1_LLM_Simplification.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "11600f58-7b2f-4b4f-a937-40c1364a37dd",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import sys\n",
    "sys.path.append('../xmen/benchmarks')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b33ea696-4fae-4420-912d-7e742db8c685",
   "metadata": {},
   "source": [
    "# Entity Simplification with OpenAI / GPT-4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a903fc26-d6b5-4fd7-b0fd-950030816598",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from pathlib import Path\n",
    "import pandas as pd\n",
    "import datasets\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a76935f5-02eb-468b-b448-0d395daa583f",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "base_path = Path.home() / '.cache' / 'xmen' / 'symptemist'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "63d553a7-d1f4-4fea-86e9-bc4a8aa3fe83",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import dataloaders\n",
    "dataset = dataloaders.load_symptemist()[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "97e2fc46-6e09-44ea-a0fb-4bd4e0c8bc1e",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from xmen import load_kb\n",
    "kb = load_kb(base_path / 'symptemist.jsonl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "321745a4-3f7e-44aa-9274-40bb53ae27b6",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from xmen.evaluation import error_analysis, evaluate, evaluate_at_k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7f7fed2a-7a92-443e-84f2-1ffba8cee5a9",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/Florian.Borchert/miniconda3/envs/symptemist/lib/python3.10/site-packages/sklearn/base.py:348: InconsistentVersionWarning: Trying to unpickle estimator TfidfTransformer from version 1.3.0 when using version 1.3.1. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n",
      "/home/Florian.Borchert/miniconda3/envs/symptemist/lib/python3.10/site-packages/sklearn/base.py:348: InconsistentVersionWarning: Trying to unpickle estimator TfidfVectorizer from version 1.3.0 when using version 1.3.1. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">[08/20/24 10:33:19] </span><span style=\"color: #000080; text-decoration-color: #000080\">INFO    </span> Loading hierarchical faiss index                                <a href=\"file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/sap_bert_linker.py\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">sap_bert_linker.py</span></a><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">:</span><a href=\"file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/sap_bert_linker.py#153\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">153</span></a>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[2;36m[08/20/24 10:33:19]\u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO    \u001b[0m Loading hierarchical faiss index                                \u001b]8;id=871778;file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/sap_bert_linker.py\u001b\\\u001b[2msap_bert_linker.py\u001b[0m\u001b]8;;\u001b\\\u001b[2m:\u001b[0m\u001b]8;id=163321;file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/sap_bert_linker.py#153\u001b\\\u001b[2m153\u001b[0m\u001b]8;;\u001b\\\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">                    </span><span style=\"color: #000080; text-decoration-color: #000080\">INFO    </span> Loading index from                                                 <a href=\"file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">faiss_indexer.py</span></a><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">:</span><a href=\"file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py#64\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">64</span></a>\n",
       "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">                    </span>         <span style=\"color: #800080; text-decoration-color: #800080\">/home/Florian.Borchert/.cache/xmen/symptemist/index/sapbert/</span><span style=\"color: #ff00ff; text-decoration-color: #ff00ff\">embed_</span> <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                   </span>\n",
       "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">                    </span>         <span style=\"color: #ff00ff; text-decoration-color: #ff00ff\">faiss_hier.pickle</span>                                                  <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                   </span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[2;36m                   \u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO    \u001b[0m Loading index from                                                 \u001b]8;id=241989;file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py\u001b\\\u001b[2mfaiss_indexer.py\u001b[0m\u001b]8;;\u001b\\\u001b[2m:\u001b[0m\u001b]8;id=765404;file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py#64\u001b\\\u001b[2m64\u001b[0m\u001b]8;;\u001b\\\n",
       "\u001b[2;36m                    \u001b[0m         \u001b[35m/home/Florian.Borchert/.cache/xmen/symptemist/index/sapbert/\u001b[0m\u001b[95membed_\u001b[0m \u001b[2m                   \u001b[0m\n",
       "\u001b[2;36m                    \u001b[0m         \u001b[95mfaiss_hier.pickle\u001b[0m                                                  \u001b[2m                   \u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">[08/20/24 10:33:21] </span><span style=\"color: #000080; text-decoration-color: #000080\">INFO    </span> Loaded index of type <span style=\"font-weight: bold\">&lt;</span><span style=\"color: #ff00ff; text-decoration-color: #ff00ff; font-weight: bold\">class</span><span style=\"color: #000000; text-decoration-color: #000000\"> </span><span style=\"color: #008000; text-decoration-color: #008000\">'faiss.swigfaiss.IndexHNSWFlat'</span><span style=\"font-weight: bold\">&gt;</span> and   <a href=\"file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">faiss_indexer.py</span></a><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">:</span><a href=\"file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py#66\" target=\"_blank\"><span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">66</span></a>\n",
