web scraping and finetuning sentence transformer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "23a6b22d",
   "metadata": {},
   "outputs": [],
   "source": [
    "from bs4 import BeautifulSoup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8d3d0455",
   "metadata": {},
   "outputs": [],
   "source": [
    "import requests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2517dda7",
   "metadata": {},
   "outputs": [],
   "source": [
    "url='https://clinicaltrials.gov/search?cond=heart%20attack'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "30b99f4a",
   "metadata": {},
   "outputs": [],
   "source": [
    "page=requests.get(url)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8ff87583",
   "metadata": {},
   "outputs": [],
   "source": [
    "soup=BeautifulSoup(page.text,'html')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "c27dfa8e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<!DOCTYPE html>\n",
       "<html data-critters-container=\"\" lang=\"en\">\n",
       "<head><link crossorigin=\"\" href=\"https://fonts.gstatic.com\" rel=\"preconnect\"/>\n",
       "<meta charset=\"utf-8\"/>\n",
       "<script type=\"text/javascript\">\n",
       "    window.ncbi_startTime = new Date();\n",
       "  </script>\n",
       "<meta content=\"IE=edge\" http-equiv=\"X-UA-Compatible\"/>\n",
       "<!-- Mobile properties -->\n",
       "<meta content=\"True\" name=\"HandheldFriendly\"/>\n",
       "<meta content=\"320\" name=\"MobileOptimized\"/>\n",
       "<meta content=\"width=device-width, initial-scale=1.0, minimum-scale=1\" name=\"viewport\"/>\n",
       "<style type=\"text/css\">@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fCRc4EsA.woff2) format('woff2');unicode-range:U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;}@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fABc4EsA.woff2) format('woff2');unicode-range:U+0301, U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;}@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fCBc4EsA.woff2) format('woff2');unicode-range:U+1F00-1FFF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fBxc4EsA.woff2) format('woff2');unicode-range:U+0370-03FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fCxc4EsA.woff2) format('woff2');unicode-range:U+0102-0103, U+0110-0111, U+0128-0129, U+0168-0169, U+01A0-01A1, U+01AF-01B0, U+0300-0301, U+0303-0304, U+0308-0309, U+0323, U+0329, U+1EA0-1EF9, U+20AB;}@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fChc4EsA.woff2) format('woff2');unicode-range:U+0100-02AF, U+0304, U+0308, U+0329, U+1E00-1E9F, U+1EF2-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:300;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmSU5fBBc4.woff2) format('woff2');unicode-range:U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu72xKOzY.woff2) format('woff2');unicode-range:U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu5mxKOzY.woff2) format('woff2');unicode-range:U+0301, U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu7mxKOzY.woff2) format('woff2');unicode-range:U+1F00-1FFF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu4WxKOzY.woff2) format('woff2');unicode-range:U+0370-03FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu7WxKOzY.woff2) format('woff2');unicode-range:U+0102-0103, U+0110-0111, U+0128-0129, U+0168-0169, U+01A0-01A1, U+01AF-01B0, U+0300-0301, U+0303-0304, U+0308-0309, U+0323, U+0329, U+1EA0-1EF9, U+20AB;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu7GxKOzY.woff2) format('woff2');unicode-range:U+0100-02AF, U+0304, U+0308, U+0329, U+1E00-1E9F, U+1EF2-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:400;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOmCnqEu92Fr1Mu4mxK.woff2) format('woff2');unicode-range:U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fCRc4EsA.woff2) format('woff2');unicode-range:U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fABc4EsA.woff2) format('woff2');unicode-range:U+0301, U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fCBc4EsA.woff2) format('woff2');unicode-range:U+1F00-1FFF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fBxc4EsA.woff2) format('woff2');unicode-range:U+0370-03FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fCxc4EsA.woff2) format('woff2');unicode-range:U+0102-0103, U+0110-0111, U+0128-0129, U+0168-0169, U+01A0-01A1, U+01AF-01B0, U+0300-0301, U+0303-0304, U+0308-0309, U+0323, U+0329, U+1EA0-1EF9, U+20AB;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fChc4EsA.woff2) format('woff2');unicode-range:U+0100-02AF, U+0304, U+0308, U+0329, U+1E00-1E9F, U+1EF2-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:500;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmEU9fBBc4.woff2) format('woff2');unicode-range:U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfCRc4EsA.woff2) format('woff2');unicode-range:U+0460-052F, U+1C80-1C88, U+20B4, U+2DE0-2DFF, U+A640-A69F, U+FE2E-FE2F;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfABc4EsA.woff2) format('woff2');unicode-range:U+0301, U+0400-045F, U+0490-0491, U+04B0-04B1, U+2116;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfCBc4EsA.woff2) format('woff2');unicode-range:U+1F00-1FFF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfBxc4EsA.woff2) format('woff2');unicode-range:U+0370-03FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfCxc4EsA.woff2) format('woff2');unicode-range:U+0102-0103, U+0110-0111, U+0128-0129, U+0168-0169, U+01A0-01A1, U+01AF-01B0, U+0300-0301, U+0303-0304, U+0308-0309, U+0323, U+0329, U+1EA0-1EF9, U+20AB;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfChc4EsA.woff2) format('woff2');unicode-range:U+0100-02AF, U+0304, U+0308, U+0329, U+1E00-1E9F, U+1EF2-1EFF, U+2020, U+20A0-20AB, U+20AD-20CF, U+2113, U+2C60-2C7F, U+A720-A7FF;}@font-face{font-family:'Roboto';font-style:normal;font-weight:700;font-display:swap;src:url(https://fonts.gstatic.com/s/roboto/v30/KFOlCnqEu92Fr1MmWUlfBBc4.woff2) format('woff2');unicode-range:U+0000-00FF, U+0131, U+0152-0153, U+02BB-02BC, U+02C6, U+02DA, U+02DC, U+0304, U+0308, U+0329, U+2000-206F, U+2074, U+20AC, U+2122, U+2191, U+2193, U+2212, U+2215, U+FEFF, U+FFFD;}</style>\n",
       "<style type=\"text/css\">@font-face{font-family:'Material Icons';font-style:normal;font-weight:400;src:url(https://fonts.gstatic.com/s/materialicons/v141/flUhRq6tzZclQEJ-Vdg-IuiaDsNc.woff2) format('woff2');}.material-icons{font-family:'Material Icons';font-weight:normal;font-style:normal;font-size:24px;line-height:1;letter-spacing:normal;text-transform:none;display:inline-block;white-space:nowrap;word-wrap:normal;direction:ltr;-webkit-font-feature-settings:'liga';-webkit-font-smoothing:antialiased;}</style>\n",
       "<script src=\"https://maps.googleapis.com/maps/api/js?key=AIzaSyC3PtSiQNBORX_rM0MRLHNA9VybME6zF1A&amp;libraries=places&amp;region=US&amp;language=en\"></script>\n",
       "<title>ClinicalTrials.gov</title>\n",
       "<base href=\"/\"/>\n",
       "<!-- Favicons -->\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"shortcut icon\" type=\"image/svg\"/>\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"icon\" type=\"image/svg\"/>\n",
       "<!-- 192x192, as recommended for Android\n",
       "  http://updates.html5rocks.com/2014/11/Support-for-theme-color-in-Chrome-39-for-Android\n",
       "  -->\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"icon\" sizes=\"192x192\" type=\"image/svg\"/>\n",
       "<!-- 57x57 (precomposed) for iPhone 3GS, pre-2011 iPod Touch and older Android devices -->\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"apple-touch-icon-precomposed\"/>\n",
       "<!-- 72x72 (precomposed) for 1st generation iPad, iPad 2 and iPad mini -->\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"apple-touch-icon-precomposed\" sizes=\"72x72\"/>\n",
       "<!-- 114x114 (precomposed) for iPhone 4, 4S, 5 and post-2011 iPod Touch -->\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"apple-touch-icon-precomposed\" sizes=\"114x114\"/>\n",
       "<!-- 144x144 (precomposed) for iPad 3rd and 4th generation -->\n",
       "<link href=\"assets/favicons/favicon.svg\" rel=\"apple-touch-icon-precomposed\" sizes=\"144x144\"/>\n",
       "<!-- Page description -->\n",
       "<meta content=\"\" name=\"description\"/>\n",
       "<meta content=\"\" property=\"og:title\"/>\n",
       "<meta content=\"\" property=\"og:description\"/>\n",
       "<meta content=\"\" name=\"twitter:title\"/>\n",
       "<meta content=\"\" name=\"twitter:description\"/>\n",
       "<meta content=\"summary\" name=\"twitter:card\"/>\n",
       "<!-- robots -->\n",
       "<meta content=\"index, follow\" name=\"robots\"/>\n",
       "<!-- Pinger -->\n",
       "<meta content=\"ctgov2\" name=\"ncbi_app\"/>\n",
       "<meta content=\"\" name=\"ncbi_db\"/>\n",
       "<meta content=\"false\" name=\"ncbi_pinger_track_stat\"/>\n",
       "<meta content=\"true\" name=\"ncbi_pinger_ga_track\"/>\n",
       "<meta content=\"UA-68648221-4\" name=\"ncbi_pinger_ga_tracking_id\"/>\n",
       "<meta content=\"lat,lng,locStr\" name=\"ncbi_pinger_url_query_strip\"/>\n",
       "<meta content=\"true\" name=\"ncbi_pinger_click_use_capture\"/>\n",
       "<meta content=\"true\" name=\"ncbi_pinger_gtm_track\"/>\n",
       "<meta content=\"\" name=\"ncbi_app_version\"/>\n",
       "<style>:root{font-family:-apple-system,BlinkMacSystemFont,Segoe UI,Roboto,Helvetica,Arial,sans-serif,\"Apple Color Emoji\",\"Segoe UI Emoji\",Segoe UI Symbol;--font-family:-apple-system, BlinkMacSystemFont, \"Segoe UI\", Roboto, Helvetica, Arial, sans-serif, \"Apple Color Emoji\", \"Segoe UI Emoji\", \"Segoe UI Symbol\";--surface-a:#ffffff;--surface-b:#f8f9fa;--surface-c:#e9ecef;--surface-d:#dee2e6;--surface-e:#ffffff;--surface-f:#ffffff;--text-color:#495057;--text-color-secondary:#6c757d;--primary-color:#2196F3;--primary-color-text:#ffffff;--surface-0:#ffffff;--surface-50:#fafafa;--surface-100:#f5f5f5;--surface-200:#eeeeee;--surface-300:#e0e0e0;--surface-400:#bdbdbd;--surface-500:#9e9e9e;--surface-600:#757575;--surface-700:#616161;--surface-800:#424242;--surface-900:#212121;--gray-50:#fafafa;--gray-100:#f5f5f5;--gray-200:#eeeeee;--gray-300:#e0e0e0;--gray-400:#bdbdbd;--gray-500:#9e9e9e;--gray-600:#757575;--gray-700:#616161;--gray-800:#424242;--gray-900:#212121;--content-padding:1rem;--inline-spacing:.5rem;--border-radius:3px;--surface-ground:#f8f9fa;--surface-section:#ffffff;--surface-card:#ffffff;--surface-overlay:#ffffff;--surface-border:#dee2e6;--surface-hover:#e9ecef;--maskbg:rgba(0, 0, 0, .4);--highlight-bg:#E3F2FD;--highlight-text-color:#495057;--focus-ring:0 0 0 .2rem #a6d5fa;color-scheme:light}:root{--blue-50:#f4fafe;--blue-100:#cae6fc;--blue-200:#a0d2fa;--blue-300:#75bef8;--blue-400:#4baaf5;--blue-500:#2196f3;--blue-600:#1c80cf;--blue-700:#1769aa;--blue-800:#125386;--blue-900:#0d3c61;--green-50:#f6fbf6;--green-100:#d4ecd5;--green-200:#b2ddb4;--green-300:#90cd93;--green-400:#6ebe71;--green-500:#4caf50;--green-600:#419544;--green-700:#357b38;--green-800:#2a602c;--green-900:#1e4620;--yellow-50:#fffcf5;--yellow-100:#fef0cd;--yellow-200:#fde4a5;--yellow-300:#fdd87d;--yellow-400:#fccc55;--yellow-500:#fbc02d;--yellow-600:#d5a326;--yellow-700:#b08620;--yellow-800:#8a6a19;--yellow-900:#644d12;--cyan-50:#f2fcfd;--cyan-100:#c2eff5;--cyan-200:#91e2ed;--cyan-300:#61d5e4;--cyan-400:#30c9dc;--cyan-500:#00bcd4;--cyan-600:#00a0b4;--cyan-700:#008494;--cyan-800:#006775;--cyan-900:#004b55;--pink-50:#fef4f7;--pink-100:#fac9da;--pink-200:#f69ebc;--pink-300:#f1749e;--pink-400:#ed4981;--pink-500:#e91e63;--pink-600:#c61a54;--pink-700:#a31545;--pink-800:#801136;--pink-900:#5d0c28;--indigo-50:#f5f6fb;--indigo-100:#d1d5ed;--indigo-200:#acb4df;--indigo-300:#8893d1;--indigo-400:#6372c3;--indigo-500:#3f51b5;--indigo-600:#36459a;--indigo-700:#2c397f;--indigo-800:#232d64;--indigo-900:#192048;--teal-50:#f2faf9;--teal-100:#c2e6e2;--teal-200:#91d2cc;--teal-300:#61beb5;--teal-400:#30aa9f;--teal-500:#009688;--teal-600:#008074;--teal-700:#00695f;--teal-800:#00534b;--teal-900:#003c36;--orange-50:#fff8f2;--orange-100:#fde0c2;--orange-200:#fbc791;--orange-300:#f9ae61;--orange-400:#f79530;--orange-500:#f57c00;--orange-600:#d06900;--orange-700:#ac5700;--orange-800:#874400;--orange-900:#623200;--bluegray-50:#f7f9f9;--bluegray-100:#d9e0e3;--bluegray-200:#bbc7cd;--bluegray-300:#9caeb7;--bluegray-400:#7e96a1;--bluegray-500:#607d8b;--bluegray-600:#526a76;--bluegray-700:#435861;--bluegray-800:#35454c;--bluegray-900:#263238;--purple-50:#faf4fb;--purple-100:#e7cbec;--purple-200:#d4a2dd;--purple-300:#c279ce;--purple-400:#af50bf;--purple-500:#9c27b0;--purple-600:#852196;--purple-700:#6d1b7b;--purple-800:#561561;--purple-900:#3e1046;--red-50:#fff5f5;--red-100:#ffd1ce;--red-200:#ffada7;--red-300:#ff8980;--red-400:#ff6459;--red-500:#ff4032;--red-600:#d9362b;--red-700:#b32d23;--red-800:#8c231c;--red-900:#661a14;--primary-50:#f4fafe;--primary-100:#cae6fc;--primary-200:#a0d2fa;--primary-300:#75bef8;--primary-400:#4baaf5;--primary-500:#2196f3;--primary-600:#1c80cf;--primary-700:#1769aa;--primary-800:#125386;--primary-900:#0d3c61}@layer primeng{*{box-sizing:border-box}}@layer primeng{button{border-radius:0}}:root{--fa-style-family-brands:\"Font Awesome 6 Brands\";--fa-font-brands:normal 400 1em/1 \"Font Awesome 6 Brands\"}@font-face{font-family:\"Font Awesome 6 Brands\";font-style:normal;font-weight:400;font-display:block;src:url(fa-brands-400.e465758e13df1d72.woff2) format(\"woff2\"),url(fa-brands-400.8bf6ccf0176a7567.ttf) format(\"truetype\")}:root{--fa-style-family-classic:\"Font Awesome 6 Free\";--fa-font-regular:normal 400 1em/1 \"Font Awesome 6 Free\"}@font-face{font-family:\"Font Awesome 6 Free\";font-style:normal;font-weight:400;font-display:block;src:url(fa-regular-400.f386b6b7c6bf65a5.woff2) format(\"woff2\"),url(fa-regular-400.a937b755f79cb30d.ttf) format(\"truetype\")}:root{--fa-style-family-classic:\"Font Awesome 6 Free\";--fa-font-solid:normal 900 1em/1 \"Font Awesome 6 Free\"}@font-face{font-family:\"Font Awesome 6 Free\";font-style:normal;font-weight:900;font-display:block;src:url(fa-solid-900.ce4938a30389d90a.woff2) format(\"woff2\"),url(fa-solid-900.081655f2111aa667.ttf) format(\"truetype\")}html{line-height:1.15;-webkit-text-size-adjust:100%}body{margin:0}a{background-color:transparent}button,input{font-family:inherit;font-size:100%;line-height:1.15;margin:0}button,input{overflow:visible}button{text-transform:none}button{-webkit-appearance:button}button::-moz-focus-inner{border-style:none;padding:0}button:-moz-focusring{outline:1px dotted ButtonText}[type=search]{-webkit-appearance:textfield;outline-offset:-2px}[type=search]::-webkit-search-decoration{-webkit-appearance:none}body{color:#1b1b1b;background-color:#fff;overflow-x:hidden}input:not([disabled]):focus,button:not([disabled]):focus{outline:.25rem solid #2491ff;outline-offset:0}[href]:focus{outline:.25rem solid #2491ff;outline-offset:0}*,*:before,*:after{box-sizing:inherit}html{font-feature-settings:\"kern\" 1;font-kerning:normal;font-family:Roboto,sans-serif;font-size:100%}.usa-sr-only{position:absolute;left:-999em;right:auto}.usa-dark-background{-moz-osx-font-smoothing:grayscale;-webkit-font-smoothing:antialiased;background-color:#3d4551}.usa-dark-background p,.usa-dark-background span,.usa-dark-background a,.usa-dark-background a:visited{color:#fff}.usa-dark-background a:hover{color:#f0f0f0}.usa-icon{display:inline-block;fill:currentColor;height:1em;position:relative;width:1em}input[type=search]{box-sizing:border-box;appearance:none}[type=search]{padding-bottom:0;padding-top:0;border-bottom-right-radius:0;border-right:none;border-top-right-radius:0;box-sizing:border-box;float:left;font-size:.93rem;height:2rem;margin:0}:root{--ctgl-color-base-lightest:#f0f0f0;--ctgl-color-base-lighter:#dfe1e2;--ctgl-color-base-light:#a9aeb1;--ctgl-color-base:#71767a;--ctgl-color-base-dark:#565c65;--ctgl-color-base-darker:#3d4551;--ctgl-color-base-darkest:#1b1b1b;--ctgl-color-primary-lighter:#d9e8f6;--ctgl-color-primary-light:#73b3e7;--ctgl-color-primary:#0071bc;--ctgl-color-primary-vivid:#0050d8;--ctgl-color-primary-dark:#1a4480;--ctgl-color-primary-darker:#205493;--ctgl-color-primary-darkest:#112e51;--ctgl-color-secondary-lighter:#f3e1e4;--ctgl-color-secondary-light:#f2938c;--ctgl-color-secondary:#d83933;--ctgl-color-secondary-vivid:#e41d3d;--ctgl-color-secondary-dark:#b50909;--ctgl-color-secondary-darker:#8b0a03;--ctgl-color-accent-warm-lighter:#f2e4d4;--ctgl-color-accent-warm-light:#ffbc78;--ctgl-color-accent-warm:#fa9441;--ctgl-color-accent-warm-dark:#c05600;--ctgl-color-accent-warm-darker:#775540;--ctgl-color-accent-cool-lighter:#e1f3f8;--ctgl-color-accent-cool-light:#97d4ea;--ctgl-color-accent-cool:#00bde3;--ctgl-color-accent-cool-dark:#28a0cb;--ctgl-color-accent-cool-darker:#07648d;--ctgl-color-error-lighter:#f4e3db;--ctgl-color-error-light:#f39268;--ctgl-color-error:#d54309;--ctgl-color-error-dark:#b50909;--ctgl-color-error-darker:#6f3331;--ctgl-color-warning-lighter:#faf3d1;--ctgl-color-warning-light:#fee685;--ctgl-color-warning:#ffbe2e;--ctgl-color-warning-dark:#e5a000;--ctgl-color-warning-darker:#936f38;--ctgl-color-success-lighter:#ecf3ec;--ctgl-color-success-light:#70e17b;--ctgl-color-success:#00a91c;--ctgl-color-success-dark:#008817;--ctgl-color-success-darker:#216e1f;--ctgl-color-info-lighter:#e7f6f8;--ctgl-color-info-light:#99deea;--ctgl-color-info:#00bde3;--ctgl-color-info-dark:#009ec1;--ctgl-color-info-darker:#2e6276;--ctgl-color-disabled-light:#e6e6e6;--ctgl-color-disabled:#c9c9c9;--ctgl-color-disabled-dark:#adadad;--ctgl-color-emergency:#9c3d10;--ctgl-color-emergency-dark:#332d29;--ctgl-color-red-cool-5:#f8eff1;--ctgl-color-red-cool-10:#f3e1e4;--ctgl-color-red-cool-20:#ecbec6;--ctgl-color-red-cool-30:#e09aa6;--ctgl-color-red-cool-40:#e16b80;--ctgl-color-red-cool-50:#cd425b;--ctgl-color-red-cool-60:#9e394b;--ctgl-color-red-cool-70:#68363f;--ctgl-color-red-cool-80:#40282c;--ctgl-color-red-cool-90:#1e1517;--ctgl-color-red-cool-5v:#fff2f5;--ctgl-color-red-cool-10v:#f8dfe2;--ctgl-color-red-cool-20v:#f8b9c5;--ctgl-color-red-cool-30v:#fd8ba0;--ctgl-color-red-cool-40v:#f45d79;--ctgl-color-red-cool-50v:#e41d3d;--ctgl-color-red-cool-60v:#b21d38;--ctgl-color-red-cool-70v:#822133;--ctgl-color-red-cool-80v:#4f1c24;--ctgl-color-red-5:#f9eeee;--ctgl-color-red-10:#f8e1de;--ctgl-color-red-20:#f7bbb1;--ctgl-color-red-30:#f2938c;--ctgl-color-red-40:#e9695f;--ctgl-color-red-50:#d83933;--ctgl-color-red-60:#a23737;--ctgl-color-red-70:#6f3331;--ctgl-color-red-80:#3e2927;--ctgl-color-red-90:#1b1616;--ctgl-color-red-5v:#fff3f2;--ctgl-color-red-10v:#fde0db;--ctgl-color-red-20v:#fdb8ae;--ctgl-color-red-30v:#ff8d7b;--ctgl-color-red-40v:#fb5a47;--ctgl-color-red-50v:#e52207;--ctgl-color-red-60v:#b50909;--ctgl-color-red-70v:#8b0a03;--ctgl-color-red-80v:#5c1111;--ctgl-color-red-warm-5:#f6efea;--ctgl-color-red-warm-10:#f4e3db;--ctgl-color-red-warm-20:#ecc0a7;--ctgl-color-red-warm-30:#dca081;--ctgl-color-red-warm-40:#d27a56;--ctgl-color-red-warm-50:#c3512c;--ctgl-color-red-warm-60:#805039;--ctgl-color-red-warm-70:#524236;--ctgl-color-red-warm-80:#332d29;--ctgl-color-red-warm-90:#1f1c18;--ctgl-color-red-warm-5v:#fff5ee;--ctgl-color-red-warm-10v:#fce1d4;--ctgl-color-red-warm-20v:#f6bd9c;--ctgl-color-red-warm-30v:#f39268;--ctgl-color-red-warm-40v:#ef5e25;--ctgl-color-red-warm-50v:#d54309;--ctgl-color-red-warm-60v:#9c3d10;--ctgl-color-red-warm-70v:#63340f;--ctgl-color-red-warm-80v:#3e2a1e;--ctgl-color-orange-warm-5:#faeee5;--ctgl-color-orange-warm-10:#fbe0d0;--ctgl-color-orange-warm-20:#f7bca2;--ctgl-color-orange-warm-30:#f3966d;--ctgl-color-orange-warm-40:#e17141;--ctgl-color-orange-warm-50:#bd5727;--ctgl-color-orange-warm-60:#914734;--ctgl-color-orange-warm-70:#633a32;--ctgl-color-orange-warm-80:#3d2925;--ctgl-color-orange-warm-90:#1c1615;--ctgl-color-orange-warm-5v:#fff3ea;--ctgl-color-orange-warm-10v:#ffe2d1;--ctgl-color-orange-warm-20v:#fbbaa7;--ctgl-color-orange-warm-30v:#fc906d;--ctgl-color-orange-warm-40v:#ff580a;--ctgl-color-orange-warm-50v:#cf4900;--ctgl-color-orange-warm-60v:#a72f10;--ctgl-color-orange-warm-70v:#782312;--ctgl-color-orange-warm-80v:#3d231d;--ctgl-color-orange-5:#f6efe9;--ctgl-color-orange-10:#f2e4d4;--ctgl-color-orange-20:#f3bf90;--ctgl-color-orange-30:#f09860;--ctgl-color-orange-40:#dd7533;--ctgl-color-orange-50:#a86437;--ctgl-color-orange-60:#775540;--ctgl-color-orange-70:#524236;--ctgl-color-orange-80:#332d27;--ctgl-color-orange-90:#1b1614;--ctgl-color-orange-5v:#fef2e4;--ctgl-color-orange-10v:#fce2c5;--ctgl-color-orange-20v:#ffbc78;--ctgl-color-orange-30v:#fa9441;--ctgl-color-orange-40v:#e66f0e;--ctgl-color-orange-50v:#c05600;--ctgl-color-orange-60v:#8c471c;--ctgl-color-orange-70v:#5f3617;--ctgl-color-orange-80v:#352313;--ctgl-color-gold-5:#f5f0e6;--ctgl-color-gold-10:#f1e5cd;--ctgl-color-gold-20:#dec69a;--ctgl-color-gold-30:#c7a97b;--ctgl-color-gold-40:#ad8b65;--ctgl-color-gold-50:#8e704f;--ctgl-color-gold-60:#6b5947;--ctgl-color-gold-70:#4d4438;--ctgl-color-gold-80:#322d26;--ctgl-color-gold-90:#191714;--ctgl-color-gold-5v:#fef0c8;--ctgl-color-gold-10v:#ffe396;--ctgl-color-gold-20v:#ffbe2e;--ctgl-color-gold-30v:#e5a000;--ctgl-color-gold-40v:#c2850c;--ctgl-color-gold-50v:#936f38;--ctgl-color-gold-60v:#7a591a;--ctgl-color-gold-70v:#5c410a;--ctgl-color-gold-80v:#3b2b15;--ctgl-color-yellow-5:#faf3d1;--ctgl-color-yellow-10:#f5e6af;--ctgl-color-yellow-20:#e6c74c;--ctgl-color-yellow-30:#c9ab48;--ctgl-color-yellow-40:#a88f48;--ctgl-color-yellow-50:#8a7237;--ctgl-color-yellow-60:#6b5a39;--ctgl-color-yellow-70:#504332;--ctgl-color-yellow-80:#332d27;--ctgl-color-yellow-90:#1a1614;--ctgl-color-yellow-5v:#fff5c2;--ctgl-color-yellow-10v:#fee685;--ctgl-color-yellow-20v:#face00;--ctgl-color-yellow-30v:#ddaa01;--ctgl-color-yellow-40v:#b38c00;--ctgl-color-yellow-50v:#947100;--ctgl-color-yellow-60v:#776017;--ctgl-color-yellow-70v:#5c4809;--ctgl-color-yellow-80v:#422d19;--ctgl-color-green-warm-5:#f1f4d7;--ctgl-color-green-warm-10:#e7eab7;--ctgl-color-green-warm-20:#cbd17a;--ctgl-color-green-warm-30:#a6b557;--ctgl-color-green-warm-40:#8a984b;--ctgl-color-green-warm-50:#6f7a41;--ctgl-color-green-warm-60:#5a5f38;--ctgl-color-green-warm-70:#45472f;--ctgl-color-green-warm-80:#2d2f21;--ctgl-color-green-warm-90:#171712;--ctgl-color-green-warm-5v:#f5fbc1;--ctgl-color-green-warm-10v:#e7f434;--ctgl-color-green-warm-20v:#c5d30a;--ctgl-color-green-warm-30v:#a3b72c;--ctgl-color-green-warm-40v:#7e9c1d;--ctgl-color-green-warm-50v:#6a7d00;--ctgl-color-green-warm-60v:#5a6613;--ctgl-color-green-warm-70v:#4b4e10;--ctgl-color-green-warm-80v:#38380b;--ctgl-color-green-5:#eaf4dd;--ctgl-color-green-10:#dfeacd;--ctgl-color-green-20:#b8d293;--ctgl-color-green-30:#9bb672;--ctgl-color-green-40:#7d9b4e;--ctgl-color-green-50:#607f35;--ctgl-color-green-60:#4c6424;--ctgl-color-green-70:#3c4a29;--ctgl-color-green-80:#293021;--ctgl-color-green-90:#161814;--ctgl-color-green-5v:#ddf9c7;--ctgl-color-green-10v:#c5ee93;--ctgl-color-green-20v:#98d035;--ctgl-color-green-30v:#7fb135;--ctgl-color-green-40v:#719f2a;--ctgl-color-green-50v:#538200;--ctgl-color-green-60v:#466c04;--ctgl-color-green-70v:#2f4a0b;--ctgl-color-green-80v:#243413;--ctgl-color-green-cool-5:#ecf3ec;--ctgl-color-green-cool-10:#dbebde;--ctgl-color-green-cool-20:#b4d0b9;--ctgl-color-green-cool-30:#86b98e;--ctgl-color-green-cool-40:#5e9f69;--ctgl-color-green-cool-50:#4d8055;--ctgl-color-green-cool-60:#446443;--ctgl-color-green-cool-70:#37493b;--ctgl-color-green-cool-80:#28312a;--ctgl-color-green-cool-90:#1a1f1a;--ctgl-color-green-cool-5v:#e3f5e1;--ctgl-color-green-cool-10v:#b7f5bd;--ctgl-color-green-cool-20v:#70e17b;--ctgl-color-green-cool-30v:#21c834;--ctgl-color-green-cool-40v:#00a91c;--ctgl-color-green-cool-50v:#008817;--ctgl-color-green-cool-60v:#216e1f;--ctgl-color-green-cool-70v:#154c21;--ctgl-color-green-cool-80v:#19311e;--ctgl-color-mint-5:#dbf6ed;--ctgl-color-mint-10:#c7efe2;--ctgl-color-mint-20:#92d9bb;--ctgl-color-mint-30:#5abf95;--ctgl-color-mint-40:#34a37e;--ctgl-color-mint-50:#2e8367;--ctgl-color-mint-60:#286846;--ctgl-color-mint-70:#204e34;--ctgl-color-mint-80:#193324;--ctgl-color-mint-90:#0d1a12;--ctgl-color-mint-5v:#c9fbeb;--ctgl-color-mint-10v:#83fcd4;--ctgl-color-mint-20v:#0ceda6;--ctgl-color-mint-30v:#04c585;--ctgl-color-mint-40v:#00a871;--ctgl-color-mint-50v:#008659;--ctgl-color-mint-60v:#146947;--ctgl-color-mint-70v:#0c4e29;--ctgl-color-mint-80v:#0d351e;--ctgl-color-mint-cool-5:#e0f7f6;--ctgl-color-mint-cool-10:#c4eeeb;--ctgl-color-mint-cool-20:#9bd4cf;--ctgl-color-mint-cool-30:#6fbab3;--ctgl-color-mint-cool-40:#4f9e99;--ctgl-color-mint-cool-50:#40807e;--ctgl-color-mint-cool-60:#376462;--ctgl-color-mint-cool-70:#2a4b45;--ctgl-color-mint-cool-80:#203131;--ctgl-color-mint-cool-90:#111818;--ctgl-color-mint-cool-5v:#d5fbf3;--ctgl-color-mint-cool-10v:#7efbe1;--ctgl-color-mint-cool-20v:#29e1cb;--ctgl-color-mint-cool-30v:#1dc2ae;--ctgl-color-mint-cool-40v:#00a398;--ctgl-color-mint-cool-50v:#008480;--ctgl-color-mint-cool-60v:#0f6460;--ctgl-color-mint-cool-70v:#0b4b3f;--ctgl-color-mint-cool-80v:#123131;--ctgl-color-cyan-5:#e7f6f8;--ctgl-color-cyan-10:#ccecf2;--ctgl-color-cyan-20:#99deea;--ctgl-color-cyan-30:#5dc0d1;--ctgl-color-cyan-40:#449dac;--ctgl-color-cyan-50:#168092;--ctgl-color-cyan-60:#2a646d;--ctgl-color-cyan-70:#2c4a4e;--ctgl-color-cyan-80:#203133;--ctgl-color-cyan-90:#111819;--ctgl-color-cyan-5v:#e5faff;--ctgl-color-cyan-10v:#a8f2ff;--ctgl-color-cyan-20v:#52daf2;--ctgl-color-cyan-30v:#00bde3;--ctgl-color-cyan-40v:#009ec1;--ctgl-color-cyan-50v:#0081a1;--ctgl-color-cyan-60v:#00687d;--ctgl-color-cyan-70v:#0e4f5c;--ctgl-color-cyan-80v:#093b44;--ctgl-color-blue-cool-5:#e7f2f5;--ctgl-color-blue-cool-10:#dae9ee;--ctgl-color-blue-cool-20:#adcfdc;--ctgl-color-blue-cool-30:#82b4c9;--ctgl-color-blue-cool-40:#6499af;--ctgl-color-blue-cool-50:#3a7d95;--ctgl-color-blue-cool-60:#2e6276;--ctgl-color-blue-cool-70:#224a58;--ctgl-color-blue-cool-80:#14333d;--ctgl-color-blue-cool-90:#0f191c;--ctgl-color-blue-cool-5v:#e1f3f8;--ctgl-color-blue-cool-10v:#c3ebfa;--ctgl-color-blue-cool-20v:#97d4ea;--ctgl-color-blue-cool-30v:#59b9de;--ctgl-color-blue-cool-40v:#28a0cb;--ctgl-color-blue-cool-50v:#0d7ea2;--ctgl-color-blue-cool-60v:#07648d;--ctgl-color-blue-cool-70v:#074b69;--ctgl-color-blue-cool-80v:#002d3f;--ctgl-color-blue-5:#eff6fb;--ctgl-color-blue-10:#d9e8f6;--ctgl-color-blue-20:#aacdec;--ctgl-color-blue-30:#73b3e7;--ctgl-color-blue-40:#4f97d1;--ctgl-color-blue-50:#2378c3;--ctgl-color-blue-60:#2c608a;--ctgl-color-blue-70:#274863;--ctgl-color-blue-80:#1f303e;--ctgl-color-blue-90:#11181d;--ctgl-color-blue-5v:#e8f5ff;--ctgl-color-blue-10v:#cfe8ff;--ctgl-color-blue-20v:#a1d3ff;--ctgl-color-blue-30v:#58b4ff;--ctgl-color-blue-40v:#2491ff;--ctgl-color-blue-50v:#0076d6;--ctgl-color-blue-60v:#005ea2;--ctgl-color-blue-70v:#0b4778;--ctgl-color-blue-80v:#112f4e;--ctgl-color-blue-warm-5:#ecf1f7;--ctgl-color-blue-warm-10:#e1e7f1;--ctgl-color-blue-warm-20:#bbcae4;--ctgl-color-blue-warm-30:#98afd2;--ctgl-color-blue-warm-40:#7292c7;--ctgl-color-blue-warm-50:#4a77b4;--ctgl-color-blue-warm-60:#345d96;--ctgl-color-blue-warm-70:#2f4668;--ctgl-color-blue-warm-80:#252f3e;--ctgl-color-blue-warm-90:#13171f;--ctgl-color-blue-warm-5v:#edf5ff;--ctgl-color-blue-warm-10v:#d4e5ff;--ctgl-color-blue-warm-20v:#adcdff;--ctgl-color-blue-warm-30v:#81aefc;--ctgl-color-blue-warm-40v:#5994f6;--ctgl-color-blue-warm-50v:#2672de;--ctgl-color-blue-warm-60v:#0050d8;--ctgl-color-blue-warm-70v:#1a4480;--ctgl-color-blue-warm-80v:#162e51;--ctgl-color-indigo-cool-5:#eef0f9;--ctgl-color-indigo-cool-10:#e1e6f9;--ctgl-color-indigo-cool-20:#bbc8f5;--ctgl-color-indigo-cool-30:#96abee;--ctgl-color-indigo-cool-40:#6b8ee8;--ctgl-color-indigo-cool-50:#496fd8;--ctgl-color-indigo-cool-60:#3f57a6;--ctgl-color-indigo-cool-70:#374274;--ctgl-color-indigo-cool-80:#292d42;--ctgl-color-indigo-cool-90:#151622;--ctgl-color-indigo-cool-5v:#edf0ff;--ctgl-color-indigo-cool-10v:#dee5ff;--ctgl-color-indigo-cool-20v:#b8c8ff;--ctgl-color-indigo-cool-30v:#94adff;--ctgl-color-indigo-cool-40v:#628ef4;--ctgl-color-indigo-cool-50v:#4866ff;--ctgl-color-indigo-cool-60v:#3e4ded;--ctgl-color-indigo-cool-70v:#222fbf;--ctgl-color-indigo-cool-80v:#1b2b85;--ctgl-color-indigo-5:#efeff8;--ctgl-color-indigo-10:#e5e4fa;--ctgl-color-indigo-20:#c5c5f3;--ctgl-color-indigo-30:#a5a8eb;--ctgl-color-indigo-40:#8889db;--ctgl-color-indigo-50:#676cc8;--ctgl-color-indigo-60:#4d52af;--ctgl-color-indigo-70:#3d4076;--ctgl-color-indigo-80:#2b2c40;--ctgl-color-indigo-90:#16171f;--ctgl-color-indigo-5v:#f0f0ff;--ctgl-color-indigo-10v:#e0e0ff;--ctgl-color-indigo-20v:#ccceff;--ctgl-color-indigo-30v:#a3a7fa;--ctgl-color-indigo-40v:#8289ff;--ctgl-color-indigo-50v:#656bd7;--ctgl-color-indigo-60v:#4a50c4;--ctgl-color-indigo-70v:#3333a3;--ctgl-color-indigo-80v:#212463;--ctgl-color-indigo-warm-5:#f1eff7;--ctgl-color-indigo-warm-10:#e7e3fa;--ctgl-color-indigo-warm-20:#cbc4f2;--ctgl-color-indigo-warm-30:#afa5e8;--ctgl-color-indigo-warm-40:#9287d8;--ctgl-color-indigo-warm-50:#7665d1;--ctgl-color-indigo-warm-60:#5e519e;--ctgl-color-indigo-warm-70:#453c7b;--ctgl-color-indigo-warm-80:#2e2c40;--ctgl-color-indigo-warm-90:#18161d;--ctgl-color-indigo-warm-5v:#f5f2ff;--ctgl-color-indigo-warm-10v:#e4deff;--ctgl-color-indigo-warm-20v:#cfc4fd;--ctgl-color-indigo-warm-30v:#b69fff;--ctgl-color-indigo-warm-40v:#967efb;--ctgl-color-indigo-warm-50v:#745fe9;--ctgl-color-indigo-warm-60v:#5942d2;--ctgl-color-indigo-warm-70v:#3d2c9d;--ctgl-color-indigo-warm-80v:#261f5b;--ctgl-color-violet-5:#f4f1f9;--ctgl-color-violet-10:#ebe3f9;--ctgl-color-violet-20:#d0c3e9;--ctgl-color-violet-30:#b8a2e3;--ctgl-color-violet-40:#9d84d2;--ctgl-color-violet-50:#8168b3;--ctgl-color-violet-60:#665190;--ctgl-color-violet-70:#4c3d69;--ctgl-color-violet-80:#312b3f;--ctgl-color-violet-90:#18161d;--ctgl-color-violet-5v:#f7f2ff;--ctgl-color-violet-10v:#ede3ff;--ctgl-color-violet-20v:#d5bfff;--ctgl-color-violet-30v:#c39deb;--ctgl-color-violet-40v:#ad79e9;--ctgl-color-violet-50v:#9355dc;--ctgl-color-violet-60v:#783cb9;--ctgl-color-violet-70v:#54278f;--ctgl-color-violet-80v:#39215e;--ctgl-color-violet-warm-5:#f8f0f9;--ctgl-color-violet-warm-10:#f6dff8;--ctgl-color-violet-warm-20:#e2bee4;--ctgl-color-violet-warm-30:#d29ad8;--ctgl-color-violet-warm-40:#bf77c8;--ctgl-color-violet-warm-50:#b04abd;--ctgl-color-violet-warm-60:#864381;--ctgl-color-violet-warm-70:#5c395a;--ctgl-color-violet-warm-80:#382936;--ctgl-color-violet-warm-90:#1b151b;--ctgl-color-violet-warm-5v:#fef2ff;--ctgl-color-violet-warm-10v:#fbdcff;--ctgl-color-violet-warm-20v:#f4b2ff;--ctgl-color-violet-warm-30v:#ee83ff;--ctgl-color-violet-warm-40v:#d85bef;--ctgl-color-violet-warm-50v:#be32d0;--ctgl-color-violet-warm-60v:#93348c;--ctgl-color-violet-warm-70v:#711e6c;--ctgl-color-violet-warm-80v:#481441;--ctgl-color-magenta-5:#f9f0f2;--ctgl-color-magenta-10:#f6e1e8;--ctgl-color-magenta-20:#f0bbcc;--ctgl-color-magenta-30:#e895b3;--ctgl-color-magenta-40:#e0699f;--ctgl-color-magenta-50:#c84281;--ctgl-color-magenta-60:#8b4566;--ctgl-color-magenta-70:#66364b;--ctgl-color-magenta-80:#402731;--ctgl-color-magenta-90:#1b1617;--ctgl-color-magenta-5v:#fff2f5;--ctgl-color-magenta-10v:#ffddea;--ctgl-color-magenta-20v:#ffb4cf;--ctgl-color-magenta-30v:#ff87b2;--ctgl-color-magenta-40v:#fd4496;--ctgl-color-magenta-50v:#d72d79;--ctgl-color-magenta-60v:#ab2165;--ctgl-color-magenta-70v:#731f44;--ctgl-color-magenta-80v:#4f172e;--ctgl-color-gray-cool-1:#fbfcfd;--ctgl-color-gray-cool-2:#f7f9fa;--ctgl-color-gray-cool-3:#f5f6f7;--ctgl-color-gray-cool-4:#f1f3f6;--ctgl-color-gray-cool-5:#edeff0;--ctgl-color-gray-cool-10:#dfe1e2;--ctgl-color-gray-cool-20:#c6cace;--ctgl-color-gray-cool-30:#a9aeb1;--ctgl-color-gray-cool-40:#8d9297;--ctgl-color-gray-cool-50:#71767a;--ctgl-color-gray-cool-60:#565c65;--ctgl-color-gray-cool-70:#3d4551;--ctgl-color-gray-cool-80:#2d2e2f;--ctgl-color-gray-cool-90:#1c1d1f;--ctgl-color-gray-1:#fcfcfc;--ctgl-color-gray-2:#f9f9f9;--ctgl-color-gray-3:#f6f6f6;--ctgl-color-gray-4:#f3f3f3;--ctgl-color-gray-5:#f0f0f0;--ctgl-color-gray-10:#e6e6e6;--ctgl-color-gray-20:#c9c9c9;--ctgl-color-gray-30:#adadad;--ctgl-color-gray-40:#919191;--ctgl-color-gray-50:#757575;--ctgl-color-gray-60:#5c5c5c;--ctgl-color-gray-70:#454545;--ctgl-color-gray-80:#2e2e2e;--ctgl-color-gray-90:#1b1b1b;--ctgl-color-gray-100:#000000;--ctgl-color-gray-warm-1:#fcfcfb;--ctgl-color-gray-warm-2:#f9f9f7;--ctgl-color-gray-warm-3:#f6f6f2;--ctgl-color-gray-warm-4:#f5f5f0;--ctgl-color-gray-warm-5:#f0f0ec;--ctgl-color-gray-warm-10:#e6e6e2;--ctgl-color-gray-warm-20:#cac9c0;--ctgl-color-gray-warm-30:#afaea2;--ctgl-color-gray-warm-40:#929285;--ctgl-color-gray-warm-50:#76766a;--ctgl-color-gray-warm-60:#5d5d52;--ctgl-color-gray-warm-70:#454540;--ctgl-color-gray-warm-80:#2e2e2a;--ctgl-color-gray-warm-90:#171716;--ctgl-color-black-transparent-5:rgba(0, 0, 0, .01);--ctgl-color-black-transparent-10:rgba(0, 0, 0, .1);--ctgl-color-black-transparent-20:rgba(0, 0, 0, .2);--ctgl-color-black-transparent-30:rgba(0, 0, 0, .3);--ctgl-color-black-transparent-40:rgba(0, 0, 0, .4);--ctgl-color-black-transparent-50:rgba(0, 0, 0, .5);--ctgl-color-black-transparent-60:rgba(0, 0, 0, .6);--ctgl-color-black-transparent-70:rgba(0, 0, 0, .7);--ctgl-color-black-transparent-80:rgba(0, 0, 0, .8);--ctgl-color-black-transparent-90:rgba(0, 0, 0, .9);--ctgl-color-white-transparent-5:rgba(255, 255, 255, .01);--ctgl-color-white-transparent-10:rgba(255, 255, 255, .1);--ctgl-color-white-transparent-20:rgba(255, 255, 255, .2);--ctgl-color-white-transparent-30:rgba(255, 255, 255, .3);--ctgl-color-white-transparent-40:rgba(255, 255, 255, .4);--ctgl-color-white-transparent-50:rgba(255, 255, 255, .5);--ctgl-color-white-transparent-60:rgba(255, 255, 255, .6);--ctgl-color-white-transparent-70:rgba(255, 255, 255, .7);--ctgl-color-white-transparent-80:rgba(255, 255, 255, .8);--ctgl-color-white-transparent-90:rgba(255, 255, 255, .9)}:root{--ctgl-breakpoint-card:10rem;--ctgl-breakpoint-card-lg:15rem;--ctgl-breakpoint-mobile:20rem;--ctgl-breakpoint-mobile-lg:30rem;--ctgl-breakpoint-tablet:40rem;--ctgl-breakpoint-tablet-lg:55rem;--ctgl-breakpoint-desktop:64rem;--ctgl-breakpoint-desktop-lg:75rem;--ctgl-breakpoint-widescreen:87.5rem}:root{font-family:Roboto,sans-serif}html{--mat-ripple-color:rgba(0, 