visualize_advantage.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.colors as mcolors\n",
    "from matplotlib.cm import ScalarMappable\n",
    "import textwrap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"output/testset_baseline_mistral_large_chunksize256.jsonl\") as f:\n",
    "    lines = f.readlines()\n",
    "    baseline_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "with open(\"output/testset_summary_only_mistral_large_chunksize256.jsonl\") as f:\n",
    "    lines = f.readlines()\n",
    "    summarization_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "with open(\"output/testset_summary_union_mistral_large_chunksize256.jsonl\") as f:\n",
    "    lines = f.readlines()\n",
    "    s_plus_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "with open(\"output/testset_questions_mistral_large_chunksize256.jsonl\") as f:\n",
    "    lines = f.readlines()\n",
    "    question_generation_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "with open(\"output/testset_question_union_mistral_large_chunksize256.jsonl\") as f:\n",
    "    lines = f.readlines()\n",
    "    q_plus_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "with open(\"output/testset_union_mistral_large_chunksize256.jsonl\") as f:\n",
    "    lines = f.readlines()\n",
    "    union_outputs = [json.loads(line) for line in lines]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 3200x800 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "correct_mat = np.array(\n",
    "    [\n",
    "        [output[\"correct\"] for output in union_outputs],\n",
    "        [output[\"correct\"] for output in q_plus_outputs],\n",
    "        [output[\"correct\"] for output in question_generation_outputs],\n",
    "        [output[\"correct\"] for output in s_plus_outputs],\n",
    "        [output[\"correct\"] for output in summarization_outputs],\n",
    "        [output[\"correct\"] for output in baseline_outputs],\n",
    "    ]\n",
    ").T\n",
    "\n",
    "\n",
    "def sort_rows(array):\n",
    "    order = []\n",
    "    baseline_correct = [i for i, correct in enumerate(array[:, -1]) if correct]\n",
    "    order += baseline_correct\n",
    "\n",
    "    incorrect_2 = [i for i, correct in enumerate(array[:, -2]) if not correct and array[i, -1]]\n",
    "    for i in incorrect_2:\n",
    "        if i in order:\n",
    "            order.remove(i)\n",
    "    order += incorrect_2\n",
    "\n",
    "    correct_2 = [i for i, correct in enumerate(array[:, -2]) if correct and i not in order]\n",
    "    order += correct_2\n",
    "\n",
    "    correct_3 = [i for i, correct in enumerate(array[:, -3]) if correct and i not in order]\n",
    "    order += correct_3\n",
    "\n",
    "    correct_4 = [i for i, correct in enumerate(array[:, -4]) if correct and i not in order]\n",
    "    order += correct_4\n",
    "\n",
    "    correct_5 = [i for i, correct in enumerate(array[:, -5]) if correct and i not in order]\n",
    "    order += correct_5\n",
    "\n",
    "    correct_6 = [i for i, correct in enumerate(array[:, -6]) if correct and i not in order]\n",
    "    order += correct_6\n",
    "\n",
    "    # incorrect = [i for i in range(array.shape[0]) if i not in order]\n",
    "    # order += incorrect\n",
    "    return order\n",
    "\n",
    "\n",
    "row_order = sort_rows(correct_mat)\n",
    "correct_mat = correct_mat[row_order]\n",
    "\n",
    "correct_mat_colorful = np.zeros(correct_mat.shape + (3,))\n",
    "for question_num in range(correct_mat.shape[0]):\n",
    "    for model_num in range(correct_mat.shape[1]):\n",
    "        if correct_mat[question_num, model_num] == correct_mat[question_num, -1]:\n",
    "            if correct_mat[question_num, model_num] == True:\n",
    "                correct_mat_colorful[question_num, model_num] = [129 / 255, 206 / 255, 255 / 255]\n",
    "            else:\n",
    "                correct_mat_colorful[question_num, model_num] = [249 / 255, 248 / 255, 255 / 255]\n",
    "        elif correct_mat[question_num, model_num] == True:\n",
    "            correct_mat_colorful[question_num, model_num] = [134 / 255, 255 / 255, 73 / 255]\n",
    "        else:\n",
    "            correct_mat_colorful[question_num, model_num] = [255 / 255, 93 / 255, 93 / 255]\n",
    "\n",
    "correct_mat_colorful = correct_mat_colorful.transpose((1, 0, 2))\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "\n",
    "# Make the plot wider than the default\n",
    "fig.set_size_inches(32, 8)\n",
    "# Make edge colors transparent\n",
    "ax.pcolormesh(correct_mat_colorful, edgecolors=[252 / 255, 252 / 255, 252 / 255], linewidth=1)\n",