       "<span style=\"color: #7fbfbf; text-decoration-color: #7fbfbf\">                    </span>         size <span style=\"color: #008080; text-decoration-color: #008080; font-weight: bold\">1079623</span>                                                       <span style=\"color: #7f7f7f; text-decoration-color: #7f7f7f\">                   </span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[2;36m[08/20/24 10:33:21]\u001b[0m\u001b[2;36m \u001b[0m\u001b[34mINFO    \u001b[0m Loaded index of type \u001b[1m<\u001b[0m\u001b[1;95mclass\u001b[0m\u001b[39m \u001b[0m\u001b[32m'faiss.swigfaiss.IndexHNSWFlat'\u001b[0m\u001b[1m>\u001b[0m and   \u001b]8;id=74981;file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py\u001b\\\u001b[2mfaiss_indexer.py\u001b[0m\u001b]8;;\u001b\\\u001b[2m:\u001b[0m\u001b]8;id=682005;file:///mnt/nfs/home/Florian.Borchert/workspace/xmen/xmen/linkers/faiss_indexer.py#66\u001b\\\u001b[2m66\u001b[0m\u001b]8;;\u001b\\\n",
       "\u001b[2;36m                    \u001b[0m         size \u001b[1;36m1079623\u001b[0m                                                       \u001b[2m                   \u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from xmen.linkers import default_ensemble\n",
    "linker = default_ensemble(base_path / 'index', cuda=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "68f069a6-68d9-4270-8337-bed423ce251b",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "SYMPTEMIST_FEW_SHOT_EXAMPLES = [\n",
    "    (\"afebril\", \"temperatura corporal normal\"),\n",
    "    (\"induración de la vaginal testicular\", \"trastorno de testículo\"),\n",
    "    (\"formaciones mamelonadas en su interior a nivel de lóbulo superior\", \"lesión de pulmón\"),\n",
    "    (\"disnea\", \"disnea\"),\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "a45ed375-17d8-4f53-a5a8-954eb61403e5",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def count_differences(simplified, original):\n",
    "    cnt_all = 0\n",
    "    cnt_diff = 0\n",
    "    for ds, do in zip(simplified, original):\n",
    "        for es, eo in zip(ds['entities'], do['entities']):\n",
    "            cnt_all += 1\n",
    "            if es['text'] != eo['text']:\n",
    "                cnt_diff +=1\n",
    "    return cnt_diff, cnt_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "6d82557a-0471-4461-834c-692ea12dfd4a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "MODEL = 'gpt-4-0125-preview'\n",
    "# Pre-computed for SympTEMIST to save API calls\n",
    "table_file = 'symptemist_lookup_gpt-4-0125-preview_20240214-205237_prompt1.pkl'\n",
    "prompt = SYMPTEMIST_FEW_SHOT_EXAMPLES\n",
    "\n",
    "lookup_table = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4180cb40-b73b-4fbc-b413-266588ec3266",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pickle\n",
    "lookup_table = pickle.load(open(table_file, 'rb')) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5202b834-70d8-4844-9034-ad020315b501",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from xmen.data.simplification import GPTSimplifier, EntitySimplification, SimplifierWrapper"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "a5231a4b-df97-46a4-b438-98ccaf52be35",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "text_simplifier = GPTSimplifier(\n",
    "    model=MODEL, \n",
    "    open_ai_api_key=\"insert_api_key\", \n",
    "    fixed_few_shot_examples = prompt,\n",
    "    table=lookup_table\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "153fdaa7-5865-4f20-991d-f125e108e5a1",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "simplifier = EntitySimplification(text_simplifier, set_long_form=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "8c7ac146-d98f-4314-a1c2-38300a1d7d4a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Your task is to simplify expressions, such that the simplified version is more suitable for looking up concepts in a medical terminology. If the input is already simple enough, just return the input. \n",
      "\n",
      "Here are some examples:\n",
      "Input: ```afebril```\n",
      "Simplified: ```temperatura corporal normal```\n",
      "\n",
      "Input: ```induración de la vaginal testicular```\n",
      "Simplified: ```trastorno de testículo```\n",
      "\n",
      "Input: ```formaciones mamelonadas en su interior a nivel de lóbulo superior```\n",
      "Simplified: ```lesión de pulmón```\n",
      "\n",
      "Input: ```disnea```\n",
      "Simplified: ```disnea```\n",
      "\n",
      "Input: ```aumento de densidad en lóbulo inferior```\n",
      "Simplified:\n"
     ]
    }
   ],
   "source": [
    "print(text_simplifier.prompt_template.format('aumento de densidad en lóbulo inferior'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "bad2a5fd-2a92-4e4f-92f8-76167d3dab90",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'lesión de pulmón'"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text_simplifier.simplify('aumento de densidad en lóbulo inferior')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66f43922-8396-47a4-9429-40ab1e0a7867",
   "metadata": {},
   "source": [
    "# Generate Candidates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b182b5d9-7b4c-4567-849f-28b0259e0d8d",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a470315376864c3fbc5e418a7e9a6196",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "64fae84a544e43e780966dcab5a64fb9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "251252384826489994227fa6d2c240a9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "candidates = linker.predict_batch(dataset, top_k=64, batch_size=128)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "5be0bdb1-b419-4a79-b8d0-09eb4e9c5dc3",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.4388426128890837\n",
      "Recall@2 0.5610112523746895\n",
      "Recall@4 0.6226801110623995\n",
      "Recall@8 0.6786497150372643\n",
      "Recall@16 0.7219055969603975\n",
      "Recall@32 0.7558088557650153\n",
      "Recall@64 0.7837205903843344\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['train'], candidates['train'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "b5fe7abd-ab13-4b32-92a1-3aebed027cec",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.4647592463363573\n",
      "Recall@2 0.5757152826238661\n",
      "Recall@4 0.6427076064200977\n",
      "Recall@8 0.6929518492672715\n",
      "Recall@16 0.7264480111653873\n",
      "Recall@32 0.7571528262386602\n",
