",
    "\n",
    "print(f\"\\n{'longueur':>9s} {'η apparent':>11s}\")\n",
    "print(\"-\" * 22)\n",
    "for length in (6, 10, 15, 20, 30, 50):\n",
    "    print(f\"{length:9d} {naive_severity_fit(train, maximum_rank=length, feed_length=length):11.3f}\")\n",
    "print(\"\\nÀ longueur fixée, la courbe est plate — et sa pente change de signe d'une longueur à\")\n",
    "print(\"l'autre. Les 0,39 de la courbe agrégée ne mesuraient que le mélange des longueurs.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "92e9a85a",
   "metadata": {},
   "source": [
    "## 3. Le test d'échangeabilité\n",
    "\n",
    "Le contrôle par longueur fixée est parlant mais partiel : il jette la plupart des fils, et\n",
    "laisse encore jouer la qualité moyenne des contenus d'un fil ou l'appétit de clic de son lecteur.\n",
    "\n",
    "Le test exact conditionne au fil. Pour un fil de longueur $L$ portant $k$ clics, on somme les\n",
    "rangs normalisés $u_R = (R - 1/2)/L$ des contenus cliqués. Sous l'hypothèse que **les clics sont\n",
    "indifférents à la position**, ces $k$ positions sont un tirage sans remise parmi les $L$\n",
    "positions du fil, d'espérance et de variance connues exactement :\n",
    "\n",
    "$$\\mathbb{E} = k\\,\\bar{u}, \\qquad \\mathbb{V} = \\frac{k(L-k)}{L-1}\\,\\sigma^2_u.$$\n",
    "\n",
    "Un biais de position concentre les clics en haut : il rend l'écart réduit **négatif**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3e18161c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:46.256784Z",
     "iopub.status.busy": "2026-08-23T13:40:46.256707Z",
     "iopub.status.idle": "2026-08-23T13:40:46.410354Z",
     "shell.execute_reply": "2026-08-23T13:40:46.409857Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " découpage  fils utiles  somme observée     attendue        z       p\n",
      "----------------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     train       156965        118187.9     118172.0    0.116   0.908\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       dev        73152         55717.7      55691.5    0.278   0.781\n",
      "\n",
      "Les clics tombent exactement là où le hasard les mettrait, sur les deux découpages.\n"
     ]
    }
   ],
   "source": [
    "print(f\"{'découpage':>10s} {'fils utiles':>12s} {'somme observée':>15s} {'attendue':>12s}\"\n",
    "      f\" {'z':>8s} {'p':>7s}\")\n",
    "print(\"-\" * 70)\n",
    "verdicts = {}\n",
    "for split, impressions in ((\"train\", train), (\"dev\", dev)):\n",
    "    verdict = exchangeability_test(impressions)\n",
    "    verdicts[split] = verdict\n",
    "    print(f\"{split:>10s} {verdict.feeds_used:12d} {verdict.statistic:15.1f}\"\n",
    "          f\" {verdict.expectation:12.1f} {verdict.deviation:8.3f} {verdict.p_value:7.3f}\")\n",
    "\n",
    "print(\"\\nLes clics tombent exactement là où le hasard les mettrait, sur les deux découpages.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "144c7b07",
   "metadata": {},
   "source": [
    "### Un test qui ne rejette rien ne dit rien — tant qu'on ignore ce qu'il rejetterait\n",
    "\n",
    "C'est l'étalonnage indispensable d'un résultat négatif. On simule des journaux de **même\n",
    "structure de fils** que MIND — mêmes longueurs, même distribution — sous un biais de position de\n",
    "sévérité connue, et on relève ce que le test en dit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "2900e088",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:46.411155Z",
     "iopub.status.busy": "2026-08-23T13:40:46.411085Z",
     "iopub.status.idle": "2026-08-23T13:40:49.122564Z",
     "shell.execute_reply": "2026-08-23T13:40:49.122062Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " η simulé          z   η naïf agrégé\n",