0, 0, .1)}html{--mat-option-selected-state-label-text-color:#0071bc;--mat-option-label-text-color:rgba(0, 0, 0, .87);--mat-option-hover-state-layer-color:rgba(0, 0, 0, .04);--mat-option-focus-state-layer-color:rgba(0, 0, 0, .04);--mat-option-selected-state-layer-color:rgba(0, 0, 0, .04)}html{--mat-optgroup-label-text-color:rgba(0, 0, 0, .87)}html{--mat-full-pseudo-checkbox-selected-icon-color:#448aff;--mat-full-pseudo-checkbox-selected-checkmark-color:#fafafa;--mat-full-pseudo-checkbox-unselected-icon-color:rgba(0, 0, 0, .54);--mat-full-pseudo-checkbox-disabled-selected-checkmark-color:#fafafa;--mat-full-pseudo-checkbox-disabled-unselected-icon-color:#b0b0b0;--mat-full-pseudo-checkbox-disabled-selected-icon-color:#b0b0b0;--mat-minimal-pseudo-checkbox-selected-checkmark-color:#448aff;--mat-minimal-pseudo-checkbox-disabled-selected-checkmark-color:#b0b0b0}html{--mat-app-background-color:#fafafa;--mat-app-text-color:rgba(0, 0, 0, .87)}html{--mdc-elevated-card-container-shape:4px;--mdc-outlined-card-container-shape:4px;--mdc-outlined-card-outline-width:1px}html{--mdc-elevated-card-container-color:white;--mdc-elevated-card-container-elevation:0px 2px 1px -1px rgba(0, 0, 0, .2), 0px 1px 1px 0px rgba(0, 0, 0, .14), 0px 1px 3px 0px rgba(0, 0, 0, .12);--mdc-outlined-card-container-color:white;--mdc-outlined-card-outline-color:rgba(0, 0, 0, .12);--mdc-outlined-card-container-elevation:0px 0px 0px 0px rgba(0, 0, 0, .2), 0px 0px 0px 0px rgba(0, 0, 0, .14), 0px 0px 0px 0px rgba(0, 0, 0, .12);--mat-card-subtitle-text-color:rgba(0, 0, 0, .54)}html{--mdc-linear-progress-active-indicator-height:4px;--mdc-linear-progress-track-height:4px;--mdc-linear-progress-track-shape:0}html{--mdc-plain-tooltip-container-shape:4px;--mdc-plain-tooltip-supporting-text-line-height:16px}html{--mdc-plain-tooltip-container-color:#616161;--mdc-plain-tooltip-supporting-text-color:#fff}html{--mdc-filled-text-field-active-indicator-height:1px;--mdc-filled-text-field-focus-active-indicator-height:2px;--mdc-filled-text-field-container-shape:4px;--mdc-outlined-text-field-outline-width:1px;--mdc-outlined-text-field-focus-outline-width:2px;--mdc-outlined-text-field-container-shape:4px}html{--mdc-filled-text-field-caret-color:#0071bc;--mdc-filled-text-field-focus-active-indicator-color:#0071bc;--mdc-filled-text-field-focus-label-text-color:rgba(0, 113, 188, .87);--mdc-filled-text-field-container-color:whitesmoke;--mdc-filled-text-field-disabled-container-color:#fafafa;--mdc-filled-text-field-label-text-color:rgba(0, 0, 0, .6);--mdc-filled-text-field-disabled-label-text-color:rgba(0, 0, 0, .38);--mdc-filled-text-field-input-text-color:rgba(0, 0, 0, .87);--mdc-filled-text-field-disabled-input-text-color:rgba(0, 0, 0, .38);--mdc-filled-text-field-input-text-placeholder-color:rgba(0, 0, 0, .6);--mdc-filled-text-field-error-focus-label-text-color:#f44336;--mdc-filled-text-field-error-label-text-color:#f44336;--mdc-filled-text-field-error-caret-color:#f44336;--mdc-filled-text-field-active-indicator-color:rgba(0, 0, 0, .42);--mdc-filled-text-field-disabled-active-indicator-color:rgba(0, 0, 0, .06);--mdc-filled-text-field-hover-active-indicator-color:rgba(0, 0, 0, .87);--mdc-filled-text-field-error-active-indicator-color:#f44336;--mdc-filled-text-field-error-focus-active-indicator-color:#f44336;--mdc-filled-text-field-error-hover-active-indicator-color:#f44336;--mdc-outlined-text-field-caret-color:#0071bc;--mdc-outlined-text-field-focus-outline-color:#0071bc;--mdc-outlined-text-field-focus-label-text-color:rgba(0, 113, 188, .87);--mdc-outlined-text-field-label-text-color:rgba(0, 0, 0, .6);--mdc-outlined-text-field-disabled-label-text-color:rgba(0, 0, 0, .38);--mdc-outlined-text-field-input-text-color:rgba(0, 0, 0, .87);--mdc-outlined-text-field-disabled-input-text-color:rgba(0, 0, 0, .38);--mdc-outlined-text-field-input-text-placeholder-color:rgba(0, 0, 0, .6);--mdc-outlined-text-field-error-caret-color:#f44336;--mdc-outlined-text-field-error-focus-label-text-color:#f44336;--mdc-outlined-text-field-error-label-text-color:#f44336;--mdc-outlined-text-field-outline-color:rgba(0, 0, 0, .38);--mdc-outlined-text-field-disabled-outline-color:rgba(0, 0, 0, .06);--mdc-outlined-text-field-hover-outline-color:rgba(0, 0, 0, .87);--mdc-outlined-text-field-error-focus-outline-color:#f44336;--mdc-outlined-text-field-error-hover-outline-color:#f44336;--mdc-outlined-text-field-error-outline-color:#f44336;--mat-form-field-focus-select-arrow-color:rgba(0, 113, 188, .87);--mat-form-field-disabled-input-text-placeholder-color:rgba(0, 0, 0, .38);--mat-form-field-state-layer-color:rgba(0, 0, 0, .87);--mat-form-field-error-text-color:#f44336;--mat-form-field-select-option-text-color:inherit;--mat-form-field-select-disabled-option-text-color:GrayText;--mat-form-field-enabled-select-arrow-color:rgba(0, 0, 0, .54);--mat-form-field-disabled-select-arrow-color:rgba(0, 0, 0, .38);--mat-form-field-hover-state-layer-opacity:.04;--mat-form-field-focus-state-layer-opacity:.12}html{--mat-form-field-container-height:56px;--mat-form-field-filled-label-display:block;--mat-form-field-container-vertical-padding:16px;--mat-form-field-filled-with-label-container-padding-top:24px;--mat-form-field-filled-with-label-container-padding-bottom:8px}html{--mat-select-panel-background-color:white;--mat-select-enabled-trigger-text-color:rgba(0, 0, 0, .87);--mat-select-disabled-trigger-text-color:rgba(0, 0, 0, .38);--mat-select-placeholder-text-color:rgba(0, 0, 0, .6);--mat-select-enabled-arrow-color:rgba(0, 0, 0, .54);--mat-select-disabled-arrow-color:rgba(0, 0, 0, .38);--mat-select-focused-arrow-color:rgba(0, 113, 188, .87);--mat-select-invalid-arrow-color:rgba(244, 67, 54, .87)}html{--mat-autocomplete-background-color:white}html{--mdc-dialog-container-elevation-shadow:0px 11px 15px -7px rgba(0, 0, 0, .2), 0px 24px 38px 3px rgba(0, 0, 0, .14), 0px 9px 46px 8px rgba(0, 0, 0, .12);--mdc-dialog-container-shadow-color:#000;--mdc-dialog-container-shape:4px}html{--mdc-dialog-container-color:white;--mdc-dialog-subhead-color:rgba(0, 0, 0, .87);--mdc-dialog-supporting-text-color:rgba(0, 0, 0, .6)}html{--mdc-radio-disabled-selected-icon-opacity:.38;--mdc-radio-disabled-unselected-icon-opacity:.38;--mdc-radio-state-layer-size:40px}html{--mdc-radio-state-layer-size:40px}html{--mat-slider-value-indicator-width:auto;--mat-slider-value-indicator-height:32px;--mat-slider-value-indicator-caret-display:block;--mat-slider-value-indicator-border-radius:4px;--mat-slider-value-indicator-padding:0 12px;--mat-slider-value-indicator-text-transform:none;--mat-slider-value-indicator-container-transform:translateX(-50%);--mdc-slider-active-track-height:6px;--mdc-slider-active-track-shape:9999px;--mdc-slider-handle-height:20px;--mdc-slider-handle-shape:50%;--mdc-slider-handle-width:20px;--mdc-slider-inactive-track-height:4px;--mdc-slider-inactive-track-shape:9999px;--mdc-slider-with-overlap-handle-outline-width:1px;--mdc-slider-with-tick-marks-active-container-opacity:.6;--mdc-slider-with-tick-marks-container-shape:50%;--mdc-slider-with-tick-marks-container-size:2px;--mdc-slider-with-tick-marks-inactive-container-opacity:.6}html{--mdc-slider-handle-color:#0071bc;--mdc-slider-focus-handle-color:#0071bc;--mdc-slider-hover-handle-color:#0071bc;--mdc-slider-active-track-color:#0071bc;--mdc-slider-inactive-track-color:#0071bc;--mdc-slider-with-tick-marks-inactive-container-color:#0071bc;--mdc-slider-with-tick-marks-active-container-color:white;--mdc-slider-disabled-active-track-color:#000;--mdc-slider-disabled-handle-color:#000;--mdc-slider-disabled-inactive-track-color:#000;--mdc-slider-label-container-color:#000;--mdc-slider-label-label-text-color:#fff;--mdc-slider-with-overlap-handle-outline-color:#fff;--mdc-slider-with-tick-marks-disabled-container-color:#000;--mdc-slider-handle-elevation:0px 2px 1px -1px rgba(0, 0, 0, .2), 0px 1px 1px 0px rgba(0, 0, 0, .14), 0px 1px 3px 0px rgba(0, 0, 0, .12);--mat-slider-ripple-color:#0071bc;--mat-slider-hover-state-layer-color:rgba(0, 113, 188, .05);--mat-slider-focus-state-layer-color:rgba(0, 113, 188, .2);--mat-slider-value-indicator-opacity:.6}html{--mat-menu-container-shape:4px}html{--mat-menu-item-label-text-color:rgba(0, 0, 0, .87);--mat-menu-item-icon-color:rgba(0, 0, 0, .87);--mat-menu-item-hover-state-layer-color:rgba(0, 0, 0, .04);--mat-menu-item-focus-state-layer-color:rgba(0, 0, 0, .04);--mat-menu-container-color:white}html{--mdc-list-list-item-container-shape:0;--mdc-list-list-item-leading-avatar-shape:50%;--mdc-list-list-item-container-color:transparent;--mdc-list-list-item-selected-container-color:transparent;--mdc-list-list-item-leading-avatar-color:transparent;--mdc-list-list-item-leading-icon-size:24px;--mdc-list-list-item-leading-avatar-size:40px;--mdc-list-list-item-trailing-icon-size:24px;--mdc-list-list-item-disabled-state-layer-color:transparent;--mdc-list-list-item-disabled-state-layer-opacity:0;--mdc-list-list-item-disabled-label-text-opacity:.38;--mdc-list-list-item-disabled-leading-icon-opacity:.38;--mdc-list-list-item-disabled-trailing-icon-opacity:.38}html{--mdc-list-list-item-label-text-color:rgba(0, 0, 0, .87);--mdc-list-list-item-supporting-text-color:rgba(0, 0, 0, .54);--mdc-list-list-item-leading-icon-color:rgba(0, 0, 0, .38);--mdc-list-list-item-trailing-supporting-text-color:rgba(0, 0, 0, .38);--mdc-list-list-item-trailing-icon-color:rgba(0, 0, 0, .38);--mdc-list-list-item-selected-trailing-icon-color:rgba(0, 0, 0, .38);--mdc-list-list-item-disabled-label-text-color:black;--mdc-list-list-item-disabled-leading-icon-color:black;--mdc-list-list-item-disabled-trailing-icon-color:black;--mdc-list-list-item-hover-label-text-color:rgba(0, 0, 0, .87);--mdc-list-list-item-hover-leading-icon-color:rgba(0, 0, 0, .38);--mdc-list-list-item-hover-trailing-icon-color:rgba(0, 0, 0, .38);--mdc-list-list-item-focus-label-text-color:rgba(0, 0, 0, .87);--mdc-list-list-item-hover-state-layer-color:black;--mdc-list-list-item-hover-state-layer-opacity:.04;--mdc-list-list-item-focus-state-layer-color:black;--mdc-list-list-item-focus-state-layer-opacity:.12}html{--mdc-list-list-item-one-line-container-height:48px;--mdc-list-list-item-two-line-container-height:64px;--mdc-list-list-item-three-line-container-height:88px}html{--mat-paginator-container-text-color:rgba(0, 0, 0, .87);--mat-paginator-container-background-color:white;--mat-paginator-enabled-icon-color:rgba(0, 0, 0, .54);--mat-paginator-disabled-icon-color:rgba(0, 0, 0, .12)}html{--mat-paginator-container-size:56px}html{--mdc-tab-indicator-active-indicator-height:2px;--mdc-tab-indicator-active-indicator-shape:0;--mdc-secondary-navigation-tab-container-height:48px;--mat-tab-header-divider-color:transparent;--mat-tab-header-divider-height:0}html{--mdc-checkbox-disabled-selected-checkmark-color:#fff;--mdc-checkbox-selected-focus-state-layer-opacity:.16;--mdc-checkbox-selected-hover-state-layer-opacity:.04;--mdc-checkbox-selected-pressed-state-layer-opacity:.16;--mdc-checkbox-unselected-focus-state-layer-opacity:.16;--mdc-checkbox-unselected-hover-state-layer-opacity:.04;--mdc-checkbox-unselected-pressed-state-layer-opacity:.16}html{--mdc-checkbox-disabled-selected-icon-color:rgba(0, 0, 0, .38);--mdc-checkbox-disabled-unselected-icon-color:rgba(0, 