      "Recall@64 0.782274947662247\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['validation'], candidates['validation'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "d800c439-0384-4328-abcc-ee4763dfa050",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.4474885844748858\n",
      "Recall@2 0.5602388479100808\n",
      "Recall@4 0.6199508254302775\n",
      "Recall@8 0.6656129258868985\n",
      "Recall@16 0.7007376185458377\n",
      "Recall@32 0.7323498419388831\n",
      "Recall@64 0.7632595714787496\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['test'], candidates['test'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "8de92a40-48fb-4f7c-9380-c00faa0dbdd0",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "95242f144975406d9d8a297b9ef2ef2b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5c382e47097946ceb5add4a707b87af1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2331ce9a7cd74b26ad79111fc6e7adfe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "09d83c5da83245d48a418f0888ec79ea",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "af1dd52291504410a281584eedc26813",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bc2fad8613aa4f5ba2baf69fce220834",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "simplified_ds = simplifier.transform_batch(candidates)\n",
    "simplified_candidates = linker.predict_batch(simplified_ds, top_k=64, batch_size=128)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71e27c35-54fc-4add-928b-9000c868c732",
   "metadata": {},
   "source": [
    "# Determine Optimal Cutoff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "a4e0ff81-cc97-4122-a754-7308dc62ddf7",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "SPLIT = 'train'\n",
    "eval_candidates = candidates[SPLIT]\n",
    "eval_ds = dataset[SPLIT]\n",
    "eval_simplified_ds = simplified_candidates[SPLIT]\n",
    "\n",
    "def select_candidates(d, idx, filter_fn=lambda e: True):\n",
    "    result = []\n",
    "    for ei, ec in zip(d['entities'], eval_candidates[idx]['entities']):\n",
    "        if filter_fn(ec):\n",
    "            result.append(ec)\n",
    "        else:\n",
    "            result.append(ei)\n",
    "    return { 'entities' : result }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "de89b873-d4e0-4389-9f67-cd10a674cab2",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "baseline = evaluate_at_k(eval_ds, eval_candidates, silent=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "b1d7717f-1b91-4fbd-859b-a21d1b2c9315",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.55\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a631968f2f89419aa17545f2395c726a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(9, 6843)\n",
      "0.6000000000000001\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "74f45be0b0eb459a96c43c273d82c14d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(104, 6843)\n",
      "0.6500000000000001\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fcc3b90fb27d44f5bca8e8997e70bf33",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(469, 6843)\n",
      "0.7000000000000002\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9d2d17182bfb4e4a8cd5162394d9d4df",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1091, 6843)\n",
      "0.7500000000000002\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2f8ae75c728548d4b0f07386515bcc07",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1871, 6843)\n",
      "0.8000000000000003\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b03764a5c95943da9520f027783c35b0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2586, 6843)\n",
      "0.8500000000000003\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0954f398c5694bfbaf209223cb78fb0d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3239, 6843)\n",
      "0.9000000000000004\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "76b65552b71545d8a8db09ba692f2884",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3864, 6843)\n",
      "0.9500000000000004\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8aeab79d37d84fa3a2dace13622d402b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4357, 6843)\n",
      "1.0000000000000004\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "302484a4b7fd44b086c6e21a4746d555",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5841, 6843)\n"
     ]
    }
   ],
   "source": [
    "cutoff_eval = []\n",
    "\n",
    "for cutoff in np.arange(0.55, 1.05, 0.05):\n",
    "    print(cutoff)\n",
    "    fn = lambda e: e['normalized'][0]['score'] >= cutoff\n",
    "    best_candidates = eval_simplified_ds.map(lambda d, i: select_candidates(d, i, fn), with_indices=True)\n",
    "    diffs = count_differences(best_candidates, eval_candidates)\n",
    "    print(diffs)\n",
    "    eval_res = evaluate_at_k(eval_ds, best_candidates, silent=True)\n",
    "    cutoff_eval_i = {'cutoff' : cutoff, 'num_changed' : diffs[0], 'num_all' : diffs[1]}\n",
    "    for i, es in eval_res.items():\n",
    "        cutoff_eval_i[i] = es['strict']['recall']\n",
    "    cutoff_eval.append(cutoff_eval_i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "3dc40460-a72e-436f-964e-e2dd6a39fe92",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "cutoff_eval = pd.DataFrame(cutoff_eval).round(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "36fece38-7cc2-467b-b0cb-fdfad7fc9d2d",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>cutoff</th>\n",
       "      <th>num_changed</th>\n",
       "      <th>num_all</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>4</th>\n",
       "      <th>8</th>\n",
       "      <th>16</th>\n",
       "      <th>32</th>\n",
       "      <th>64</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.55</td>\n",
       "      <td>9</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.439</td>\n",
       "      <td>0.561</td>\n",
       "      <td>0.623</td>\n",
       "      <td>0.679</td>\n",
       "      <td>0.722</td>\n",
       "      <td>0.756</td>\n",
       "      <td>0.784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.60</td>\n",
       "      <td>104</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.439</td>\n",
       "      <td>0.561</td>\n",
       "      <td>0.623</td>\n",
       "      <td>0.678</td>\n",
       "      <td>0.722</td>\n",
       "      <td>0.756</td>\n",