      "-------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     0.00       0.94          -0.003\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     0.02      -9.02           0.017\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     0.05     -26.95           0.049\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     0.10     -49.22           0.101\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     0.25    -104.44           0.245\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     0.50    -161.23           0.510\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     1.00    -199.34           1.018\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "écart réduit observé sur MIND : +0.116\n",
      "sévérité minimale détectable  : η ≈ 0.0038\n",
      "\n",
      "Le test voit η = 0,02 à douze écarts-types. Sur MIND il ne voit rien : l'ordre\n",
      "enregistré ne porte aucune information de placement au-delà de η ≈ 0,005.\n"
     ]
    }
   ],
   "source": [
    "generator = np.random.default_rng(16)\n",
    "print(f\"{'η simulé':>9s} {'z':>10s} {'η naïf agrégé':>15s}\")\n",
    "print(\"-\" * 37)\n",
    "power = []\n",
    "for severity in (0.0, 0.02, 0.05, 0.10, 0.25, 0.50, 1.00):\n",
    "    simulated = simulate_feeds(train.feed_lengths, severity=severity, rng=generator)\n",
    "    deviation = exchangeability_test(simulated).deviation\n",
    "    power.append((severity, deviation))\n",
    "    print(f\"{severity:9.2f} {deviation:10.2f} {naive_severity_fit(simulated, 20):15.3f}\")\n",
    "\n",
    "threshold = detectable_severity(train.feed_lengths, probe=0.02, rng=np.random.default_rng(17))\n",
    "print(f\"\\nécart réduit observé sur MIND : {verdicts['train'].deviation:+.3f}\")\n",
    "print(f\"sévérité minimale détectable  : η ≈ {threshold:.4f}\")\n",
    "print(\"\\nLe test voit η = 0,02 à douze écarts-types. Sur MIND il ne voit rien : l'ordre\")\n",
    "print(\"enregistré ne porte aucune information de placement au-delà de η ≈ 0,005.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f3225601",
   "metadata": {},
   "source": [
    "La documentation de MIND le disait, en une ligne : *« the orders of news in a impressions have\n",
    "been shuffled »*. Une ligne de documentation ne dit toutefois ni ce qu'il en reste, ni ce que la\n",
    "mesure donne quand on l'ignore. Les deux se mesurent — et la suite montre que les ignorer ne\n",
    "produit pas une erreur visible, mais un chiffre confiant.\n",
    "\n",
    "## 4. Ce que l'estimateur du notebook 15 fait de ce jeu\n",
    "\n",
    "`estimate_position_bias` vérifie l'identifiabilité avant de répondre : il refuse d'estimer quand\n",
    "aucun contenu n'a changé de rang. Sur MIND, des milliers de contenus ont changé de rang. Il\n",
    "répond donc — et le seuil d'impressions, simple paramètre de nuisance, décide de la réponse."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "994aa7cb",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:40:49.123369Z",
     "iopub.status.busy": "2026-08-23T13:40:49.123294Z",
     "iopub.status.idle": "2026-08-23T13:41:01.684625Z",
     "shell.execute_reply": "2026-08-23T13:41:01.684098Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " seuil   η estimé  erreur type  contenus  identifiable\n",
      "--------------------------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     5    -0.1307       0.0020      2655  oui\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    10    -0.0465       0.0022      2050  oui\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    20     0.0533       0.0027      1429  oui\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    50     0.1941       0.0043       705  oui\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   100     0.2546       0.0065       369  oui\n",