    "# Set the y ticks\n",
    "ticks = [\n",
    "    f\"RAIDD-U: {sum(correct_mat[:, 0])}%\",\n",
    "    f\"RAIDD-Q+: {sum(correct_mat[:, 1])}%\",\n",
    "    f\"RAIDD-Q: {sum(correct_mat[:, 2])}%\",\n",
    "    f\"RAIDD-S+: {sum(correct_mat[:, 3])}%\",\n",
    "    f\"RAIDD-S: {sum(correct_mat[:, 4])}%\",\n",
    "    f\"RAG Baseline: {sum(correct_mat[:, 5])}%\",\n",
    "]\n",
    "ax.set_yticks(np.arange(0.5, len(ticks), 1), ticks)\n",
    "plt.yticks(fontsize=32)\n",
    "ax.set_xticks([], [])\n",
    "# Remove borders\n",
    "ax.spines[\"top\"].set_visible(False)\n",
    "ax.spines[\"right\"].set_visible(False)\n",
    "ax.spines[\"left\"].set_visible(False)\n",
    "ax.spines[\"bottom\"].set_visible(False)\n",
    "# Add background color to the plot\n",
    "ax.set_facecolor([0.95, 0, 0.95])\n",
    "# Set the x and y axis labels\n",
    "ax.set_xlabel(\"Question\", fontsize=32)\n",
    "\n",
    "\n",
    "# Add legend\n",
    "cmap = mcolors.ListedColormap(\n",
    "    [[255 / 255, 93 / 255, 93 / 255], [134 / 255, 255 / 255, 73 / 255], [129 / 255, 206 / 255, 255 / 255]]\n",
    ")\n",
    "bounds = [0, 1, 2, 3]\n",
    "norm = mcolors.BoundaryNorm(bounds, cmap.N)\n",
    "sm = ScalarMappable(cmap=cmap, norm=norm)\n",
    "sm.set_array([])\n",
    "cbar = fig.colorbar(sm, ax=ax, boundaries=bounds, ticks=[0.5, 1.5, 2.5], spacing=\"proportional\")\n",
    "cbar.set_ticklabels([\"Worse than baseline\", \"Better than baseline\", \"Same as baseline\"], fontsize=32)\n",
    "# remove minor ticks\n",
    "cbar.ax.yaxis.set_tick_params(which=\"minor\", length=0)\n",
    "# Remove the colorbar border\n",
    "cbar.outline.set_visible(False)\n",
    "\n",
    "# Make background transparent\n",
    "fig.patch.set_alpha(0)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/confusion_comparison.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "incorrect_baseline_indices = [i for i, output in enumerate(baseline_outputs) if output[\"correct\"] == False]\n",
    "correct_summarization_indices = [i for i, output in enumerate(summarization_outputs) if output[\"correct\"] == True]\n",
    "correct_question_generation_indices = [i for i, output in enumerate(question_generation_outputs) if output[\"correct\"] == True]\n",
    "correct_union_indices = [i for i, output in enumerate(union_outputs) if output[\"correct\"] == True]\n",
    "\n",
    "union_advantage_indices = set(correct_summarization_indices) & set(incorrect_baseline_indices)\n",
    "union_advantage_indices = list(union_advantage_indices)\n",
    "union_advantage_indices.sort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def print_multiple_lines(text, width=100):\n",
    "    lines = textwrap.wrap(text.strip(), width)\n",
    "    for line in lines:\n",
    "        print(line)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Question:\n",
      "How many years did Picardo work for Parador from his first of Parador projects until the bankruptcy on Parador?\n",
      "\n",
      "\n",
      "Ground truth:\n",
      "Twenty-nine years.\n",
      "\n",
      "\n",
      "Baseline answer:\n",
      "Picardo worked for Parador from the early 1960s to 1985, which is approximately 25 years. However,\n",
      "the question mentions a bankruptcy on Parador, which is not mentioned in the given context.\n",
      "Therefore, I cannot provide a precise answer based on the given information.\n",
      "\n",
      "\n",
      "Baseline evidence:\n",
      "Despite these further works, the Parador has kept its original character as established by Picardo.\n",
      "Most of the improvements have been only to modernise the services and facilities. Other Parador\n",
      "projects. In the 1960s and 1970s Picardo was called upon by the Ministry of Information and Tourism\n",
      "to investigate and report on a number of other old buildings for possible conversion into Paradores.\n",
      "He drew up proposals and plans for a number of these buildings but, despite detailed work on some of\n",
      "them, they did not become Paradores within his working life or were completed by other architects.\n",
      "Picardo was also asked to review proposed works for similar buildings to be restored by other\n",
      "architects, and to develop ideas for improvements to existing Paradores.. Among the most advanced\n",
      "plans Picardo drew up were in 1969 for the renovation and conversion into a Parador of the castle at\n",
      "Puebla de Alcocer, a small municipality 70 miles east of Mérida in the Province of Badajoz in\n",