       "      <td>0.784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.65</td>\n",
       "      <td>469</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.442</td>\n",
       "      <td>0.565</td>\n",
       "      <td>0.626</td>\n",
       "      <td>0.681</td>\n",
       "      <td>0.726</td>\n",
       "      <td>0.760</td>\n",
       "      <td>0.789</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.70</td>\n",
       "      <td>1091</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.449</td>\n",
       "      <td>0.572</td>\n",
       "      <td>0.635</td>\n",
       "      <td>0.688</td>\n",
       "      <td>0.733</td>\n",
       "      <td>0.768</td>\n",
       "      <td>0.796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.75</td>\n",
       "      <td>1871</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.451</td>\n",
       "      <td>0.577</td>\n",
       "      <td>0.640</td>\n",
       "      <td>0.694</td>\n",
       "      <td>0.738</td>\n",
       "      <td>0.773</td>\n",
       "      <td>0.802</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.80</td>\n",
       "      <td>2586</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.453</td>\n",
       "      <td>0.577</td>\n",
       "      <td>0.638</td>\n",
       "      <td>0.692</td>\n",
       "      <td>0.739</td>\n",
       "      <td>0.773</td>\n",
       "      <td>0.803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.85</td>\n",
       "      <td>3239</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.453</td>\n",
       "      <td>0.575</td>\n",
       "      <td>0.638</td>\n",
       "      <td>0.694</td>\n",
       "      <td>0.747</td>\n",
       "      <td>0.781</td>\n",
       "      <td>0.811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.90</td>\n",
       "      <td>3864</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.446</td>\n",
       "      <td>0.568</td>\n",
       "      <td>0.635</td>\n",
       "      <td>0.691</td>\n",
       "      <td>0.745</td>\n",
       "      <td>0.781</td>\n",
       "      <td>0.811</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.95</td>\n",
       "      <td>4357</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.433</td>\n",
       "      <td>0.556</td>\n",
       "      <td>0.627</td>\n",
       "      <td>0.686</td>\n",
       "      <td>0.743</td>\n",
       "      <td>0.781</td>\n",
       "      <td>0.810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.00</td>\n",
       "      <td>5841</td>\n",
       "      <td>6843</td>\n",
       "      <td>0.428</td>\n",
       "      <td>0.552</td>\n",
       "      <td>0.623</td>\n",
       "      <td>0.686</td>\n",
       "      <td>0.743</td>\n",
       "      <td>0.781</td>\n",
       "      <td>0.810</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   cutoff  num_changed  num_all      1      2      4      8     16     32  \\\n",
       "0    0.55            9     6843  0.439  0.561  0.623  0.679  0.722  0.756   \n",
       "1    0.60          104     6843  0.439  0.561  0.623  0.678  0.722  0.756   \n",
       "2    0.65          469     6843  0.442  0.565  0.626  0.681  0.726  0.760   \n",
       "3    0.70         1091     6843  0.449  0.572  0.635  0.688  0.733  0.768   \n",
       "4    0.75         1871     6843  0.451  0.577  0.640  0.694  0.738  0.773   \n",
       "5    0.80         2586     6843  0.453  0.577  0.638  0.692  0.739  0.773   \n",
       "6    0.85         3239     6843  0.453  0.575  0.638  0.694  0.747  0.781   \n",
       "7    0.90         3864     6843  0.446  0.568  0.635  0.691  0.745  0.781   \n",
       "8    0.95         4357     6843  0.433  0.556  0.627  0.686  0.743  0.781   \n",
       "9    1.00         5841     6843  0.428  0.552  0.623  0.686  0.743  0.781   \n",
       "\n",
       "      64  \n",
       "0  0.784  \n",
       "1  0.784  \n",
       "2  0.789  \n",
       "3  0.796  \n",
       "4  0.802  \n",
       "5  0.803  \n",
       "6  0.811  \n",
       "7  0.811  \n",
       "8  0.810  \n",
       "9  0.810  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cutoff_eval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "f6ffc479-d063-4aa9-a2b7-7b86895d4437",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "import matplotlib\n",
    "\n",
    "matplotlib.rcParams.update({'font.size': 14, 'font.family' : 'serif'})\n",
    "\n",
    "fig, axs = plt.subplots(2, 1, figsize=(10,5), sharex=True, gridspec_kw={'height_ratios': [2, 1]}, squeeze=True)\n",
    "\n",
    "ax1 = axs[0]\n",
    "\n",
    "x_range = np.arange(0, len(cutoff_eval))\n",
    "width = 0.3       \n",
    "\n",
    "ax1.bar(x_range - 0.5 * width, \n",
    "        (cutoff_eval.set_index('cutoff')[1] - baseline[1]['strict']['recall']).values, \n",
    "        width, color='darkblue', linewidth=0.5, edgecolor='black', label='$\\Delta$ Recall@1 (pp.)')\n",
    "ax1.bar(x_range + 0.5 * width, \n",
    "        (cutoff_eval.set_index('cutoff')[64] - baseline[64]['strict']['recall']).values, \n",
    "        width, color='lightblue', linewidth=0.5, edgecolor='black', label='$\\Delta$ Recall@64 (pp.)')\n",
    "\n",
    "ax1.grid(axis='y')\n",
    "ax1.set_yticks(np.arange(-0.005, 0.03, 0.005))\n",
    "ax1.set_yticklabels([f'+{t * 100:.1f}' if t > 0 else f'{t * 100:.1f}' for t in ax1.get_yticks()])\n",
    "\n",
    "ax1.set_xticks(x_range)\n",
    "ax1.set_xticklabels(cutoff_eval.cutoff.values)\n",
    "\n",
    "ax0 = axs[1]\n",
    "ax0.grid(axis='y')\n",
    "ax0.plot(x_range, cutoff_eval.num_changed / cutoff_eval.num_all, color='red', label='Simplified Mentions (%)', marker='o')\n",
    "ax0.set_yticks(np.arange(0,1.02,0.2))\n",
    "ax0.set_yticklabels([f\"{int(t * 100)}%\" if t >= 0 and t <=1 else \"\" for t in ax0.get_yticks()])\n",
    "\n",
    "ax0.set_xlabel('Confidence Threshold for Entity Simplification')\n",
    "ax0.legend()\n",
    "\n",
    "ax0.set_xticks(x_range)\n",
    "ax0.set_xticklabels(cutoff_eval.cutoff.values)\n",
    "\n",
    "ax1.legend()\n",
    "\n",