      "\n",
      "La sévérité estimée est une fonction croissante du seuil — un simple paramètre de\n",
      "nuisance — et parcourt −0,13 à +0,25 sans jamais cesser d'être « significative ».\n",
      "Une sévérité négative voudrait dire que les positions basses reçoivent plus de clics.\n"
     ]
    }
   ],
   "source": [
    "print(f\"{'seuil':>6s} {'η estimé':>10s} {'erreur type':>12s} {'contenus':>9s}  identifiable\")\n",
    "print(\"-\" * 56)\n",
    "estimates = {}\n",
    "for minimum in (5, 10, 20, 50, 100):\n",
    "    estimate = estimate_position_bias(items, ranks, clicks, minimum_impressions=minimum)\n",
    "    estimates[minimum] = estimate\n",
    "    print(f\"{minimum:6d} {estimate.severity:10.4f} {estimate.standard_error:12.4f}\"\n",
    "          f\" {estimate.items_with_variation:9d}  {'oui' if estimate.identifiable else 'NON'}\")\n",
    "\n",
    "print(\"\\nLa sévérité estimée est une fonction croissante du seuil — un simple paramètre de\")\n",
    "print(\"nuisance — et parcourt −0,13 à +0,25 sans jamais cesser d'être « significative ».\")\n",
    "print(\"Une sévérité négative voudrait dire que les positions basses reçoivent plus de clics.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46f3c70a",
   "metadata": {},
   "source": [
    "!!! failure \"Correction du notebook 15\"\n",
    "    Le contrôle d'identifiabilité y était présenté comme la garde à passer avant d'estimer\n",
    "    $\\eta$. Il est **nécessaire et non suffisant** : il compte la variation de rang sans dire\n",
    "    d'où elle vient, et une variation **artificielle** la satisfait mieux que n'importe quelle\n",
    "    exploration réelle. Le test d'échangeabilité est le contrôle manquant, et il doit précéder\n",
    "    l'estimation, non la suivre.\n",
    "\n",
    "## 5. Le mélange ne débiaise pas les clics\n",
    "\n",
    "Une lecture répandue veut que mélanger l'ordre enregistré *protège* du biais de position. Elle\n",
    "confond deux choses.\n",
    "\n",
    "Les clics de MIND ont été produits par des lecteurs qui voyaient un fil **ordonné** — l'ordre\n",
    "réel de Microsoft News, celui que le jeu n'a pas conservé. Ils portent donc le biais de position\n",
    "en entier. Ce que le mélange a retiré, c'est le **rang**, seul régresseur qui aurait permis d'en\n",
    "tenir compte.\n",
    "\n",
    "L'expérience ci-dessous le montre sur un journal simulé dont on connaît la vérité : on l'estime\n",
    "une fois tel quel, une fois après avoir mélangé l'ordre à l'intérieur de chaque fil."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f42da2b6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:41:01.685441Z",
     "iopub.status.busy": "2026-08-23T13:41:01.685368Z",
     "iopub.status.idle": "2026-08-23T13:41:02.145977Z",
     "shell.execute_reply": "2026-08-23T13:41:02.145428Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sévérité vraie du journal simulé   : 1.000\n",
      "estimée sur l'ordre conservé       : 1.003 ± 0.008\n",
      "estimée après mélange de l'ordre   : -0.003 ± 0.006  (identifiable=True)\n",
      "\n",
      "Les clics sont les mêmes dans les deux colonnes : le biais de position n'a pas bougé.\n",
      "Seule la variable qui permettait de le corriger a disparu.\n"
     ]
    }
   ],
   "source": [
    "TRUE_SEVERITY = 1.0\n",
    "rng = np.random.default_rng(23)\n",
    "catalogue = rng.uniform(0.15, 0.9, 12)\n",
    "logged_items, logged_ranks_raw, logged_clicks = simulate_ranked_feedback(\n",
    "    catalogue, 40_000, TRUE_SEVERITY, exploration=0.5, rng=rng\n",
    ")\n",
    "\n",
    "intact = estimate_position_bias(logged_items, logged_ranks_raw, logged_clicks)\n",