0, 0, .38);--mdc-checkbox-selected-checkmark-color:#fff;--mdc-checkbox-selected-focus-icon-color:#448aff;--mdc-checkbox-selected-hover-icon-color:#448aff;--mdc-checkbox-selected-icon-color:#448aff;--mdc-checkbox-selected-pressed-icon-color:#448aff;--mdc-checkbox-unselected-focus-icon-color:#212121;--mdc-checkbox-unselected-hover-icon-color:#212121;--mdc-checkbox-unselected-icon-color:rgba(0, 0, 0, .54);--mdc-checkbox-unselected-pressed-icon-color:rgba(0, 0, 0, .54);--mdc-checkbox-selected-focus-state-layer-color:#448aff;--mdc-checkbox-selected-hover-state-layer-color:#448aff;--mdc-checkbox-selected-pressed-state-layer-color:#448aff;--mdc-checkbox-unselected-focus-state-layer-color:black;--mdc-checkbox-unselected-hover-state-layer-color:black;--mdc-checkbox-unselected-pressed-state-layer-color:black}html{--mdc-checkbox-state-layer-size:40px}html{--mdc-text-button-container-shape:4px;--mdc-text-button-keep-touch-target:false;--mdc-filled-button-container-shape:4px;--mdc-filled-button-keep-touch-target:false;--mdc-protected-button-container-shape:4px;--mdc-protected-button-keep-touch-target:false;--mdc-outlined-button-keep-touch-target:false;--mdc-outlined-button-outline-width:1px;--mdc-outlined-button-container-shape:4px}html{--mdc-text-button-label-text-color:#000;--mdc-text-button-disabled-label-text-color:rgba(0, 0, 0, .38);--mat-text-button-state-layer-color:#000;--mat-text-button-disabled-state-layer-color:#000;--mat-text-button-ripple-color:rgba(0, 0, 0, .1);--mat-text-button-hover-state-layer-opacity:.04;--mat-text-button-focus-state-layer-opacity:.12;--mat-text-button-pressed-state-layer-opacity:.12;--mdc-filled-button-container-color:white;--mdc-filled-button-label-text-color:#000;--mdc-filled-button-disabled-container-color:rgba(0, 0, 0, .12);--mdc-filled-button-disabled-label-text-color:rgba(0, 0, 0, .38);--mat-filled-button-state-layer-color:#000;--mat-filled-button-disabled-state-layer-color:#000;--mat-filled-button-ripple-color:rgba(0, 0, 0, .1);--mat-filled-button-hover-state-layer-opacity:.04;--mat-filled-button-focus-state-layer-opacity:.12;--mat-filled-button-pressed-state-layer-opacity:.12;--mdc-protected-button-container-color:white;--mdc-protected-button-label-text-color:#000;--mdc-protected-button-disabled-container-color:rgba(0, 0, 0, .12);--mdc-protected-button-disabled-label-text-color:rgba(0, 0, 0, .38);--mdc-protected-button-container-elevation-shadow:0px 3px 1px -2px rgba(0, 0, 0, .2), 0px 2px 2px 0px rgba(0, 0, 0, .14), 0px 1px 5px 0px rgba(0, 0, 0, .12);--mdc-protected-button-disabled-container-elevation-shadow:0px 0px 0px 0px rgba(0, 0, 0, .2), 0px 0px 0px 0px rgba(0, 0, 0, .14), 0px 0px 0px 0px rgba(0, 0, 0, .12);--mdc-protected-button-focus-container-elevation-shadow:0px 2px 4px -1px rgba(0, 0, 0, .2), 0px 4px 5px 0px rgba(0, 0, 0, .14), 0px 1px 10px 0px rgba(0, 0, 0, .12);--mdc-protected-button-hover-container-elevation-shadow:0px 2px 4px -1px rgba(0, 0, 0, .2), 0px 4px 5px 0px rgba(0, 0, 0, .14), 0px 1px 10px 0px rgba(0, 0, 0, .12);--mdc-protected-button-pressed-container-elevation-shadow:0px 5px 5px -3px rgba(0, 0, 0, .2), 0px 8px 10px 1px rgba(0, 0, 0, .14), 0px 3px 14px 2px rgba(0, 0, 0, .12);--mdc-protected-button-container-shadow-color:#000;--mat-protected-button-state-layer-color:#000;--mat-protected-button-disabled-state-layer-color:#000;--mat-protected-button-ripple-color:rgba(0, 0, 0, .1);--mat-protected-button-hover-state-layer-opacity:.04;--mat-protected-button-focus-state-layer-opacity:.12;--mat-protected-button-pressed-state-layer-opacity:.12;--mdc-outlined-button-disabled-outline-color:rgba(0, 0, 0, .12);--mdc-outlined-button-disabled-label-text-color:rgba(0, 0, 0, .38);--mdc-outlined-button-label-text-color:#000;--mdc-outlined-button-outline-color:rgba(0, 0, 0, .12);--mat-outlined-button-state-layer-color:#000;--mat-outlined-button-disabled-state-layer-color:#000;--mat-outlined-button-ripple-color:rgba(0, 0, 0, .1);--mat-outlined-button-hover-state-layer-opacity:.04;--mat-outlined-button-focus-state-layer-opacity:.12;--mat-outlined-button-pressed-state-layer-opacity:.12}html{--mdc-text-button-container-height:36px;--mdc-filled-button-container-height:36px;--mdc-outlined-button-container-height:36px;--mdc-protected-button-container-height:36px}html{--mdc-icon-button-state-layer-size:48px;--mdc-icon-button-icon-size:24px}html{--mdc-icon-button-icon-color:inherit;--mdc-icon-button-disabled-icon-color:rgba(0, 0, 0, .38);--mat-icon-button-state-layer-color:#000;--mat-icon-button-disabled-state-layer-color:#000;--mat-icon-button-ripple-color:rgba(0, 0, 0, .1);--mat-icon-button-hover-state-layer-opacity:.04;--mat-icon-button-focus-state-layer-opacity:.12;--mat-icon-button-pressed-state-layer-opacity:.12}html{--mdc-fab-container-shape:50%;--mdc-fab-icon-size:24px}html{--mdc-fab-container-color:white;--mdc-fab-container-elevation-shadow:0px 3px 5px -1px rgba(0, 0, 0, .2), 0px 6px 10px 0px rgba(0, 0, 0, .14), 0px 1px 18px 0px rgba(0, 0, 0, .12);--mdc-fab-focus-container-elevation-shadow:0px 5px 5px -3px rgba(0, 0, 0, .2), 0px 8px 10px 1px rgba(0, 0, 0, .14), 0px 3px 14px 2px rgba(0, 0, 0, .12);--mdc-fab-hover-container-elevation-shadow:0px 5px 5px -3px rgba(0, 0, 0, .2), 0px 8px 10px 1px rgba(0, 0, 0, .14), 0px 3px 14px 2px rgba(0, 0, 0, .12);--mdc-fab-pressed-container-elevation-shadow:0px 7px 8px -4px rgba(0, 0, 0, .2), 0px 12px 17px 2px rgba(0, 0, 0, .14), 0px 5px 22px 4px rgba(0, 0, 0, .12);--mdc-fab-container-shadow-color:#000;--mat-fab-foreground-color:black;--mat-fab-state-layer-color:#000;--mat-fab-disabled-state-layer-color:#000;--mat-fab-ripple-color:rgba(0, 0, 0, .1);--mat-fab-hover-state-layer-opacity:.04;--mat-fab-focus-state-layer-opacity:.12;--mat-fab-pressed-state-layer-opacity:.12;--mat-fab-disabled-state-container-color:rgba(0, 0, 0, .12);--mat-fab-disabled-state-foreground-color:rgba(0, 0, 0, .38);--mdc-extended-fab-container-elevation-shadow:0px 3px 5px -1px rgba(0, 0, 0, .2), 0px 6px 10px 0px rgba(0, 0, 0, .14), 0px 1px 18px 0px rgba(0, 0, 0, .12);--mdc-extended-fab-focus-container-elevation-shadow:0px 5px 5px -3px rgba(0, 0, 0, .2), 0px 8px 10px 1px rgba(0, 0, 0, .14), 0px 3px 14px 2px rgba(0, 0, 0, .12);--mdc-extended-fab-hover-container-elevation-shadow:0px 5px 5px -3px rgba(0, 0, 0, .2), 0px 8px 10px 1px rgba(0, 0, 0, .14), 0px 3px 14px 2px rgba(0, 0, 0, .12);--mdc-extended-fab-pressed-container-elevation-shadow:0px 7px 8px -4px rgba(0, 0, 0, .2), 0px 12px 17px 2px rgba(0, 0, 0, .14), 0px 5px 22px 4px rgba(0, 0, 0, .12);--mdc-extended-fab-container-shadow-color:#000}html{--mdc-snackbar-container-shape:4px}html{--mdc-snackbar-container-color:#333333;--mdc-snackbar-supporting-text-color:rgba(255, 255, 255, .87);--mat-snack-bar-button-color:#448aff}html{--mat-table-row-item-outline-width:1px}html{--mat-table-background-color:white;--mat-table-header-headline-color:rgba(0, 0, 0, .87);--mat-table-row-item-label-text-color:rgba(0, 0, 0, .87);--mat-table-row-item-outline-color:rgba(0, 0, 0, .12)}html{--mat-table-header-container-height:56px;--mat-table-footer-container-height:52px;--mat-table-row-item-container-height:52px}html{--mdc-circular-progress-active-indicator-width:4px;--mdc-circular-progress-size:48px}html{--mdc-circular-progress-active-indicator-color:#0071bc}html{--mat-badge-background-color:#0071bc;--mat-badge-text-color:white;--mat-badge-disabled-state-background-color:#b9b9b9;--mat-badge-disabled-state-text-color:rgba(0, 0, 0, .38)}html{--mat-bottom-sheet-container-shape:4px}html{--mat-bottom-sheet-container-text-color:rgba(0, 0, 0, .87);--mat-bottom-sheet-container-background-color:white}html{--mat-legacy-button-toggle-height:36px;--mat-legacy-button-toggle-shape:2px;--mat-legacy-button-toggle-focus-state-layer-opacity:1;--mat-standard-button-toggle-shape:4px;--mat-standard-button-toggle-hover-state-layer-opacity:.04;--mat-standard-button-toggle-focus-state-layer-opacity:.12}html{--mat-legacy-button-toggle-text-color:rgba(0, 0, 0, .38);--mat-legacy-button-toggle-state-layer-color:rgba(0, 0, 0, .12);--mat-legacy-button-toggle-selected-state-text-color:rgba(0, 0, 0, .54);--mat-legacy-button-toggle-selected-state-background-color:#e0e0e0;--mat-legacy-button-toggle-disabled-state-text-color:rgba(0, 0, 0, .26);--mat-legacy-button-toggle-disabled-state-background-color:#eeeeee;--mat-legacy-button-toggle-disabled-selected-state-background-color:#bdbdbd;--mat-standard-button-toggle-text-color:rgba(0, 0, 0, .87);--mat-standard-button-toggle-background-color:white;--mat-standard-button-toggle-state-layer-color:black;--mat-standard-button-toggle-selected-state-background-color:#e0e0e0;--mat-standard-button-toggle-selected-state-text-color:rgba(0, 0, 0, .87);--mat-standard-button-toggle-disabled-state-text-color:rgba(0, 0, 0, .26);--mat-standard-button-toggle-disabled-state-background-color:white;--mat-standard-button-toggle-disabled-selected-state-text-color:rgba(0, 0, 0, .87);--mat-standard-button-toggle-disabled-selected-state-background-color:#bdbdbd;--mat-standard-button-toggle-divider-color:#e0e0e0}html{--mat-standard-button-toggle-height:48px}html{--mat-datepicker-calendar-date-selected-state-text-color:white;--mat-datepicker-calendar-date-selected-state-background-color:#0071bc;--mat-datepicker-calendar-date-selected-disabled-state-background-color:rgba(0, 113, 188, .4);--mat-datepicker-calendar-date-today-selected-state-outline-color:white;--mat-datepicker-calendar-date-focus-state-background-color:rgba(0, 113, 188, .3);--mat-datepicker-calendar-date-hover-state-background-color:rgba(0, 113, 188, .3);--mat-datepicker-toggle-active-state-icon-color:#0071bc;--mat-datepicker-calendar-date-in-range-state-background-color:rgba(0, 113, 188, .2);--mat-datepicker-calendar-date-in-comparison-range-state-background-color:rgba(249, 171, 0, .2);--mat-datepicker-calendar-date-in-overlap-range-state-background-color:#a8dab5;--mat-datepicker-calendar-date-in-overlap-range-selected-state-background-color:#46a35e;--mat-datepicker-toggle-icon-color:rgba(0, 0, 0, .54);--mat-datepicker-calendar-body-label-text-color:rgba(0, 0, 0, .54);--mat-datepicker-calendar-period-button-icon-color:rgba(0, 0, 0, .54);--mat-datepicker-calendar-navigation-button-icon-color:rgba(0, 0, 0, .54);--mat-datepicker-calendar-header-divider-color:rgba(0, 0, 0, .12);--mat-datepicker-calendar-header-text-color:rgba(0, 0, 0, .54);--mat-datepicker-calendar-date-today-outline-color:rgba(0, 0, 0, .38);--mat-datepicker-calendar-date-today-disabled-state-outline-color:rgba(0, 0, 0, .18);--mat-datepicker-calendar-date-text-color:rgba(0, 0, 0, .87);--mat-datepicker-calendar-date-outline-color:transparent;--mat-datepicker-calendar-date-disabled-state-text-color:rgba(0, 0, 0, .38);--mat-datepicker-calendar-date-preview-state-outline-color:rgba(0, 0, 0, .24);--mat-datepicker-range-input-separator-color:rgba(0, 0, 0, .87);--mat-datepicker-range-input-disabled-state-separator-color:rgba(0, 0, 0, .38);--mat-datepicker-range-input-disabled-state-text-color:rgba(0, 0, 0, .38);--mat-datepicker-calendar-container-background-color:white;--mat-datepicker-calendar-container-text-color:rgba(0, 0, 0, .87)}html{--mat-divider-width:1px}html{--mat-divider-color:rgba(0, 0, 0, .12)}html{--mat-expansion-container-shape:4px}html{--mat-expansion-container-background-color:white;--mat-expansion-container-text-color:rgba(0, 