    "plt.savefig('diff_cutoff.png', dpi=1200, bbox_inches='tight')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "864aa5af-5cd6-44cf-9dfd-e933ff9116de",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "best_cutoff = 0.85"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "0dda3df6-f5d5-4c54-b0c3-439f5808ddef",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1ee895c87b2844bea15ce4eca535ba61",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "218294ab6e5a46929b4f18f191729faa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4a3484ca54a4c729d3ce1d3528d6e61",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e370cecf0c2c49ab82de27e606cda5da",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a0096a518580436d92a85160c198e256",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "de18f39562244285b2e950ed77559efd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "62dba6cd4ec84cfca7efedfeb7e5b48f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "55bf24d23f654ca58226aeaa0f7b8233",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "55bcf545d1c04ccea3749d7fc97939f8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# we use a SimplifierWrapper so that the original spans can be restored after candidate generatio\n",
    "simplifier_wrapper = SimplifierWrapper(\n",
    "    linker,\n",
    "    text_simplifier, \n",
    "    filter_fn=lambda c: c['normalized'][0]['score'] < best_cutoff, \n",
    "    set_long_form=True\n",
    ")\n",
    "simplified_candidates_cutoff = simplifier_wrapper.predict_batch(candidates, top_k=64, batch_size=128)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "aa1b9ec3-5a2e-45d2-b93b-b4845b047ea4",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(0, 6843)\n",
      "(0, 1433)\n",
      "(0, 2848)\n"
     ]
    }
   ],
   "source": [
    "print(count_differences(simplified_candidates_cutoff['train'], dataset['train']))\n",
    "print(count_differences(simplified_candidates_cutoff['validation'], dataset['validation']))\n",
    "print(count_differences(simplified_candidates_cutoff['test'], dataset['test']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "908c1fb6-ef49-4d4d-9b34-2a5ad25cd0c0",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.48499651081646894\n",
      "Recall@2 0.5903698534542917\n",
      "Recall@4 0.6566643405443127\n",
      "Recall@8 0.7083042568039079\n",
      "Recall@16 0.7522679692951849\n",
      "Recall@32 0.7773900907187719\n",
      "Recall@64 0.8066992323796232\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['validation'], simplified_candidates_cutoff['validation'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "c8a9a5ff-73f7-446e-aa3e-b283d777612d",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.46224095539164034\n",
      "Recall@2 0.5760449596066034\n",
      "Recall@4 0.6434843695117668\n",
      "Recall@8 0.6880927291886196\n",
      "Recall@16 0.7288373726729891\n",
      "Recall@32 0.7629083245521602\n",
      "Recall@64 0.794169301018616\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['test'], simplified_candidates_cutoff['test'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8e693279-cec7-4e0b-b3c6-29a23258d7f8",
   "metadata": {},
   "source": [
    "# Save"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "2ea15cb9-8d1e-4022-8116-388bd97eacb3",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b1410be903e34305ad6bc14f320d79ac",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Saving the dataset (0/1 shards):   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "730ce7968f164114bdbd969dfaa9cdfa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Saving the dataset (0/1 shards):   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "26f15d7ce7b34066ae97125b1671e32e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Saving the dataset (0/1 shards):   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "simplified_candidates_cutoff.save_to_disk('symptemist_candidates_simplified_cutoff')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71ddb5c4-cf8a-4180-8883-3c99ad8872b4",
   "metadata": {},
   "source": [
    "# Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "75546489-f9fc-40c1-930a-e5cb2a9f1254",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from xmen.evaluation import error_analysis, evaluate_at_k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "b5b5f71f-87f1-405a-8ec4-c729e641ecfd",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.4474885844748858\n",
      "Recall@2 0.5602388479100808\n",
      "Recall@4 0.6199508254302775\n",
      "Recall@8 0.6656129258868985\n",
      "Recall@16 0.7007376185458377\n",
      "Recall@32 0.7323498419388831\n",
      "Recall@64 0.7632595714787496\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['test'], candidates['test'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "116dcde7-e04b-4d42-999d-f9383620f2ab",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall@1 0.46224095539164034\n",
      "Recall@2 0.5760449596066034\n",
      "Recall@4 0.6434843695117668\n",
      "Recall@8 0.6880927291886196\n",
      "Recall@16 0.7288373726729891\n",
      "Recall@32 0.7629083245521602\n",
      "Recall@64 0.794169301018616\n"
     ]
    }
   ],
   "source": [
    "_ = evaluate_at_k(dataset['test'], simplified_candidates_cutoff['test'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "1fd74a01-9067-4e28-93b9-00f554198f85",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fd6bba55b15341aeb0d38d4b1b6f4639",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/608 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "010d4e3bb0da45119f0530aef0297bc8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1ba5721154d1492ca3181eac6a239550",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/136 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "simplified_ds = simplifier.transform_batch(dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "a83f15e9-1d9e-41f2-8002-c45e728d45bf",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "ea_before_simple_test = error_analysis(dataset['test'], candidates['test'])\n",
    "ea_after_simple_test = error_analysis(simplified_ds['test'], simplified_candidates_cutoff['test'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "9037936b-fbb4-4fa3-a298-ca49757d443a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "ea_after_no_thresh_test = error_analysis(simplified_ds['test'], simplified_candidates['test'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "7ca8bc69-23d1-42de-9fa1-8b0fb0b13090",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "SPLIT = 'test'\n",
    "eval_candidates = candidates[SPLIT]\n",
    "eval_ds = dataset[SPLIT]\n",
    "eval_simplified_ds = simplified_candidates[SPLIT]\n",