      "Extremadura. His draft plans show that a multitude of openings would have to have been made in the\n",
      "outside walls for windows.\n",
      "The book is long out of print and virtually unknown in Spain, and not at all elsewhere, but is\n",
      "available second-hand. Paradores de Turismo. From the early 1960s to 1985 Picardo dedicated much of\n",
      "his professional life to the state-run hotel chain, Paradores de Turismo de España. He had for some\n",
      "time carried out minor work for the Ministerio de Información y Turismo which controlled the hotel\n",
      "network. For the purposes of tourism the Ministry and its forebears had for over 30 years\n",
      "rehabilitated rundown and sometimes ruined historic buildings such as castles and convents and\n",
      "converted them into luxury hotels in a style that went beyond ordinary hotel use.\n",
      "He was to employ Arroyo repeatedly in his Parador projects over the following twenty years and as a\n",
      "result the ceramicist's work became an identifying feature of Picardo's work.Restoration at\n",
      "Guadalupe started in November 1963 and the hotel, with twenty double rooms, opened on 11 December\n",
      "1965.In 1981 Picardo was invited back to Guadalupe to add a new wing of guest rooms. He built them\n",
      "in understated but similar style to the rest of the Parador, providing views of the rooftops and\n",
      "towers of the monastery and of the surrounding mountains. The new wing increased the number of guest\n",
      "rooms to 41. As a result of being an afterthought to the original design, access to the new wing was\n",
      "complicated and required an abundance of staircases and lifts. Parador de Jaén: Castillo de Santa\n",
      "Catalina. At the same time as preparing his restoration at Guadalupe, Picardo was commissioned by\n",
      "the Ministry to design and build a Parador at the Castillo de Santa Catalina (Castle of Santa\n",
      "Catalina) in Jaén in Andalucía.\n",
      "For the purposes of tourism the Ministry and its forebears had for over 30 years rehabilitated\n",
      "rundown and sometimes ruined historic buildings such as castles and convents and converted them into\n",
      "luxury hotels in a style that went beyond ordinary hotel use. In the early 1960s, as Spanish tourism\n",
      "increased, the Ministry decided to rapidly expand its Parador operation (which would within a decade\n",
      "grow from 40 to 83 establishments) and Picardo, with his previous experience of historical\n",
      "restoration and his abiding interest in historical and vernacular buildings, was seen by the\n",
      "Ministry be a suitable architect to take on much of this type of work.. Picardo began working for\n",
      "Paradores on a series of restorations of old, monumental buildings and sometimes building new\n",
      "establishments adjacent to ruined monuments in a style that faithfully copied their original\n",
      "designs. His hybrid conversions maintained and often embellished the monuments' ancient appearance\n",
      "while at the same time finding inspiration in them for the style of luxurious modern hotel\n",
      "arrangements the authorities required.. A wealth of Picardo's drawings for his Paradores projects\n",
      "survive. There are large collections of extensively detailed plans which cover his designs from\n",
      "whole Paradores to the smallest detail of door furniture.\n",
      "In 1975 he also developed improvements for one of the earliest existing Paradores, opened in 1929,\n",
      "the castle at Ciudad Rodrigo in the Province of Salamanca in Castilla y León. Controversial legacy\n",
      "of Picardo's Paradores. It was not unusual in Spain in the 1960s and 1970s for the rehabilitation of\n",
      "castles and convents (not all destined to be Paradores) to be carried out without archeological\n",
      "research either before work began, which would have added to expense and delayed the project, or\n",
      "while work was being carried out. Instead, Picardo's rebuilding projects were planned mostly on the\n",
      "basis of his own historical and architectural research. The hotel conversions and the demolition of\n",
      "large parts of monumental buildings without detailed investigation and record-keeping was somewhat\n",
      "frowned upon in the 1960s and 1970s, and over half a century later is seen by archeologists and\n",
      "historians as a matter of significant controversy and regret.\n",
      "There are large collections of extensively detailed plans which cover his designs from whole\n",