0, 0, .87);--mat-expansion-actions-divider-color:rgba(0, 0, 0, .12);--mat-expansion-header-hover-state-layer-color:rgba(0, 0, 0, .04);--mat-expansion-header-focus-state-layer-color:rgba(0, 0, 0, .04);--mat-expansion-header-disabled-state-text-color:rgba(0, 0, 0, .26);--mat-expansion-header-text-color:rgba(0, 0, 0, .87);--mat-expansion-header-description-color:rgba(0, 0, 0, .54);--mat-expansion-header-indicator-color:rgba(0, 0, 0, .54)}html{--mat-expansion-header-collapsed-state-height:48px;--mat-expansion-header-expanded-state-height:64px}html{--mat-icon-color:inherit}html{--mat-sidenav-container-shape:0}html{--mat-sidenav-container-divider-color:rgba(0, 0, 0, .12);--mat-sidenav-container-background-color:white;--mat-sidenav-container-text-color:rgba(0, 0, 0, .87);--mat-sidenav-content-background-color:#fafafa;--mat-sidenav-content-text-color:rgba(0, 0, 0, .87);--mat-sidenav-scrim-color:rgba(0, 0, 0, .6)}html{--mat-stepper-header-icon-foreground-color:white;--mat-stepper-header-selected-state-icon-background-color:#0071bc;--mat-stepper-header-selected-state-icon-foreground-color:white;--mat-stepper-header-done-state-icon-background-color:#0071bc;--mat-stepper-header-done-state-icon-foreground-color:white;--mat-stepper-header-edit-state-icon-background-color:#0071bc;--mat-stepper-header-edit-state-icon-foreground-color:white;--mat-stepper-container-color:white;--mat-stepper-line-color:rgba(0, 0, 0, .12);--mat-stepper-header-hover-state-layer-color:rgba(0, 0, 0, .04);--mat-stepper-header-focus-state-layer-color:rgba(0, 0, 0, .04);--mat-stepper-header-label-text-color:rgba(0, 0, 0, .54);--mat-stepper-header-optional-label-text-color:rgba(0, 0, 0, .54);--mat-stepper-header-selected-state-label-text-color:rgba(0, 0, 0, .87);--mat-stepper-header-error-state-label-text-color:#f44336;--mat-stepper-header-icon-background-color:rgba(0, 0, 0, .54);--mat-stepper-header-error-state-icon-foreground-color:#f44336;--mat-stepper-header-error-state-icon-background-color:transparent}html{--mat-stepper-header-height:72px}html{--mat-sort-arrow-color:#757575}html{--mat-toolbar-container-background-color:whitesmoke;--mat-toolbar-container-text-color:rgba(0, 0, 0, .87)}html{--mat-toolbar-standard-height:64px;--mat-toolbar-mobile-height:56px}html{--mat-tree-container-background-color:white;--mat-tree-node-text-color:rgba(0, 0, 0, .87)}html{--mat-tree-node-min-height:48px}p{line-height:1.62!important}html{box-sizing:border-box}*,*:after,*:before{box-sizing:inherit}@media print{.glossary{display:none}.hide-glossary .usa-icon{height:1.5rem;width:1.5rem;margin-top:4px}}@charset \"UTF-8\";.glossary{background:#112e51;color:#fff;width:300px;position:fixed;top:0;bottom:0;right:0;overflow-y:auto;padding:2.4rem;transition:right .2s;z-index:10001}@media screen and (min-width: 481px){.glossary{width:364px}}.glossary ul{padding:0}.glossary input[type=search]{box-sizing:inherit;margin-bottom:4rem;min-width:220px}.glossary label{margin-top:2.6rem}.glossary a{color:#fff;text-decoration:underline}.glossary[aria-hidden=true]{right:-300px}@media screen and (min-width: 481px){.glossary[aria-hidden=true]{right:-364px}}.glossary-heading{display:inline;font-size:22px;margin-top:0}.js-glossary-toggle{border-bottom:none;color:#0071bc;font-size:1.5rem;font-weight:700;line-height:initial;width:initial}.js-glossary-toggle:hover{color:#205493}.js-glossary-toggle svg{margin-right:2px}.js-glossary-close{border:none;font-size:20px;background-color:transparent;float:right;width:initial}.js-glossary-search::-webkit-search-cancel-button{-webkit-appearance:searchfield-cancel-button!important}.usa-sr-only{position:absolute;left:-999em}.clearable-input{position:relative;display:inline-block}.clearable-input>input{padding-right:1.4em}.clearable-input>input::-ms-clear{display:none}</style><link href=\"styles.09afa4e08ac8f7f2.css\" media=\"print\" onload=\"this.media='all'\" rel=\"stylesheet\"/><noscript><link href=\"styles.09afa4e08ac8f7f2.css\" rel=\"stylesheet\"/></noscript></head>\n",
       "<body>\n",
       "<div aria-describedby=\"glossary-result\" aria-hidden=\"true\" class=\"glossary usa-dark-background\" id=\"glossary\">\n",
       "<button class=\"js-glossary-close button button--close js-glossary-toggle usa-button-unstyled hide-glossary\" title=\"Close glossary\"><span class=\"usa-sr-only\">Hide glossary</span>\n",
       "<svg aria-hidden=\"true\" class=\"usa-icon\" focusable=\"false\" role=\"img\">\n",
       "<use href=\"/assets/uswds/img/sprite.svg#close\"></use>\n",
       "</svg>\n",
       "</button>\n",
       "<h4 class=\"glossary-heading\">Glossary</h4>\n",
       "<p>\n",
       "  Study record managers: refer to the <a href=\"https://clinicaltrials.gov/ct2/manage-recs/resources#DataElement\"><span style=\"text-decoration:underline\">Data Element Definitions</span></a> if submitting registration or results information.\n",
       "  </p>\n",
       "<p>\n",
       "<label class=\"label\" for=\"glossary-search\">Search for terms</label>\n",
       "</p>\n",
       "<span class=\"clearable-input\">\n",
       "<input class=\"js-glossary-search\" id=\"glossary-search\" type=\"search\"/>\n",
       "</span>\n",
       "<div class=\"glossary__content\" id=\"glossary-result\">\n",
       "<ul class=\"js-glossary-list\"></ul>\n",
       "</div>\n",
       "</div>\n",
       "<app-root></app-root>\n",
       "<script src=\"https://www.ncbi.nlm.nih.gov/core/pinger/pinger.js\" type=\"text/javascript\"></script>\n",
       "<script src=\"runtime.4f8d0037c1387325.js\" type=\"module\"></script><script src=\"polyfills.ea43fa2cc74daa7d.js\" type=\"module\"></script><script defer=\"\" src=\"scripts.3c9d5dd337c20601.js\"></script><script src=\"main.0aa3758a8e7739cf.js\" type=\"module\"></script></body>\n",
       "<!-- JavaScript -->\n",
       "<script crossorigin=\"anonymous\" integrity=\"sha256-9/aliU8dGd2tb6OSsuzixeV4y/faTqgFtohetphbbj0=\" src=\"https://code.jquery.com/jquery-3.5.1.min.js\">\n",
       "</script>\n",
       "<script>\n",
       "  var fallbackJquery = \"https://www.ncbi.nlm.nih.gov/style-guide/static/base/js/jquery-3.5.0.min.js\";\n",
       "  window.jQuery || document.write(\"<script src=\" + fallbackJquery + \">\\x3C/script>\")\n",
       "</script>\n",
       "<!--<script  type=\"text/javascript\" src=\"assets/core/alerts/alerts.js\"> </script>-->\n",
       "<!--<script  type=\"text/javascript\" src=\"https://www.ncbi.nlm.nih.gov/core/pinger/pinger.js\"> </script>-->\n",
       "<!-- - - - - -\n",
       "\n",
       "- - - - - -->\n",
       "<script type=\"text/javascript\">\n",
       "  var alertsUrl = \"/core/alerts/alerts.js\";\n",
       "  if (typeof ncbiBaseUrl !== 'undefined') {\n",
       "    alertsUrl = ncbiBaseUrl + alertsUrl;\n",
       "  } else {\n",
       "    alertsUrl = \"https://www.ncbi.nlm.nih.gov\" + alertsUrl;\n",
       "  }\n",
       "</script>\n",
       "<!--\n",
       "<script src=\"https://www.ncbi.nlm.nih.gov/coreutils/ncbi-banner/ncbi-banner-experiment.js\"></script>\n",
       "-->\n",
       "</html>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "soup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "4a05901d",
   "metadata": {},
   "outputs": [
    {
     "ename": "SyntaxError",
     "evalue": "invalid syntax (1782545805.py, line 1)",
     "output_type": "error",
     "traceback": [
      "\u001b[1;36m  Cell \u001b[1;32mIn[11], line 1\u001b[1;36m\u001b[0m\n\u001b[1;33m    <table class=\"radius-md margin-right-05 condition-text-mark\" >\u001b[0m\n\u001b[1;37m    ^\u001b[0m\n\u001b[1;31mSyntaxError\u001b[0m\u001b[1;31m:\u001b[0m invalid syntax\n"
     ]
    }
   ],
   "source": [
    "<table class=\"radius-md margin-right-05 condition-text-mark\" >\n",
    "<caption>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "4129242d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "None\n"
     ]
    },
    {
     "ename": "TypeError",
     "evalue": "'NoneType' object is not subscriptable",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[28], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m \u001b[38;5;28;43mprint\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43msoup\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfind\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[38;5;124;43mtable\u001b[39;49m\u001b[38;5;124;43m'\u001b[39;49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\u001b[43m[\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\n",
      "\u001b[1;31mTypeError\u001b[0m: 'NoneType' object is not subscriptable"
     ]
    }
   ],
   "source": [
    "print(soup.find('table'))[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4853a716",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"Conditions\">\n",
       "    \n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%html\n",
    "<table class=\"Conditions\">\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "48bbf5ae",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        CUI  LAT  TS       LUI  STT       SUI  ISPREF        AUI  SAUI  \\\n",
      "0       CUI  LAT  TS       LUI  STT       SUI  ISPREF        AUI  SAUI   \n",
      "1  C0000005  ENG   P  L0000005   PF  S0007492       Y  A26634265   NaN   \n",
      "2  C0000005  ENG   S  L0270109   PF  S0007491       Y  A26634266   NaN   \n",
      "3  C0000005  FRE   P  L6220710   PF  S7133957       Y  A13433185   NaN   \n",
      "4  C0000005  FRE   S  L6215648   PF  S7133916       Y  A27488794   NaN   \n",
      "\n",
      "       SCUI     SDUI     SAB  TTY     CODE  \\\n",
      "0      SCUI     SDUI     SAB  TTY     CODE   \n",
      "1  M0019694  D012711     MSH  PEP  D012711   \n",
      "2  M0019694  D012711     MSH   ET  D012711   \n",
      "3  M0019694  D012711  MSHFRE  PEP  D012711   \n",
      "4  M0019694  D012711  MSHFRE   ET  D012711   \n",
      "\n",
      "                                             STR  SRL  SUPPRESS  SVF  \n",
      "0                                            STR  SRL  SUPPRESS  SVF  \n",
      "1                 (131)I-Macroaggregated Albumin    0         N  256  \n",
      "2                                     (131)I-MAA    0         N  256  \n",
      "3  Macroagrégats d'albumine marquée à l'iode 131    3         N  NaN  \n",
      "4                                      MAA-I 131    3         N  NaN  \n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "rrf_file_path = r\"C:\\Users\\annem\\Downloads\\MRCONSO.rrf\"  # Use 'r' before the string to indicate a raw string literal\n",
    "\n",
    "# Specify delimiter and column names based on the RRF format\n",
    "column_names = [\n",
    "    \"CUI\", \"LAT\", \"TS\", \"LUI\", \"STT\", \"SUI\", \"ISPREF\", \"AUI\", \"SAUI\", \"SCUI\", \n",
    "    \"SDUI\", \"SAB\", \"TTY\", \"CODE\", \"STR\", \"SRL\", \"SUPPRESS\", \"SVF\"\n",
    "]\n",
    "\n",
    "df_rrf = pd.read_csv(rrf_file_path, delimiter=\"|\", header=None, quoting=3, engine='python')\n",
    "\n",
    "\n",
    "if len(df_rrf.columns) == 19:\n",
    "    df_rrf = df_rrf.iloc[:, :-1]\n",
    "\n",
    "df_rrf.columns = column_names\n",
    "\n",
    "print(df_rrf.head())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "5ba1496d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of rows in the DataFrame: 16132275\n"
     ]
    }
   ],
   "source": [