    "\n",
    "def select_candidates(d, idx, filter_fn=lambda e: True):\n",
    "    result = []\n",
    "    for ei, ec in zip(d['entities'], eval_candidates[idx]['entities']):\n",
    "        if filter_fn(ec):\n",
    "            result.append(ec)\n",
    "        else:\n",
    "            result.append(ei)\n",
    "    return { 'entities' : result }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "c3ad311c-bd58-464c-922d-4c5e47a55495",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c7c80916a7ae4925890f41c03cf1f351",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Map:   0%|          | 0/246 [00:00<?, ? examples/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fn = lambda e: e['normalized'][0]['score'] >= best_cutoff\n",
    "best_candidates = eval_simplified_ds.map(lambda d, i: select_candidates(d, i, fn), with_indices=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "238db9fe-1638-4210-9ad2-f6a9a459b814",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_len(t):\n",
    "    return len(t.split(' '))\n",
    "\n",
    "def get_stats(ea_df, name, ds):\n",
    "    res = {}\n",
    "    res['name'] = name\n",
    "    word_lens = pd.Series([get_len(e['text'][0]) for d in ds for e in d['entities']])\n",
    "    res['max_length'] = int(word_lens.max())\n",
    "    res['mean_length'] = word_lens.mean().round(2)\n",
    "    res['recall_1'] = (ea_df.pred_index == 0).sum() / len(ea_df)\n",
    "    res['recall_64'] = (ea_df.pred_index >= 0).sum() / len(ea_df)\n",
    "    res['mean_score'] = ea_df.pred_top_score.mean()\n",
    "    for i in np.linspace(0.5, 1.0, 6):\n",
    "        begin, end = i, i + 0.1\n",
    "        sub_ea = ea_df[(ea_df.pred_top_score >= begin) & (ea_df.pred_top_score < end)]\n",
    "        prefix = f\"{begin}-{end}\"\n",
    "        res[f'{prefix}_len'] = len(sub_ea)\n",
    "        res[f'{prefix}_recall_1'] = (sub_ea.pred_index == 0).sum() / len(sub_ea)\n",
    "        res[f'{prefix}_recall_64'] = (sub_ea.pred_index >= 0).sum() / len(sub_ea)\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "a8124dd6-039f-47f3-8087-06a257700110",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>name</th>\n",
       "      <th>no_simpl</th>\n",
       "      <th>simpl_thresh</th>\n",
       "      <th>delta</th>\n",
       "      <th>simpl_no_thresh</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>max_length</th>\n",
       "      <td>23.000</td>\n",
       "      <td>13.000</td>\n",
       "      <td>-10.000</td>\n",
       "      <td>13.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_length</th>\n",
       "      <td>3.630</td>\n",
       "      <td>2.610</td>\n",
       "      <td>-1.020</td>\n",
       "      <td>2.830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall_1</th>\n",
       "      <td>0.447</td>\n",
       "      <td>0.462</td>\n",
       "      <td>0.015</td>\n",
       "      <td>0.432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall_64</th>\n",
       "      <td>0.763</td>\n",
       "      <td>0.794</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean_score</th>\n",
       "      <td>0.848</td>\n",
       "      <td>0.833</td>\n",
       "      <td>-0.015</td>\n",
       "      <td>0.815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5-0.6_len</th>\n",
       "      <td>37.000</td>\n",
       "      <td>83.000</td>\n",
       "      <td>46.000</td>\n",
       "      <td>83.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5-0.6_recall_1</th>\n",
       "      <td>0.027</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.021</td>\n",
       "      <td>0.048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.5-0.6_recall_64</th>\n",
       "      <td>0.243</td>\n",
       "      <td>0.241</td>\n",
       "      <td>-0.002</td>\n",
       "      <td>0.241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.6-0.7_len</th>\n",
       "      <td>443.000</td>\n",
       "      <td>615.000</td>\n",
       "      <td>172.000</td>\n",
       "      <td>638.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.6-0.7_recall_1</th>\n",
       "      <td>0.063</td>\n",
       "      <td>0.098</td>\n",
       "      <td>0.034</td>\n",
       "      <td>0.097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.6-0.7_recall_64</th>\n",
       "      <td>0.336</td>\n",
       "      <td>0.488</td>\n",
       "      <td>0.151</td>\n",
       "      <td>0.506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.7-0.7999999999999999_len</th>\n",
       "      <td>654.000</td>\n",
       "      <td>578.000</td>\n",
       "      <td>-76.000</td>\n",
       "      <td>743.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.7-0.7999999999999999_recall_1</th>\n",
       "      <td>0.220</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.110</td>\n",
       "      <td>0.342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.7-0.7999999999999999_recall_64</th>\n",
       "      <td>0.625</td>\n",
       "      <td>0.744</td>\n",
       "      <td>0.119</td>\n",
       "      <td>0.782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.8-0.9_len</th>\n",
       "      <td>555.000</td>\n",
       "      <td>411.000</td>\n",
       "      <td>-144.000</td>\n",
       "      <td>507.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.8-0.9_recall_1</th>\n",
       "      <td>0.438</td>\n",
       "      <td>0.494</td>\n",
       "      <td>0.056</td>\n",
       "      <td>0.523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.8-0.9_recall_64</th>\n",
       "      <td>0.845</td>\n",
       "      <td>0.910</td>\n",
       "      <td>0.065</td>\n",
       "      <td>0.933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.9-1.0_len</th>\n",
       "      <td>447.000</td>\n",
       "      <td>447.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>217.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.9-1.0_recall_1</th>\n",
       "      <td>0.682</td>\n",
       "      <td>0.682</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0.9-1.0_recall_64</th>\n",
       "      <td>0.978</td>\n",
       "      <td>0.978</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1.1_len</th>\n",
       "      <td>710.000</td>\n",
       "      <td>710.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>656.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1.1_recall_1</th>\n",