    "\n",
    "# Le mélange de MIND : à l'intérieur de chaque fil, les rangs sont redistribués au hasard.\n",
    "per_feed = logged_ranks_raw.reshape(-1, catalogue.size)\n",
    "shuffled_ranks = rng.permuted(per_feed, axis=1).ravel()\n",
    "erased = estimate_position_bias(logged_items, shuffled_ranks, logged_clicks)\n",
    "\n",
    "print(f\"sévérité vraie du journal simulé   : {TRUE_SEVERITY:.3f}\")\n",
    "print(f\"estimée sur l'ordre conservé       : {intact.severity:.3f} ± {intact.standard_error:.3f}\")\n",
    "print(f\"estimée après mélange de l'ordre   : {erased.severity:.3f} ± {erased.standard_error:.3f}\"\n",
    "      f\"  (identifiable={erased.identifiable})\")\n",
    "print(\"\\nLes clics sont les mêmes dans les deux colonnes : le biais de position n'a pas bougé.\")\n",
    "print(\"Seule la variable qui permettait de le corriger a disparu.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab33a5f1",
   "metadata": {},
   "source": [
    "### Ce que cette perte coûte à l'évaluation\n",
    "\n",
    "Le [notebook 15](15_rang_adverse_et_severite.ipynb) a chiffré ce que coûte un $\\eta$ posé de\n",
    "travers. Un journal mélangé conduit l'analyste à $\\hat\\eta \\approx 0$ — c'est-à-dire à supposer\n",
    "qu'aucune position n'est plus vue qu'une autre, donc à ne rien corriger du tout."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "90386e4a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:41:02.146790Z",
     "iopub.status.busy": "2026-08-23T13:41:02.146718Z",
     "iopub.status.idle": "2026-08-23T13:41:02.163725Z",
     "shell.execute_reply": "2026-08-23T13:41:02.163254Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "coût réel du filtre de diversité              :   6.61 %\n",
      "estimé avec η lu sur l'ordre conservé (1.00) :   6.68 %\n",
      "estimé avec η lu sur l'ordre mélangé (-0.00) :   0.00 %\n",
      "\n",
      "erreur relative après mélange : -100 %\n",
      "\n",
      "Le zéro n'est pas une coïncidence numérique. Sous η = 0, toutes les positions sont\n",
      "réputées également vues : deux politiques qui ne diffèrent que par l'ordre reçoivent\n",
      "alors la même valeur estimée, quoi qu'elles fassent. L'évaluation ne se trompe pas\n",
      "de peu — dans ce cadre, elle est vide par construction.\n"
     ]
    }
   ],
   "source": [
    "ITEMS, IMPRESSIONS = 20, 400_000\n",
    "rng = np.random.default_rng(11)\n",
    "\n",
    "relevance = rng.uniform(0.05, 0.95, ITEMS)\n",
    "diversity = rng.uniform(0.0, 1.0, ITEMS)\n",
    "platform_ranks = np.argsort(np.argsort(-relevance)) + 1\n",
    "target_ranks = np.argsort(np.argsort(-(0.4 * relevance + 0.6 * diversity))) + 1\n",
    "\n",
    "logged = rank_propensities(platform_ranks, TRUE_SEVERITY)\n",
    "target = rank_propensities(target_ranks, TRUE_SEVERITY)\n",
    "examined, clicks_logged = simulate_logged_feedback(relevance, logged, IMPRESSIONS, rng)\n",
    "true_cost = 1 - value_under_policy(relevance, target) / value_under_policy(relevance, logged)\n",
    "\n",
    "\n",
    "def cost_assuming(severity):\n",
    "    assumed_logged = rank_propensities(platform_ranks, severity)\n",
    "    assumed_target = rank_propensities(target_ranks, severity)\n",
    "    return 1 - snips(clicks_logged, assumed_target[examined],\n",
    "                     assumed_logged[examined]) / naive(clicks_logged)\n",
    "\n",
    "\n",
    "print(f\"coût réel du filtre de diversité              : {100 * true_cost:6.2f} %\")\n",
    "print(f\"estimé avec η lu sur l'ordre conservé ({intact.severity:.2f}) : \"\n",
    "      f\"{100 * cost_assuming(intact.severity):6.2f} %\")\n",
    "print(f\"estimé avec η lu sur l'ordre mélangé ({erased.severity:+.2f}) : \"\n",