    "row_count = df_rrf.shape[0]\n",
    "print(\"Number of rows in the DataFrame:\", row_count)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "72879bc0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "206d30cd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "6eed86a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "abc\n",
      "car\n",
      "ada\n",
      "racecar\n",
      "cool\n"
     ]
    }
   ],
   "source": [
    "words = [\"abc\",\"car\",\"ada\",\"racecar\",\"cool\"]\n",
    "for i in words:\n",
    "    print(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "9b02588c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ada\n"
     ]
    }
   ],
   "source": [
    "for i in words:\n",
    "    j = i\n",
    "    if j == i[::-1]:\n",
    "        print(j)\n",
    "        break\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cc6c8f31",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "beb3b98e",
   "metadata": {},
   "outputs": [],
   "source": [
    "nums1 = [1,2,2,1]\n",
    "nums2 = [2,2]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "9c845210",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n"
     ]
    }
   ],
   "source": [
    "print(len(nums1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "b4254825",
   "metadata": {},
   "outputs": [],
   "source": [
    "nums1=set(nums1)\n",
    "nums2=set(nums2)\n",
    "# print(nums1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "de9477d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{2}"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nums2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "fa00dd31",
   "metadata": {},
   "outputs": [],
   "source": [
    "re=[]\n",
    "for i in nums1:\n",
    "    for j in nums2:\n",
    "        if i == j:\n",
    "            re.append(i)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "427897f6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n"
     ]
    }
   ],
   "source": [
    "print(len(nums1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "858fb23e",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(nums+1)):\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d986d84",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(k+1):\n",
    "    nums1[]\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "d17212ba",
   "metadata": {},
   "outputs": [],
   "source": [
    "n = nums1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2e84e15d",
   "metadata": {},
   "outputs": [],
   "source": [
    "class Solution:\n",
    "    def isMonotonic(self, nums: List[int]) -> bool:\n",
    "        n = nums\n",
    "        increasing = decreasing = True\n",
    "\n",
    "        for i in range(len(n) - 1):\n",
    "            if n[i] > n[i + 1]:\n",
    "                increasing = False\n",
    "            if n[i] < n[i + 1]:\n",
    "                decreasing = False\n",
    "\n",
    "        return increasing or decreasing\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2c48ba7d",
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(nums1) - 1):\n",
    "    if n[i] > n[i + 1]:\n",
    "        increasing = False\n",
    "    if n[i] < n[i + 1]:\n",
    "        decreasing = False\n",
    "        \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "8c62dc52",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1, 2, 1, [...]]\n"
     ]
    }
   ],
   "source": [
    "n=[1,2,1]\n",
    "m=n\n",
    "m.append(n)\n",
    "print(m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "3a595d69",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "KeyboardInterrupt\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in n:\n",
    "    n.append(i)\n",
    "    n=n\n",
    "print(n)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c70d4122",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c69ebad5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SentenceTransformer(\n",
       "  (0): Transformer({'max_seq_length': 384, 'do_lower_case': False}) with Transformer model: MPNetModel \n",
       "  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False})\n",
       "  (2): Normalize()\n",
       ")"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sentence_transformers import SentenceTransformer, InputExample\n",
    "from torch.utils.data import DataLoader\n",
    "import torch\n",
    "from tqdm import tqdm\n",
    "excel_file_path1 =r\"C:\\Users\\annem\\Downloads\\cleaned_data.xlsx\"\n",
    "excel_file_path = r\"C:\\Users\\annem\\OneDrive\\Desktop\\ohdsi concepts\\Usagi_eval.xlsx\"\n",
    "df = pd.read_excel(excel_file_path)\n",
    "\n",
    "pretrained_model = \"sentence-transformers/all-mpnet-base-v2\"\n",
    "model = SentenceTransformer(pretrained_model)\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "model.to(device)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b06f7809",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ba72ce49",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 1/10: 100%|██████████| 2727/2727 [19:41<00:00,  2.31it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10, Average Loss: 0.08855246034367653\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 2/10: 100%|██████████| 2727/2727 [19:32<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2/10, Average Loss: 0.08855246080201717\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 3/10: 100%|██████████| 2727/2727 [19:34<00:00,  2.32it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 3/10, Average Loss: 0.0885524606597638\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 4/10: 100%|██████████| 2727/2727 [19:30<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 4/10, Average Loss: 0.08855246057951748\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 5/10: 100%|██████████| 2727/2727 [19:31<00:00,  2.33it/s] \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 5/10, Average Loss: 0.08855246076214482\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 6/10: 100%|██████████| 2727/2727 [19:30<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 6/10, Average Loss: 0.08855246067879281\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 7/10: 100%|██████████| 2727/2727 [19:31<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 7/10, Average Loss: 0.08855246091734323\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 8/10: 100%|██████████| 2727/2727 [19:31<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8/10, Average Loss: 0.08855246047538617\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 9/10: 100%|██████████| 2727/2727 [19:31<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 9/10, Average Loss: 0.08855246075399438\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 10/10: 100%|██████████| 2727/2727 [29:48<00:00,  1.52it/s]    "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 10/10, Average Loss: 0.08855246055970203\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "pretrained_model = \"sentence-transformers/all-mpnet-base-v2\"\n",
    "model = SentenceTransformer(pretrained_model)\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "model.to(device)\n",
    "\n",
    "class CustomDataset:\n",
    "    def __init__(self, entities, concepts):\n",
    "        self.examples = [InputExample(texts=[str(entity), str(concept)]) for entity, concept in zip(entities, concepts)]\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.examples)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        example = self.examples[idx]\n",
    "        entity_text = str(example.texts[0])\n",
    "        concept_text = str(example.texts[1])\n",
    "        return entity_text, concept_text\n",
    "\n",
    "def custom_loss(entity_embeddings, concept_embeddings, margin=0.5):\n",
    "    entity_embeddings = entity_embeddings.requires_grad_()\n",
    "    concept_embeddings = concept_embeddings.requires_grad_()\n",
    "    \n",
    "    distance = torch.nn.functional.pairwise_distance(entity_embeddings, concept_embeddings)\n",
    "    loss = torch.nn.functional.relu(margin - distance).mean()\n",
    "    return loss\n",
    "\n",
    "custom_dataset = CustomDataset(entities=df['Entity'].tolist(), concepts=df['Concept(gold)'].tolist())\n",
    "batch_size = 3\n",
    "dataloader = DataLoader(custom_dataset, batch_size=batch_size, shuffle=True)\n",
    "num_epochs = 10\n",
    "learning_rate = 0.01\n",
    "optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)\n",
    "\n",
    "model.train()\n",
    "for epoch in range(num_epochs):\n",
    "    total_loss = 0.0\n",
    "    for batch in tqdm(dataloader, desc=f\"Epoch {epoch + 1}/{num_epochs}\"):\n",
    "        entity_texts, concept_texts = batch\n",
    "        entity_embeddings = model.encode(entity_texts, convert_to_tensor=True, device=device)\n",
    "        concept_embeddings = model.encode(concept_texts, convert_to_tensor=True, device=device)\n",
    "\n",
    "        loss = custom_loss(entity_embeddings, concept_embeddings)\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        total_loss += loss.item()\n",
    "\n",
    "    average_loss = total_loss / len(dataloader)\n",
    "    print(f\"Epoch {epoch + 1}/{num_epochs}, Average Loss: {average_loss}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "734edaae",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SentenceTransformer(\n",
       "  (0): Transformer({'max_seq_length': 384, 'do_lower_case': False}) with Transformer model: MPNetModel \n",
       "  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False})\n",
       "  (2): Normalize()\n",
       ")"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.save(model.state_dict(), 'trained_model.pth')\n",
    "loaded_model = SentenceTransformer(pretrained_model)\n",
    "loaded_model.load_state_dict(torch.load('trained_model.pth'))\n",
    "loaded_model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "cb5cb7ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[0.7127]])\n",
      "tensor([[0.7563]])\n",
      "tensor([[0.6611]])\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "from sentence_transformers import SentenceTransformer, InputExample, losses, util\n",
    "from torch.utils.data import DataLoader\n",
    "import torch\n",
    "from tqdm import tqdm\n",
    "model=\n",
    "entity=loaded_model.encode(\"lobes of the lung\")\n",
    "concept1=loaded_model.encode(\"fusion of lobes of lung\")\n",
    "Gold_concept=loaded_model.encode(\"history of total lobectomy of lung\")\n",
    "concpt2=loaded_model.encode(\"Accessory lobe of lung\")\n",
    "\n",
    "print(util.cos_sim(entity, concept1))\n",
    "print(util.cos_sim(entity, concpt2))\n",
    "print(util.cos_sim(entity, Gold_concept))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "e7bd75ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.scatterplot(x=entity_embeddings.detach().numpy()[:, 0], y=entity_embeddings.detach().numpy()[:, 1], label='Entity')\n",
    "sns.scatterplot(x=concept_embeddings.detach().numpy()[:, 0], y=concept_embeddings.detach().numpy()[:, 1], label='Gold Concept')\n",
    "\n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "5e5c89d3",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 1/3: 100%|██████████| 2727/2727 [09:13<00:00,  4.93it/s]\n",
      "Epoch 2/3: 100%|██████████| 2727/2727 [10:26<00:00,  4.35it/s]\n",