       "      <td>0.779</td>\n",
       "      <td>0.779</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1.0-1.1_recall_64</th>\n",
       "      <td>0.986</td>\n",
       "      <td>0.986</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.992</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "name                              no_simpl  simpl_thresh    delta  \\\n",
       "max_length                          23.000        13.000  -10.000   \n",
       "mean_length                          3.630         2.610   -1.020   \n",
       "recall_1                             0.447         0.462    0.015   \n",
       "recall_64                            0.763         0.794    0.031   \n",
       "mean_score                           0.848         0.833   -0.015   \n",
       "0.5-0.6_len                         37.000        83.000   46.000   \n",
       "0.5-0.6_recall_1                     0.027         0.048    0.021   \n",
       "0.5-0.6_recall_64                    0.243         0.241   -0.002   \n",
       "0.6-0.7_len                        443.000       615.000  172.000   \n",
       "0.6-0.7_recall_1                     0.063         0.098    0.034   \n",
       "0.6-0.7_recall_64                    0.336         0.488    0.151   \n",
       "0.7-0.7999999999999999_len         654.000       578.000  -76.000   \n",
       "0.7-0.7999999999999999_recall_1      0.220         0.330    0.110   \n",
       "0.7-0.7999999999999999_recall_64     0.625         0.744    0.119   \n",
       "0.8-0.9_len                        555.000       411.000 -144.000   \n",
       "0.8-0.9_recall_1                     0.438         0.494    0.056   \n",
       "0.8-0.9_recall_64                    0.845         0.910    0.065   \n",
       "0.9-1.0_len                        447.000       447.000    0.000   \n",
       "0.9-1.0_recall_1                     0.682         0.682    0.000   \n",
       "0.9-1.0_recall_64                    0.978         0.978    0.000   \n",
       "1.0-1.1_len                        710.000       710.000    0.000   \n",
       "1.0-1.1_recall_1                     0.779         0.779    0.000   \n",
       "1.0-1.1_recall_64                    0.986         0.986    0.000   \n",
       "\n",
       "name                              simpl_no_thresh  \n",
       "max_length                                 13.000  \n",
       "mean_length                                 2.830  \n",
       "recall_1                                    0.432  \n",
       "recall_64                                   0.794  \n",
       "mean_score                                  0.815  \n",
       "0.5-0.6_len                                83.000  \n",
       "0.5-0.6_recall_1                            0.048  \n",
       "0.5-0.6_recall_64                           0.241  \n",
       "0.6-0.7_len                               638.000  \n",
       "0.6-0.7_recall_1                            0.097  \n",
       "0.6-0.7_recall_64                           0.506  \n",
       "0.7-0.7999999999999999_len                743.000  \n",
       "0.7-0.7999999999999999_recall_1             0.342  \n",
       "0.7-0.7999999999999999_recall_64            0.782  \n",
       "0.8-0.9_len                               507.000  \n",
       "0.8-0.9_recall_1                            0.523  \n",
       "0.8-0.9_recall_64                           0.933  \n",
       "0.9-1.0_len                               217.000  \n",
       "0.9-1.0_recall_1                            0.659  \n",
       "0.9-1.0_recall_64                           0.982  \n",
       "1.0-1.1_len                               656.000  \n",
       "1.0-1.1_recall_1                            0.764  \n",
       "1.0-1.1_recall_64                           0.992  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame([\n",
    "    get_stats(ea_before_simple_test, 'no_simpl', candidates['test']),\n",
    "    get_stats(ea_after_simple_test, 'simpl_thresh', best_candidates),\n",
    "    get_stats(ea_after_no_thresh_test, 'simpl_no_thresh', simplified_ds['test']),\n",
    "]).set_index('name').T\n",
    "df['delta'] = df.simpl_thresh - df.no_simpl\n",
    "df.round(3)[['no_simpl', 'simpl_thresh', 'delta', 'simpl_no_thresh']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "7ad38acd-77c2-4fdf-9413-6680eb27c10d",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "ea_before_simpl = error_analysis(dataset['train'], candidates['train'])\n",
    "ea_after_simple = error_analysis(simplified_ds['train'], simplified_candidates_cutoff['train'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "902d75a1-967a-41ab-aff5-0fb95b25dbf8",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.294201861130995 2.5755189692197566\n"
     ]
    }
   ],
   "source": [
    "both_correct = (ea_before_simpl.pred_index == 0) & (ea_after_simple.pred_index == 0)\n",
    "print(ea_before_simpl[both_correct]._word_len.mean(), ea_after_simple[both_correct]._word_len.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "d699c479-b51b-4ca2-a1da-97e565c22bf0",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5.005181347150259 3.354922279792746\n"
     ]
    }
   ],
   "source": [
    "better_recall = (ea_before_simpl.pred_index == -1) & (ea_after_simple.pred_index >= 0)\n",
    "print(ea_before_simpl[better_recall]._word_len.mean(), ea_after_simple[better_recall]._word_len.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "625d5ebd-84dc-4e59-92f6-29b22945de15",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.5849056603773586 3.339622641509434\n"
     ]
    }
   ],
   "source": [
    "better_recall_1 = (ea_before_simpl.pred_index == -1) & (ea_after_simple.pred_index == 0)\n",
    "print(ea_before_simpl[better_recall_1]._word_len.mean(), ea_after_simple[better_recall_1]._word_len.mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "97e880fd-6fb8-4d80-8b98-02fedf980978",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.339622641509434"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ea_after_simple[(ea_before_simpl.pred_index == -1) & (ea_after_simple.pred_index == 0)]._word_len.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "0e23118a-95c6-4098-8697-c622dd715822",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.339622641509434"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ea_after_simple[(ea_before_simpl.pred_index == -1) & (ea_after_simple.pred_index == 0)]._word_len.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e064855-7dc0-4c40-a8a6-489838b690e4",