      "Paradores to the smallest detail of door furniture. There are axonometric before-and-after drawings\n",
      "of the buildings and the landscapes around them. There are bird's eye views exercising his mastery\n",
      "of perspective and his spatial vision. They all show meticulous skill.For nearly twenty years, from\n",
      "the early 1960s to his last work for the Paradores in the 1980s, Picardo carried out eleven major\n",
      "reconstructions of historical buildings and/or erected sympathetic and imitative new constructions\n",
      "abutting them or rising from their ruined foundations. With a number he returned to build additions\n",
      "to his earlier work. He also worked on a number of other Parador projects which for various reasons\n",
      "did not reach fruition. His eleven Parador masterworks encouraged other Spanish architects to work\n",
      "in the same vein, and Portuguese architects, too, in the similar state-run chain of hotels in\n",
      "Portugal, the Pousadas de Portugal. Picardo's work for Paradores de Turismo is highly regarded by\n",
      "other professionals, and also by hotel guests who revel in the historical imagery and romance of his\n",
      "work. Parador de Guadalupe: Zurbarán.\n",
      "Neither idea was acceptable to him.Using the elongated site at the top of the hill, Picardo planned\n",
      "a dining room, a lounge, service accommodation and guest rooms. He styled his new building on the\n",
      "layout and dimensions of the old castle and on what had been discovered during his research of its\n",
      "surviving interior designs. Work started in 1963, and the Parador opened to guests on 11 September\n",
      "1965.The first phase, built only as a hostería, had on the first floor 7 double guest rooms with\n",
      "fireplaces and with wooden balconies of a design that Picardo would repeat in a number of his later\n",
      "Parador designs. A mezzanine floor housed a cafetería and a bar with an outside terrace, and on the\n",
      "ground floor was the reception area, the lounge and the restaurant, together with the service areas.\n",
      "There were also four single rooms for drivers, and a mechanical workshop.Picardo's Parador at Jaén\n",
      "was a pastiche, which paid homage to the neighbouring castle.\n",
      "Its influence, particularly in his work on Paradores, can be clearly seen.. During the 1950s Picardo\n",
      "pursued his personal architectural ambitions, entering competitions and publishing plans and\n",
      "drawings of uncommissioned buildings. In 1951 in  company with his fellow architect Carlos de Miguel\n",
      "he designed a centre for the Cofradía de Pescadores (Fishermen's Brotherhood) of Altea in the\n",
      "province of Alicante which attracted much attention but remained unbuilt.Picardo, working alone,\n",
      "designed a small hotel which could be built on the Costa del Sol in Andalucía. He described it as\n",
      "\"un modesto parador (a modest inn), resolved in a simple and attractive way\".In competition, Picardo\n",
      "entered plans and drawings of a preliminary draft for the Delegación de Hacienda de Gerona (Treasury\n",
      "Delegation in Gerona) but came second behind the Spanish architect Carlos Sobrini who had been a\n",
      "year behind Picardo at college.\n",
      "\n",
      "\n",
      "RAIDD answer:\n",
      "Picardo worked for Parador from his first of Parador projects in July 1963 until the bankruptcy on\n",
      "Parador in 1992, which is a total of 29 years.\n",
      "\n",
      "\n",
      "RAIDD evidence:\n",
      "There are large collections of extensively detailed plans which cover his designs from whole\n",
      "Paradores to the smallest detail of door furniture. There are axonometric before-and-after drawings\n",
      "of the buildings and the landscapes around them. There are bird's eye views exercising his mastery\n",
      "of perspective and his spatial vision. They all show meticulous skill.For nearly twenty years, from\n",
      "the early 1960s to his last work for the Paradores in the 1980s, Picardo carried out eleven major\n",
      "reconstructions of historical buildings and/or erected sympathetic and imitative new constructions\n",
      "abutting them or rising from their ruined foundations. With a number he returned to build additions\n",
      "to his earlier work. He also worked on a number of other Parador projects which for various reasons\n",
      "did not reach fruition. His eleven Parador masterworks encouraged other Spanish architects to work\n",