    "      f\"{100 * cost_assuming(max(erased.severity, 0.0)):6.2f} %\")\n",
    "print(f\"\\nerreur relative après mélange : \"\n",
    "      f\"{100 * (cost_assuming(max(erased.severity, 0.0)) / true_cost - 1):+.0f} %\")\n",
    "print(\"\\nLe zéro n'est pas une coïncidence numérique. Sous η = 0, toutes les positions sont\")\n",
    "print(\"réputées également vues : deux politiques qui ne diffèrent que par l'ordre reçoivent\")\n",
    "print(\"alors la même valeur estimée, quoi qu'elles fassent. L'évaluation ne se trompe pas\")\n",
    "print(\"de peu — dans ce cadre, elle est vide par construction.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "68cb57dd",
   "metadata": {},
   "source": [
    "## 6. La figure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "37bdd19a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2026-08-23T13:41:02.164541Z",
     "iopub.status.busy": "2026-08-23T13:41:02.164468Z",
     "iopub.status.idle": "2026-08-23T13:41:02.623059Z",
     "shell.execute_reply": "2026-08-23T13:41:02.622473Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1488x1008 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "figure, axes = plt.subplots(2, 2, figsize=(12.4, 8.4))\n",
    "\n",
    "# (a) la courbe agrégée et son démenti à longueur fixée\n",
    "ax = axes[0, 0]\n",
    "positions = np.arange(1, 21)\n",
    "ax.plot(positions, aggregate, \"o-\", color=PALETTE[\"disorder\"], markersize=4,\n",
    "        label=f\"tous fils confondus (η = {severity_aggregate:.2f})\")\n",
    "for length, colour in ((10, PALETTE[\"order\"]), (30, PALETTE[\"remedy\"])):\n",
    "    rates = fixed[length][:20]\n",
    "    ax.plot(np.arange(1, rates.size + 1), rates, \"o-\", color=colour, markersize=3,\n",
    "            alpha=0.9, label=f\"fils de longueur {length} exactement\")\n",
    "ax.set_xlabel(\"position dans la liste enregistrée\")\n",
    "ax.set_ylabel(\"taux de clic\")\n",
    "ax.set_title(\"(a) une décroissance qui n'est qu'un mélange de longueurs\")\n",
    "ax.set_xticks([1, 5, 10, 15, 20])\n",
    "ax.legend(loc=\"upper right\", fontsize=8)\n",
    "\n",
    "# (b) puissance du test\n",
    "ax = axes[0, 1]\n",
    "severities = np.array([value for value, _ in power])\n",
    "deviations = np.array([value for _, value in power])\n",
    "ax.plot(severities, -deviations, \"o-\", color=PALETTE[\"order\"], markersize=4,\n",
    "        label=\"journaux simulés, structure de MIND\")\n",
    "ax.axhline(1.96, color=PALETTE[\"neutral\"], linestyle=\":\", linewidth=1.2)\n",
    "ax.text(0.62, 2.6, \"seuil de rejet à 5 %\", fontsize=8, color=PALETTE[\"neutral\"])\n",
    "ax.plot([0.0], [-verdicts[\"train\"].deviation], \"*\", color=PALETTE[\"disorder\"], markersize=16,\n",
    "        zorder=5, label=f\"MIND (z = {verdicts['train'].deviation:+.2f})\")\n",
    "ax.annotate(f\"MIND : η indétectable\\nau-delà de {threshold:.3f}\", xy=(0.0, 0.0),\n",
    "            xytext=(0.16, -0.6), fontsize=8, color=PALETTE[\"disorder\"],\n",
    "            arrowprops={\"arrowstyle\": \"->\", \"color\": PALETTE[\"disorder\"], \"linewidth\": 1.0})\n",
    "ax.set_yscale(\"symlog\", linthresh=10)\n",
    "ax.set_xlabel(\"sévérité $\\\\eta$ du biais de position simulé\")\n",
    "ax.set_ylabel(\"$-z$ du test d'échangeabilité\")\n",
    "ax.set_title(\"(b) ce que le test aurait su détecter\")\n",
    "ax.legend(loc=\"lower right\", fontsize=8)\n",
    "\n",
    "# (c) les cinq chiffres tirés du même jeu\n",
    "ax = axes[1, 0]\n",
    "labels = [\"ajustement naïf\\nagrégé\", \"seuil 5\", \"seuil 20\", \"seuil 50\",\n",
    "          \"test exact\\n(échangeabilité)\"]\n",
    "values = [severity_aggregate, estimates[5].severity, estimates[20].severity,\n",
    "          estimates[50].severity, 0.0]\n",