   "metadata": {},
   "source": [
    "## Error Analysis by Mention Length"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "3d06c300-6278-46b7-99e8-d8627a1cb590",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "index = []\n",
    "labels = []\n",
    "match_any = []\n",
    "match_lost = []\n",
    "better_ranking = []\n",
    "worse_ranking = []\n",
    "n_candidates = []\n",
    "\n",
    "end = 10\n",
    "\n",
    "def get_new_candidates(i):\n",
    "    return ea_after_simple[(ea_before_simpl._word_len == i) & (ea_before_simpl.pred_index == -1) & (ea_after_simple.pred_index != -1)]\n",
    "\n",
    "def get_better_ranking(i):\n",
    "    return ea_after_simple[(ea_before_simpl._word_len == i) & (ea_after_simple.pred_index != -1) & (ea_after_simple.pred_index < ea_before_simpl.pred_index)]\n",
    "\n",
    "def get_worse_ranking(i):\n",
    "    return ea_after_simple[(ea_before_simpl._word_len == i) & (ea_before_simpl.pred_index != -1) & (ea_after_simple.pred_index > ea_before_simpl.pred_index)]\n",
    "\n",
    "def get_lost_candidates(i):\n",
    "    return ea_after_simple[(ea_before_simpl._word_len == i) & (ea_before_simpl.pred_index != -1) & (ea_after_simple.pred_index == -1)]\n",
    "\n",
    "def get_n_candidates(i):\n",
    "    return ea_after_simple[(ea_before_simpl._word_len == i)]\n",
    "\n",
    "for i in range(1, end + 1):\n",
    "    n_candidates_i = get_n_candidates(i)\n",
    "    eaany = get_new_candidates(i)\n",
    "    eabetter = get_better_ranking(i)\n",
    "    ealost = get_lost_candidates(i)\n",
    "    eaworse = get_worse_ranking(i)\n",
    "    \n",
    "    n_candidates.append(len(n_candidates_i))\n",
    "    index.append(i)\n",
    "    labels.append(str(i))\n",
    "    match_any.append(len(eaany))\n",
    "    better_ranking.append(len(eabetter))\n",
    "    match_lost.append(-len(ealost))\n",
    "    worse_ranking.append(-len(eaworse))\n",
    "    \n",
    "labels.append(str(f'>{ i }'))\n",
    "\n",
    "i += 1\n",
    "\n",
    "n_candidates_i = get_n_candidates(i)\n",
    "eaany = get_new_candidates(i)\n",
    "eabetter = get_better_ranking(i)\n",
    "ealost = get_lost_candidates(i)\n",
    "eaworse = get_worse_ranking(i)\n",
    "\n",
    "n_candidates.append(len(n_candidates_i))\n",
    "index.append(i)\n",
    "match_any.append(len(eaany))\n",
    "better_ranking.append(len(eabetter))\n",
    "match_lost.append(-len(ealost))\n",
    "worse_ranking.append(-len(eaworse))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "a9969893-7156-4dba-831b-7a2cea9e2e79",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_rel(m):\n",
    "    return [mi / ci for mi, ci in zip(m, n_candidates)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "e62c8bcf-72fd-44ed-bb64-9b71256af58e",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x1200 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib\n",
    "from matplotlib import pyplot as plt\n",
    "import matplotlib.font_manager as font_manager\n",
    "\n",
    "\n",
    "fig, axs = plt.subplots(3, 1, figsize=(8,12), gridspec_kw={'height_ratios': [0.4, 1, 1]})\n",
    "font_size_lg = 14\n",
    "font_size_sm = 12\n",
    "\n",
    "matplotlib.rcParams.update({'font.size': font_size_lg, 'font.family' : 'serif'})\n",
    "\n",
    "width = 0.2\n",
    "\n",
    "ax0 = axs[0]\n",
    "ax0.bar(np.array(index), n_candidates, width=width, label='No. Mentions', color='grey', edgecolor='black', linewidth=0.5)\n",
    "ax0.grid(axis='y')\n",
    "ax0.set_title('(a) Total number of mentions', size=font_size_lg)\n",
    "ax0.set_yticks(range(0,3000,500))\n",
    "\n",
    "ax1 = axs[1]\n",
    "\n",
    "ax1.grid(axis='y')\n",
    "\n",
    "#ax.bar(np.array(index) - width, match_0, width=width, label='')\n",
    "ax1.bar(np.array(index) - width, get_rel(match_any), width=width, label='↑ Recall', color='lightgreen', edgecolor='black', linewidth=0.5)\n",
    "ax1.bar(np.array(index), get_rel(match_lost), width=width, label='↓ Recall', color='wheat', edgecolor='black', linewidth=0.5)\n",
    "ax1.bar(np.array(index) + width, get_rel(np.array(match_any) + np.array(match_lost)), width=width, label='Difference', color='blue', edgecolor='black', linewidth=0.5)\n",
    "\n",
    "y_range = np.arange(-0.1, 0.18, 0.05)\n",
    "ax1.set_yticks(y_range)\n",
    "ax1.set_yticklabels([f'{\"+\" if i > 0 else \"\"}{round(i * 100)}pp.' for i in y_range])\n",
    "ax1.set_ylim(-0.12,0.14)\n",
    "ax1.set_title('(b) Difference in recall@64 due to entity simplification', size=font_size_lg)\n",
    "ax1.legend(loc='upper left', \n",
    "           ncol=3, fontsize=font_size_sm)    \n",
    "\n",
    "ax2 = axs[2]\n",
    "    \n",
    "ax2.bar(np.array(index) - width, get_rel(better_ranking), width=width, label='↑ Ranking', color='lightgreen', edgecolor='black', linewidth=0.5)\n",
    "ax2.bar(np.array(index), get_rel(worse_ranking), width=width, label='↓ Ranking', color='wheat', edgecolor='black', linewidth=0.5)\n",
    "ax2.bar(np.array(index) + width, get_rel(np.array(better_ranking) + np.array(worse_ranking)), width=width, label='Difference', color='blue', edgecolor='black', linewidth=0.5)\n",
    "    \n",
    "y_range_2 = np.arange(-0.2,0.3,0.05)\n",
    "ax2.set_yticks(y_range_2)\n",
    "ax2.set_yticklabels([f'{\"\" if i > 0 else \"\"}{abs(round(i * 100))}%' for i in y_range_2])\n",
    " \n",
    "ax2.legend(loc='upper left', ncol=3, fontsize=font_size_sm)\n",
    "ax2.set_title('(c) Proportion of mentions with increased / decreased ranking\\nof the ground-truth concept due to entity simplification', size=font_size_lg)\n",
    "    \n",
    "ax2.grid(axis='y')\n",
    "ax2.set_xlabel('Mention length (tokens)')\n",
    "\n",
    "for ax in axs:\n",
    "    ax.set_xlim(axs[0].get_xlim())\n",
    "    ax.set_xticks(index)\n",
    "    ax.set_xticklabels(labels)\n",
    "\n",
    "fig.tight_layout()\n",
    "    \n",
    "plt.savefig('gain_vs_length.jpg', dpi=1200, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce11b62b-561d-4345-bcf6-e7bc01545c78",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:symptemist]",
   "language": "python",
   "name": "conda-env-symptemist-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}