      "in the same vein, and Portuguese architects, too, in the similar state-run chain of hotels in\n",
      "Portugal, the Pousadas de Portugal. Picardo's work for Paradores de Turismo is highly regarded by\n",
      "other professionals, and also by hotel guests who revel in the historical imagery and romance of his\n",
      "work. Parador de Guadalupe: Zurbarán.\n",
      "His plans offered the prospect of 16 guest rooms and in November 1969 the proposal to convert the\n",
      "Hostería into a Parador was made public. However, difficulties in purchasing the neighbouring\n",
      "property made the project impossible.The Pedraza Hostería continued in operation until 15 December\n",
      "1992 when economic pressures on the Parador chain caused its closure. Parador de Alcañiz: La\n",
      "Concordia. In 1966 Picardo began the conversion into a Parador of the Palacio de los Comendadores at\n",
      "Alcañiz in the province of Teruel in Aragon. The palace - the façade of which was remodelled in\n",
      "late-Renaissance style in 1728 - stood as the most prominent additional part of the Castillo de los\n",
      "Calatravos (Castle of the Calatravos), a monastery-fortress built in 1179.\n",
      "Ninguno de ells la ha vista. Qué dirán?\" (None of them has seen it. What will they say?). At Jaen,\n",
      "and at Guadalupe, finished at much the same time, Picardo established a style of architecture and\n",
      "interior design which found favour with his clients and their guests and which he was to pursue in\n",
      "most of his further work for Paradores, refining it where required and elsewhere repeating it\n",
      "faithfully. Parador de Arcos de la Frontera: Casa del Corregidor. The Parador at Arcos de la\n",
      "Frontera in the province of Cádiz in Andalucía is located in the centre of the old town, at the top\n",
      "of the cliffs that overhang the Rio Guadalete. Picardo first visited the triangular site in February\n",
      "1964.\n",
      "For the purposes of tourism the Ministry and its forebears had for over 30 years rehabilitated\n",
      "rundown and sometimes ruined historic buildings such as castles and convents and converted them into\n",
      "luxury hotels in a style that went beyond ordinary hotel use. In the early 1960s, as Spanish tourism\n",
      "increased, the Ministry decided to rapidly expand its Parador operation (which would within a decade\n",
      "grow from 40 to 83 establishments) and Picardo, with his previous experience of historical\n",
      "restoration and his abiding interest in historical and vernacular buildings, was seen by the\n",
      "Ministry be a suitable architect to take on much of this type of work.. Picardo began working for\n",
      "Paradores on a series of restorations of old, monumental buildings and sometimes building new\n",
      "establishments adjacent to ruined monuments in a style that faithfully copied their original\n",
      "designs. His hybrid conversions maintained and often embellished the monuments' ancient appearance\n",
      "while at the same time finding inspiration in them for the style of luxurious modern hotel\n",
      "arrangements the authorities required.. A wealth of Picardo's drawings for his Paradores projects\n",
      "survive. There are large collections of extensively detailed plans which cover his designs from\n",
      "whole Paradores to the smallest detail of door furniture.\n",
      "At that point the rest of the palace was, as Picardo had recommended, restored and converted and,\n",
      "with his original Hostería, opened as a full Parador with 27 guest rooms on 10 October 1989. Picardo\n",
      "was not involved in this work.. Subsequent expansions into neighbouring buildings have turned the\n",
      "Parador into a much larger establishment. Picardo's original entrance, bar and cafetería area now\n",
      "form a sumptuous suite, though the medieval aura of his interior decoration and furnishings for that\n",
      "part of the building has been lost through modernisation. Parador de Carmona: Alcazar del Rey Don\n",
      "Pedro. In 1966, while building the Parador at Arcos de la Frontera, Picardo was commissioned to\n",
      "inspect three ancient sites near the city of Sevilla in the province of the same name in Andalucía\n",
      "with a view to constructing another Parador. After looking at the castle at Alcalá de Guadaíra, and\n",
      "the palaces of Écija, he came across the ruined castle of Carmona the Alcázar del Rey Don Pedro\n",
      "(also known as the Alcázar de Arriba y Puerta de Marchena).\n",