    "errors = [0.0, estimates[5].standard_error, estimates[20].standard_error,\n",
    "          estimates[50].standard_error, threshold]\n",
    "colours = [PALETTE[\"disorder\"]] * 4 + [PALETTE[\"remedy\"]]\n",
    "ax.barh(np.arange(5), values, xerr=errors, color=colours, alpha=0.85, height=0.6,\n",
    "        error_kw={\"ecolor\": PALETTE[\"neutral\"], \"capsize\": 3})\n",
    "ax.axvline(0.0, color=PALETTE[\"neutral\"], linewidth=1.0)\n",
    "ax.set_yticks(np.arange(5))\n",
    "ax.set_yticklabels(labels, fontsize=8)\n",
    "ax.invert_yaxis()\n",
    "ax.set_xlabel(\"sévérité $\\\\hat\\\\eta$ estimée\")\n",
    "ax.set_title(\"(c) cinq estimations du même jeu, quatre de trop\")\n",
    "\n",
    "# (d) le rang effacé\n",
    "ax = axes[1, 1]\n",
    "names = [\"coût réel\", \"η lu sur\\nl'ordre conservé\", \"η lu sur\\nl'ordre mélangé\"]\n",
    "costs = [100 * true_cost, 100 * cost_assuming(intact.severity),\n",
    "         100 * cost_assuming(max(erased.severity, 0.0))]\n",
    "ax.bar(names, costs, color=[PALETTE[\"neutral\"], PALETTE[\"remedy\"], PALETTE[\"disorder\"]],\n",
    "       alpha=0.85, width=0.55)\n",
    "for index, value in enumerate(costs):\n",
    "    ax.text(index, value + 0.4, f\"{value:.1f} %\", ha=\"center\", fontsize=9)\n",
    "ax.plot([2 - 0.275, 2 + 0.275], [0, 0], color=PALETTE[\"disorder\"], linewidth=3)\n",
    "ax.annotate(\"évaluation vide :\\nl'ordre ne compte plus,\\ndonc le réordonnancement\\nne coûte rien\",\n",
    "            xy=(2, 0.15), xytext=(1.55, 2.4), fontsize=8, color=PALETTE[\"disorder\"],\n",
    "            arrowprops={\"arrowstyle\": \"->\", \"color\": PALETTE[\"disorder\"], \"linewidth\": 1.0})\n",
    "ax.set_ylabel(\"coût d'engagement estimé du filtre de diversité (%)\")\n",
    "ax.set_title(\"(d) ce que coûte la variable détruite\")\n",
    "ax.set_ylim(0, max(costs) * 1.25)\n",
    "ax.tick_params(axis=\"x\", labelsize=8)\n",
    "\n",
    "save_figure(figure, \"fig16_exploration_mind\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b442fbb3",
   "metadata": {},
   "source": [
    "## 7. Ce que cette mesure décide\n",
    "\n",
    "**MIND ne peut pas calibrer $\\eta$.** Ce n'est pas un défaut de la méthode ni un manque de\n",
    "données — 5,8 millions de contenus servis, 156 965 fils — mais l'absence de la seule variable\n",
    "qui identifierait la sévérité. Aucun raffinement de l'estimateur n'y changera rien.\n",
    "\n",
    "**Trois conséquences, par ordre de portée.**\n",
    "\n",
    "*Pour ce dépôt.* L'évaluation de l'ADE sur MIND reste possible pour tout ce qui ne dépend pas de\n",
    "l'exposition — composition des fils, diversité servie, coût en pertinence *déclarée*. Elle est\n",
    "**impossible** pour ce qui en dépend, c'est-à-dire l'estimation contrefactuelle du coût\n",
    "d'engagement, qui était l'objet même de l'exercice. Il faut soit un jeu qui enregistre le rang\n",
    "d'affichage, soit assumer un $\\eta$ importé — et le notebook 15 a chiffré ce que cela coûte.\n",
    "\n",
    "*Pour quiconque évalue un réordonnancement sur données publiques.* Le contrôle à faire n'est pas\n",
    "« ai-je assez de variation de rang ? » mais « cette variation vient-elle de la plateforme ou de\n",
    "l'anonymisation ? ». Les deux se ressemblent parfaitement du point de vue de l'estimateur, et\n",
    "seule la seconde produit des chiffres confiants et faux.\n",
    "\n",
    "*Pour ceux qui publient des journaux.* Mélanger l'ordre d'affichage ne rend pas un jeu de\n",
    "données non biaisé : il le rend **non corrigible**. Publier le rang servi, ou à défaut la\n",
    "propension d'exposition, coûte une colonne et décide de ce qui reste mesurable."
   ]
  }
 ],
 "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
}