      "In 1987 an expansion was carried out by the architect Carlos Fernández-Cuenca Gómez which included\n",
      "additional guest rooms, and a new pavilion for conventions. He rigidly copied Picardo's style.. The\n",
      "Carmona Parador now has 9 double rooms, 51 twin rooms and 3 single rooms, making a total guest\n",
      "complement of 123. Parador de Sigüenza: Castillo de Sigüenza. In 1964 Picardo was involved, with the\n",
      "Ministry of Information and Tourism, in investigating old buildings for conversion into a new\n",
      "Parador in the Province of Guadalajara. Possible locations were the castle at Atienza and the Casa\n",
      "del Cordón, an old inn in the same town, the castle at Molina de Aragón and the castle at Sigüenza.\n",
      "He considered the last to be the best proposition despite it being comprehensively ruined. It stands\n",
      "prominently above the town and cathedral of Sigüenza and dominates the landscape. The Ministry set\n",
      "about acquiring it the same year.\n",
      "Picardo's work for Paradores de Turismo is highly regarded by other professionals, and also by hotel\n",
      "guests who revel in the historical imagery and romance of his work. Parador de Guadalupe: Zurbarán.\n",
      "For his first of many Parador projects Picardo was appointed by the Ministry of Information and\n",
      "Tourism in July 1963 to convert into a Parador two ancient neighbouring buildings in the village of\n",
      "Guadalupe in the province of Cáceres in Extremadura. One building was the Hospital de San Juan\n",
      "Bautista, also known as the Hospital de Hombres, which was built in the mid-14th-century, rebuilt in\n",
      "1402 and refurbished in the 16th century. The other building was the Colegio de Infantes, also known\n",
      "as the Colegio de Gramática, built in the early 16th-century for the education of  boys. They were\n",
      "situated close to the Monastery of Santa María de Guadalupe, one of the most important monasteries\n",
      "of medieval Spain, in the centre of the village.\n",
      "In 1975 he also developed improvements for one of the earliest existing Paradores, opened in 1929,\n",
      "the castle at Ciudad Rodrigo in the Province of Salamanca in Castilla y León. Controversial legacy\n",
      "of Picardo's Paradores. It was not unusual in Spain in the 1960s and 1970s for the rehabilitation of\n",
      "castles and convents (not all destined to be Paradores) to be carried out without archeological\n",
      "research either before work began, which would have added to expense and delayed the project, or\n",
      "while work was being carried out. Instead, Picardo's rebuilding projects were planned mostly on the\n",
      "basis of his own historical and architectural research. The hotel conversions and the demolition of\n",
      "large parts of monumental buildings without detailed investigation and record-keeping was somewhat\n",
      "frowned upon in the 1960s and 1970s, and over half a century later is seen by archeologists and\n",
      "historians as a matter of significant controversy and regret.\n"
     ]
    }
   ],
   "source": [
    "i = 0\n",
    "question_index = union_advantage_indices[i]\n",
    "\n",
    "print(\"Question:\")\n",
    "print(union_outputs[question_index][\"question\"])\n",
    "\n",
    "print(\"\\n\")\n",
    "print(\"Ground truth:\")\n",
    "print(union_outputs[question_index][\"ground_truth\"])\n",
    "\n",
    "print(\"\\n\")\n",
    "print(\"Baseline answer:\")\n",
    "print_multiple_lines(baseline_outputs[question_index][\"generated_answer\"])\n",
    "\n",
    "print(\"\\n\")\n",
    "print(\"Baseline evidence:\")\n",
    "for evidence in baseline_outputs[question_index][\"additional_info\"][\"top_chunks_text\"]:\n",
    "    print_multiple_lines(evidence)\n",
    "\n",
    "print(\"\\n\")\n",
    "print(\"RAIDD answer:\")\n",
    "print_multiple_lines(union_outputs[question_index][\"generated_answer\"])\n",
    "\n",
    "print(\"\\n\")\n",
    "print(\"RAIDD evidence:\")\n",
    "for evidence in union_outputs[question_index][\"additional_info\"][\"top_chunks_info\"]:\n",
    "    print_multiple_lines(evidence[\"text\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "from rag_components import *\n",
    "\n",
    "from llama_index.core.retrievers import VectorIndexRetriever"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_rank_of_text(query, text, index_title):\n",
    "    index = get_index_by_title(index_title)\n",
    "    top_k = 1024\n",
    "    retriever = VectorIndexRetriever(\n",
    "        index=index,\n",
    "        similarity_top_k=top_k,\n",
    "    )\n",
    "    top_nodes = retriever.retrieve(query)\n",
    "    for i, node in enumerate(top_nodes):\n",
    "        if node.text == text:\n",
    "            print(i + 1)\n",
    "            return i + 1\n",
    "    return None\n",
    "\n",
    "\n",
    "def get_index_name_from_question(question):\n",
    "    for environment in longdep_qa_ds:\n",
    "        title = environment[\"title\"]\n",
    "        qa_pairs = eval(environment[\"qa_pairs\"])\n",
    "        for question_dict in qa_pairs:\n",
    "            q = question_dict[\"Q\"]\n",
    "            if q == question:\n",
    "                return title\n",
    "    return None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"s_plus_rank_pairs.txt\", \"r\") as f:\n",
    "    s_plus_rank_pairs = f.readlines()\n",
    "    s_plus_rank_pairs = [[int(rank) for rank in pair.strip().split(\" \")] for pair in s_plus_rank_pairs]\n",
    "    s_plus_rank_pairs = np.array(s_plus_rank_pairs)\n",
    "with open(\"q_plus_rank_pairs.txt\", \"r\") as f:\n",
    "    q_plus_rank_pairs = f.readlines()\n",
    "    q_plus_rank_pairs = [[int(rank) for rank in pair.strip().split(\" \")] for pair in q_plus_rank_pairs]\n",
    "    q_plus_rank_pairs = np.array(q_plus_rank_pairs)\n",
    "with open(\"u_rank_pairs.txt\", \"r\") as f:\n",
    "    u_rank_pairs = f.readlines()\n",
    "    u_rank_pairs = [[int(rank) for rank in pair.strip().split(\" \")] for pair in u_rank_pairs]\n",
    "    u_rank_pairs = np.array(u_rank_pairs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams.update(\n",
    "    {\n",
    "        \"figure.facecolor\": (0.0, 0.0, 0.0, 0.0),  # red   with alpha = 30%\n",
    "        \"axes.facecolor\": (250 / 255, 250 / 255, 255 / 255, 1.0),  # green with alpha = 50%\n",
    "        \"savefig.facecolor\": (0.0, 0.0, 0.0, 0.0),  # blue  with alpha = 20%\n",
    "    }\n",
    ")\n",
    "\n",
    "plt.figure(figsize=(6, 6))\n",
    "plt.hist(s_plus_rank_pairs[:, 1], bins=100)\n",
    "plt.title(\"Rank of Documents Used in RAIDD-S\", fontsize=16)\n",
    "plt.xlabel(\"Rank\", fontsize=16)\n",
    "plt.ylabel(\"Frequency\", fontsize=16)\n",
    "plt.xticks(fontsize=16)\n",
    "plt.yticks(fontsize=16)\n",
    "plt.xlim([0, 150])\n",
    "plt.ylim([0, 200])\n",
    "plt.gca().spines[\"top\"].set_visible(False)\n",
    "plt.gca().spines[\"right\"].set_visible(False)\n",
    "plt.gca().spines[\"left\"].set_visible(False)\n",
    "plt.gca().spines[\"bottom\"].set_visible(False)\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/rank_histogram_s_plus.png\")\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=(6, 6))\n",
    "plt.hist(q_plus_rank_pairs[:, 1], bins=100)\n",
    "plt.title(\"Rank of Documents Used in RAIDD-Q\", fontsize=16)\n",
    "plt.xlabel(\"Rank\", fontsize=16)\n",
    "plt.ylabel(\"Frequency\", fontsize=16)\n",
    "plt.xticks(fontsize=16)\n",
    "plt.yticks(fontsize=16)\n",
    "plt.xlim([0, 150])\n",
    "plt.ylim([0, 200])\n",
    "plt.gca().spines[\"top\"].set_visible(False)\n",
    "plt.gca().spines[\"right\"].set_visible(False)\n",
    "plt.gca().spines[\"left\"].set_visible(False)\n",
    "plt.gca().spines[\"bottom\"].set_visible(False)\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/rank_histogram_q_plus.png\")\n",
    "plt.show()\n",
    "\n",
    "plt.figure(figsize=(6, 6))\n",
    "plt.hist(u_rank_pairs[:, 1], bins=100)\n",
    "plt.title(\"Rank of Documents Used in RAIDD-U\", fontsize=16)\n",
    "plt.xlabel(\"Rank\", fontsize=16)\n",
    "plt.ylabel(\"Frequency\", fontsize=16)\n",
    "plt.xticks(fontsize=16)\n",
    "plt.yticks(fontsize=16)\n",
    "plt.xlim([0, 150])\n",
    "plt.ylim([0, 200])\n",
    "plt.gca().spines[\"top\"].set_visible(False)\n",
    "plt.gca().spines[\"right\"].set_visible(False)\n",
    "plt.gca().spines[\"left\"].set_visible(False)\n",
    "plt.gca().spines[\"bottom\"].set_visible(False)\n",
    "plt.tight_layout()\n",
    "plt.savefig(\"figures/rank_histogram_u.png\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/isaac/miniforge3/envs/LongRAG/lib/python3.10/site-packages/datasets/load.py:1461: FutureWarning: The repository for bigainlco/LooGLE contains custom code which must be executed to correctly load the dataset. You can inspect the repository content at https://hf.co/datasets/bigainlco/LooGLE\n",
      "You can avoid this message in future by passing the argument `trust_remote_code=True`.\n",
      "Passing `trust_remote_code=True` will be mandatory to load this dataset from the next major release of `datasets`.\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "from rag_components import *\n",
    "import difflib"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def soft_contains_context(text, gt_sequences):\n",
    "    numer = 0\n",
    "    denom = 0\n",
    "    contains = []\n",
    "    for sequence in gt_sequences:\n",
    "        l = difflib.SequenceMatcher(None, text, sequence).find_longest_match().size\n",
    "        numer += l\n",
    "        denom += len(sequence)\n",
    "        contains.append(l > len(sequence) * 0.5)\n",
    "\n",
    "    return np.mean(contains)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.2786111111111111\n",
      "0.33558212560386474\n",
      "0.34930434782608694\n"
     ]
    }
   ],
   "source": [
    "questions = [\n",
    "    eval(longdep_qa_ds[i][\"qa_pairs\"])[j]\n",
    "    for i in range(len(longdep_qa_ds))\n",
    "    for j in range(len(eval(longdep_qa_ds[i][\"qa_pairs\"])))\n",
    "]\n",
    "question_indices = list(range(105, 205))\n",
    "\n",
    "\n",
    "with open(\"output/testset_summary_only_mistral_large_chunksize256.jsonl\", \"r\") as f:\n",
    "    lines = f.readlines()\n",
    "    s_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "retrieval_scores = []\n",
    "for question_num in question_indices:\n",
    "    i = question_num - 105\n",
    "    retrieved_text = \" \".join([retrieved[\"text\"] for retrieved in s_outputs[i][\"additional_info\"][\"top_chunks_info\"]])\n",
    "    gt_sequences = questions[question_num][\"S\"]\n",
    "    retrieval_scores.append(soft_contains_context(retrieved_text, gt_sequences))\n",
    "print(np.mean(retrieval_scores))\n",
    "\n",
    "with open(\"output/testset_summary_union_mistral_large_chunksize256.jsonl\", \"r\") as f:\n",
    "    lines = f.readlines()\n",
    "    s_plus_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "retrieval_scores = []\n",
    "for question_num in question_indices:\n",
    "    i = question_num - 105\n",
    "    retrieved_text = \" \".join([retrieved[\"text\"] for retrieved in s_plus_outputs[i][\"additional_info\"][\"top_chunks_info\"]])\n",
    "    gt_sequences = questions[question_num][\"S\"]\n",
    "    retrieval_scores.append(soft_contains_context(retrieved_text, gt_sequences))\n",
    "print(np.mean(retrieval_scores))\n",
    "\n",
    "\n",
    "# with open(\"output/question_union_with_corresponding_in_context_mistral_large_chunksize256.jsonl\", \"r\") as f:\n",
    "#     lines = f.readlines()\n",
    "#     q_plus_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "# retrieval_scores = []\n",
    "# for question_num in question_indices:\n",
    "#     retrieved_text = \" \".join(\n",
    "#         [retrieved[\"text\"] for retrieved in q_plus_outputs[question_num][\"additional_info\"][\"top_chunks_info\"]]\n",
    "#     )\n",
    "#     gt_sequences = questions[question_num][\"S\"]\n",
    "#     retrieval_scores.append(soft_contains_context(retrieved_text, gt_sequences))\n",
    "# print(np.mean(retrieval_scores))\n",
    "\n",
    "\n",
    "# with open(\"output/union_with_corresponding_in_context_mistral_large_chunksize256.jsonl\", \"r\") as f:\n",
    "#     lines = f.readlines()\n",
    "#     u_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "# retrieval_scores = []\n",
    "# for question_num in question_indices:\n",
    "#     retrieved_text = \" \".join(\n",
    "#         [retrieved[\"text\"] for retrieved in u_outputs[question_num][\"additional_info\"][\"top_chunks_info\"]]\n",
    "#     )\n",
    "#     gt_sequences = questions[question_num][\"S\"]\n",
    "#     retrieval_scores.append(soft_contains_context(retrieved_text, gt_sequences))\n",
    "# print(np.mean(retrieval_scores))\n",
    "\n",
    "\n",
    "with open(\"output/testset_baseline_mistral_large_chunksize256.jsonl\", \"r\") as f:\n",
    "    lines = f.readlines()\n",
    "    baseline_outputs = [json.loads(line) for line in lines]\n",
    "\n",
    "retrieval_scores = []\n",
    "for question_num in question_indices:\n",
    "    i = question_num - 105\n",
    "    retrieved_text = \" \".join([retrieved[\"text\"] for retrieved in baseline_outputs[i][\"additional_info\"][\"top_chunks_info\"]])\n",
    "    gt_sequences = questions[question_num][\"S\"]\n",
    "    retrieval_scores.append(soft_contains_context(retrieved_text, gt_sequences))\n",
    "print(np.mean(retrieval_scores))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "LongRAG",
   "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": "3.10.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}