      "Epoch 3/3: 100%|██████████| 2727/2727 [10:04<00:00,  4.51it/s]\n"
     ]
    },
    {
     "ename": "NameError",
     "evalue": "name 'np' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[21], line 60\u001b[0m\n\u001b[0;32m     57\u001b[0m         loss\u001b[38;5;241m.\u001b[39mbackward()\n\u001b[0;32m     58\u001b[0m         optimizer\u001b[38;5;241m.\u001b[39mstep()\n\u001b[1;32m---> 60\u001b[0m all_embeddings \u001b[38;5;241m=\u001b[39m \u001b[43mnp\u001b[49m\u001b[38;5;241m.\u001b[39mconcatenate(all_embeddings, axis\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0\u001b[39m)\n\u001b[0;32m     62\u001b[0m tsne_embeddings \u001b[38;5;241m=\u001b[39m TSNE(n_components\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m2\u001b[39m)\u001b[38;5;241m.\u001b[39mfit_transform(all_embeddings)\n\u001b[0;32m     65\u001b[0m plt\u001b[38;5;241m.\u001b[39mscatter(tsne_embeddings[:, \u001b[38;5;241m0\u001b[39m], tsne_embeddings[:, \u001b[38;5;241m1\u001b[39m], label\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mEntity\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[1;31mNameError\u001b[0m: name 'np' is not defined"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "from sentence_transformers import SentenceTransformer, InputExample\n",
    "from sklearn.manifold import TSNE\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm\n",
    "\n",
    "from tqdm import tqdm\n",
    "excel_file_path =r\"C:\\Users\\annem\\OneDrive\\Desktop\\ohdsi concepts\\CONCEPT.csv\"\n",
    "df = pd.read_csv(excel_file_path)\n",
    "pretrained_model = \"FremyCompany/BioLORD-2023\"\n",
    "model = SentenceTransformer(pretrained_model)\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "model.to(device)\n",
    "\n",
    "class CustomDataset:\n",
    "    def __init__(self, entities, concepts):\n",
    "        self.examples = [InputExample(texts=[str(entity), str(concept)]) for entity, concept in zip(entities, concepts)]\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.examples)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        example = self.examples[idx]\n",
    "        entity_text = str(example.texts[0])\n",
    "        concept_text = str(example.texts[1])\n",
    "        return entity_text, concept_text\n",
    "\n",
    "def custom_loss(entity_embeddings, concept_embeddings, margin=0.5):\n",
    "    entity_embeddings = entity_embeddings.requires_grad_()\n",
    "    concept_embeddings = concept_embeddings.requires_grad_()\n",
    "    \n",
    "    distance = torch.nn.functional.pairwise_distance(entity_embeddings, concept_embeddings)\n",
    "    loss = torch.nn.functional.relu(margin - distance).mean()\n",
    "    return loss\n",
    "\n",
    "custom_dataset = CustomDataset(entities=df['Entity'].tolist(), concepts=df['Concept(gold)'].tolist())\n",
    "batch_size = 3\n",
    "dataloader = DataLoader(custom_dataset, batch_size=batch_size, shuffle=True)\n",
    "num_epochs = 3\n",
    "learning_rate = 0.01\n",
    "optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)\n",
    "\n",
    "# Training the model and visualizing embeddings during training\n",
    "model.train()\n",
    "all_embeddings = []\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    for batch in tqdm(dataloader, desc=f\"Epoch {epoch + 1}/{num_epochs}\"):\n",
    "        entity_texts, concept_texts = batch\n",
    "        entity_embeddings = model.encode(entity_texts, convert_to_tensor=True, device=device)\n",
    "        concept_embeddings = model.encode(concept_texts, convert_to_tensor=True, device=device)\n",
    "\n",
    "        all_embeddings.append(entity_embeddings.detach().numpy())\n",
    "\n",
    "        loss = custom_loss(entity_embeddings, concept_embeddings)\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "8a7fb44e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjMAAAGdCAYAAADnrPLBAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB360lEQVR4nO3de3xTZbY//k+aJml6Sdu0SUtv3ClUAS2WcpWLDVQ4nsMMzkwZpyg6MDo6fhVnFBAqFATE84M5nHHGkSMgzByY62HGAZFCAVEKVCqCtpaLIBZaegklvaRNm+7fHzUhaXPZSfbO3knW+/XipU13dp6mafbK86xnLQnDMAwIIYQQQgJUmNADIIQQQgjxBQUzhBBCCAloFMwQQgghJKBRMEMIIYSQgEbBDCGEEEICGgUzhBBCCAloFMwQQgghJKBRMEMIIYSQgBYu9AD8oaenBzdv3kRMTAwkEonQwyGEEEIICwzDoKWlBSkpKQgLcz7/EhLBzM2bN5Geni70MAghhBDihW+//RZpaWlOvx8SwUxMTAyA3idDpVIJPBpCCCGEsGEwGJCenm69jjsTEsGMZWlJpVJRMEMIIYQEGHcpIpQATAghhJCARsEMIYQQQgIaBTOEEEIICWghkTNDCCGEeINhGHR3d8NsNgs9lKAklUoRHh7uc9kUCmYIIYQQB0wmE2pra9He3i70UIJaZGQkBgwYALlc7vU5KJghhBBC+ujp6cHVq1chlUqRkpICuVxORVc5xjAMTCYTGhoacPXqVQwfPtxlYTxXKJghhBBC+jCZTOjp6UF6ejoiIyOFHk7QUiqVkMlk+Oabb2AymRAREeHVeSgBmBBCCHHC25kCwh4XzzHNzBBCCE+MJjPWH6jEtaZ2DEqIxIo5WVDKpUIPi5CgQ8EMIYRwyGgyY837X+Av5TWw3f9y4hKw+9R16LK02LYwR7DxEeKt6dOn47777sOvf/1roYfSD82fEUIIRxbvKseoooPY2yeQsVVSWY8nd5z267hIaHniiScgkUj6/cvPz2d1/2PHjkEikaC5udnu9r///e9Yu3at9etBgwaJJrChmRlCCOHA4l3lKKmsZ3VsaXUjntxxBtsXjed5VCRU5efnY8eOHXa3KRQKn86pVqt9uj+faGaGECIYfasJszYfw31rDmHW5mPQt5qEHpJXjCYz60DGorS6AYt3lfM0IiIm5h4GZVea8I9zN1B2pQnmHob3x1QoFEhOTrb7Fx8fD6C3aeP//M//4Hvf+x4iIyMxfPhw/POf/wQAXLt2DTNmzAAAxMfHQyKR4IknngDQu8z0wgsvWP//m2++wYsvvmid+Wlra4NKpcJf//pXu7Hs27cPUVFRaGlp4e3npZkZQoggctaVoMEmeGk2diF7XQk00XKUr9QJODLPrT9Q6dX9SirrYTSZKSk4iB38ohZr3q9E7Z0O620DYiPw2iNZyL93gGDjWrNmDTZt2oQ333wT//3f/43HHnsM33zzDdLT0/G3v/0N8+fPR3V1NVQqFZRKZb/7//3vf8fYsWOxZMkSLF68GAAQFRWFgoIC7NixA48++qj1WMvXMTExvP08NDNDCPG7voGMrYZWE3LWlfh5RL651uR9hVhvAyEifge/qMUzf6iwC2QAoO5OB575QwUOflHL22P/61//QnR0tN2/9evXW7//xBNPYMGCBRg2bBjWr1+P1tZWnDlzBlKp1LqcpNVqkZycjNjY2H7nV6vVkEqliImJsc78AMBPf/pTfPjhh6it7f3Z6uvrceDAATz55JO8/awABTOEED/Tt5qcBjIWDa2mgFpyGpTgfVE1XwIhIl7mHgZr3q+EowUly21r3q/kbclpxowZOHfunN2/p59+2vr9MWPGWP8/KioKKpUK9fWeLZU6Mn78eNxzzz147733AAB/+MMfMHDgQDz44IM+n9sVCmYIIbwwmsxYte8CCt89jVX7LsBo6t3f8+jvPmZ1/+wAmp1ZMSfL6/v6EggR8TpzVd9vRsYWA6D2TgfOXNXz8vhRUVEYNmyY3T/bBF6ZTGZ3vEQiQU9PDyeP/dOf/hQ7d+4E0LvEtGjRIt5bQVDODCGEc3139lhqrKgjZdC3d7E+zw29Eanq/uv1fPO02N3zeyu8fixfAiEiXvUtzgMZb47zJ0vDR3edwuVyucNjfvKTn+Dll1/G1q1bUVlZiccff5yXcdqiYIYQwilXW5Q9CWQA4OGtx3F+NbvaGFxxFoilqBQ48ssZ/YIafavJ451MFrosLSX/BiltDLseQ2yP81RnZyfq6ursbgsPD0diYqLb+w4cOBASiQT/+te/MGfOHCiVSkRHR/c7btCgQfjoo49QUFAAhUJhPXd8fDy+//3v41e/+hVmzZqFtLQ0bn4oF2iZiRDCGW+2KLvS1un6kyHXXAViNw2dGFV00G479eJd5V4vh83MTKRKwEFs/GA1BsRGwNniigS9u5rGD+andsvBgwcxYMAAu39Tpkxhdd/U1FSsWbMGy5YtQ1JSEp577jmHxxUXF+PatWsYOnQoNBqN3feeeuopmEwm3hN/LSQMw/C/4V1gBoMBsbGxuHPnDlQqldDDISRo9F2OOVxZh1oDt4m7/ir/bzSZMaroIKtjdVlaAPA5cKPWBuLV0dGBq1evYvDgwV53crbsZgJglwhsCXB+95NsQbdn82n37t148cUXcfPmTeuylTOunmu2129aZiKEeMXRcgwfSirrsXhXOS8X/Rt6Ix7eehxtnWann6CdjYkLfP5sRHj59w7A736S3a/OTLII6szwpb29HbW1tdi4cSN+9rOfuQ1kuELBDCHEY56U7udCSWU99K0mqKM9f2N0lsw74tUDMJmFn5imwnnBLf/eAdBlJePMVT3qWzqgjeldWpKG8bu7RyibNm3C66+/jgcffBDLly/32+PSMhMhxCOeLMdwzdmyjLOAxVnQJQEc1v8QSuGEDKydN9rn81iehyv1LWhsNSEhRoFhmmi3u7FIf1wsMxF2aJmJEOJ3QlasdbQs42z3kSZa7rQ4n5gCGQDYf77W52DGYeBW34ayK3rsPnWd8nNIUKPdTISISCA0XhS6Yq1lWQZwvdzlrsqwmOjbu6w/kzfYLPtZAkFCghHNzBAiEs4aLwJAuAQ4+P+mYVhy/1oP/jYoIZK3ZF+21h+oxIo5WX7N2+Hb+gOVrGZn9K0m/PD3n+BaUzt6GEAdKUVjG7tAiPJzPBcCmRiC4+I5pmCGEIEZTWbcX3wIHd3OS4l3M0Der48D6M2vYFuZlg8r5mRh96nrfn3Mvq41tQddg0Y2M16OGnSyDWQs2AZNoc5S7r+9vd1h12jCnfb23td+3xYLnqBghhABebMryBJIWHJD+M6F6JtcW9Ns5O2x2BqUECn4chfX3PVoctVp3BPB9rzxRSqVIi4uztp8MTIykvf+QqGGYRi0t7ejvr4ecXFxkEq9/2BGwQwhAuFqezOftUr8VUvGUyvmZGH9gUrRjIcLrno0sek0zhY1tmQvOTkZADjpJk2ci4uLsz7X3qJghhABcF323zYp1pMGia74u5YMACTFyHCrxX3/Jn2rSRTLXVxx16Pph2+z6zTOBjW2ZE8ikWDAgAHQarXo6vKsrxhhRyaT+TQjY0HBDCEC4CPfQ7flGGpu360yalmGmpmZiO2Lcj06F9fBFltsAhngbgNKXZY2qJKAHVm8qxyXG7lZ2ktRKSj51wtSqZSTCy7hDxXNI8RPbHNPTl1pRJfzfF/OJUSG42zRbNbHr9p3QdSzHhIAw7VRqG8xoavbjDZ/Ppk8mZmZiNT4SLtZtef3VnAarBXkpGHj/LGcnY8QvlHRPEJERIglG1tN7d0Yt/YQzq6a5bRari2xJ4kyAC7Wtwk9DE6VVjda/98yq8a1G80d7g8iJABRMEMIz4QOZCya2rowbu0hNLXdXcpxtiNKDLVkCPco+ZcEK6oATAiPhMo9ccY2kLHVtzosJYkGJ/q9kmBFwQwhPAqkwm62O6KUcil0WVqBR0S45G7HFCGBjIIZQngk9tyTvmyDr20LcyigCRLUZJIEO8qZIYRHgZZ70jf42rYwB0aTGcXvf4k95d8KNCriqdRYBYZoYwRreUGIv1EwQwiPAq2wm6MEUaVcig3zx6CxrVNU+T/EuRt3OnH4pRkUxJCQQctMhPAo0HJPXCWIcrHsJKN3HL8pfv/LfrcZTWas2ncBhe+exqp9F6w5UoQEOiqaR4gfiGF7tiZa7rK/D9u8ir51arrNDOslqKrifAB3Wy5cbWhFDdU+4UWySoFTK/KsXzt7DVI+DREzKppHiIhYck/yNh/ze+EyRbgE54pmQymXcnJBU8qlWDtvtPVro8nMKpiZMVJjXfZYO280Fu8qp0DGT1wF044albIprEiImFAwQ4ifKOVSfLLsISz/2+fYU17D2+NEycIgC5dCGyPH3iWToI6WW7/3xvfH4lLtR7h2u/O7Y4HZ9wxAQ1sXVu274NVFy7KU5mrmSRMtx44nxlu/Flv9nWA0I1MDgN1zbdmW7yjg7VtYUd9qQsE7J1HfYnL4GiNECLTMRIifGU1mjCo6yMu5o2Rh+HTVLIefqnPWlbhcZrJwNEvD5pO6s0//04YnQBOjwPuf16LLzEAhk2CASoGvm2hWhk9VxflQyqWs+2wVTshAnaHDZeAjl0pgMve/ZGii5ShfqfNpvIQ4wvb6TcEMIQLgK4cmPAzodtBz0dlFyBnbgMaTpam+Qc+N2+12PYeIf9jOokx+4zCMXe5/95OGxOPk17e9fkwKaAgfKGeGkAAWHxkOTbQCidEKDNVGo+xKIy43uC/A5yiQAeBRIAP0LjsMWrbf7TF9cy1s82kW7yqnQEYAlkCG7UycRaMHxzrS0GqCvtVES05EEBTMEOJnbHIYbrd34+SyPOtSjr7VhOx1Jf4Ynkdscy1sUU6MfyVEyjBteAIOf1WP0ir3gajDc8QoAB87kRe8cxKHlk736RyEeIOqPhDiZ2z7Ndkep46WQyPST7xP/+HTfrcFUk+qQDRCG4WpwxNROCEDVcX5MHR04e+f18HQ2QMPJ+GsvmnwLZABgPoWE9WyIYKgmRlC/Ixtv6a+x5Wv1DldOpCFSdDVI0z62/GLjTj4RS3y7x1gvS3QelIFmn88N9U6GzZ42X5w8Zu/aej0+Rxd3fbJ7ZadUOpIGeaOGUBbvAlvKJghhKW+ya0v5mViy+Fqj2txsO3X5Ki1QPlKncOtsVsOVwvaNmHN+5XQZSVDGiYBEHg9qQJJWlyE9f8f23aSk0CGK21djpO29O1d2H3quvU1KpNKMFCtxJ9/NplybAgnaDcTISyw3X3Epvgc263Zlq21bPC53ZutPYsnYOLQBADizfEh4kO7oIgrbK/fvObMfPTRR3jkkUeQkpICiUSCffv22X2fYRgUFRVhwIABUCqVyMvLw6VL9h/n9Ho9HnvsMahUKsTFxeGpp55Ca2srn8MmxI4n26hLKusxcf1hLNn1KbZ99DVMDrYXsenXpMvSejQdL4YeUPUtd+vGbDlcLeBISCBpaDUhR4SBr7vcH32rCbM2H8N9aw5h1uZj0Pu4G4z4htdgpq2tDWPHjsVbb73l8PubNm3C1q1b8fbbb+P06dOIiorC7Nmz0dFx903xsccew5dffomSkhL861//wkcffYQlS5bwOWwSohy9eXmzK6fW0IlDlbfw+oEqZK76ABscJMO6atroba8cLhpB+uJ/Tnxtfd4oZ4Z4wrKtWywW7yrHqKKD2H3qOk5casTuU9cxquggFu8qBwDkrCtB9roSXKxvQ7OxCxfr25C9rkSUQVmo8Nsyk0Qiwf/93/9h3rx5AHpnZVJSUvDSSy/hl7/8JQDgzp07SEpKws6dO1FQUICqqipkZWWhvLwcDzzwAADg4MGDmDNnDmpqapCSksLqsWmZibjjbPYlShbmNA/AEz97cDCWO+hIzUcPnEU7z+DoVw0+ncNXafERqLlNFX4Je8MSI3H4lzP89ni2f3tpcUoADGqaO/B1Q6vL/mnuClDSshm3RF807+rVq6irq0Ne3t2urrGxscjNzUVZWRkKCgpQVlaGuLg4ayADAHl5eQgLC8Pp06fxve99z+G5Ozs70dl5NzPfYDDw94OQgLdw+yl8dLHJ4fe4CGQAYNuJq3hp1kjIw+0nQ/s2bfSV0WQWPJABQIGMyD2QoUKdwSSqRp/f3Dbycl5HHxie31vhdR0kdwUoqXigMAQLZurq6gAASUlJdrcnJSVZv1dXVwet1n7aPDw8HGq12nqMIxs2bMCaNWs4HjEJRuPWfoimtm7eH6eHAXaXXcNTU4c4PYaLBn5U3yVwpcVFoLXTjGZjF++P1Wzs5jSQsSxv+lIoMfy7nXBcctY0k29UPND/gnJr9vLly7F06VLr1waDAenp6QKOiIhRzroSvwQyFt/oneeR9K0f02zsQva6EiREyTBn9ADWy1D+ylUJA8BmzipFpeCkfkkoqGnuwIKcdOwp/5b3x/pGz90syIAYuds+XmzMHZ3M2Zh8HYuv6lvEk/8TKgSrAJyc3PvCvXXrlt3tt27dsn4vOTkZ9fX2L8bu7m7o9XrrMY4oFAqoVCq7f4TY0reaPOpbw4WB6v51Y4D+gYytpraufkmIU9444rCqqtFkxq07/EzVW6gipKhYqcPk4YmsjqdAxkN+qpTB5SxIbYsJ44o/BNCbhF5VnI/CCRmQefgYa+eN4WxMQrfT0MbQEpO/CRbMDB48GMnJyThy5Ij1NoPBgNOnT2PixIkAgIkTJ6K5uRlnz561HlNaWoqenh7k5ub6fcwkeBS8c9KvjxcmAR4dl95vt5Q3QVXN7Q67nRXA3d0XF33srePO6RU6qKPlDgv6Ed99WHXL/UEc+DeOZ0Ga2rsx4tUDAO7mgV1aPwdJ0TJW9x8QI+O0MrDQy617l0wS9PFDEa/LTK2trbh8+bL166tXr+LcuXNQq9XIyMjACy+8gHXr1mH48OEYPHgwVq1ahZSUFOuOp1GjRiE/Px+LFy/G22+/ja6uLjz33HMoKChgvZOJkL6MJjOuNvJ70e8rXa3E2OJD1q8ta/dRcu8/T1i6Vlv+n2+2tW9WzMkStOJwsNK38Z8vo4mWo3jeGPyl4ian5zWZGTz+7im899QE623HXn6IVTHH0l89xOlYhCwNIJdKKPlXALwGM59++ilmzLi71c6Sx/L4449j586dePnll9HW1oYlS5agubkZU6ZMwcGDBxERcbdc9x//+Ec899xzeOihhxAWFob58+dj69atfA6bBABvtzQLtY7+TZPj5Z82k2+7pfz1s/StfWMp0kedsQOLBEBzuwnZaw8hWSVHnYHbpdbjl5rsuqizeZ14WiCSDSHbaZjMjMNO8oRf1M6ABBxnAYm7YnNCJgQGIoU0DPFRMkwZloBwaRhuNHf0Cxy5eE4X5KTjlqEDpdXCbykXUkKkDE3tns/MOKt7opJLkBwXifoWEwzGLlYJ21wonJDRr9yAs7wwvmqyCN3eY4Q2yq4ZKPGeKNoZEMK1he+ecnrxtF126UvfaqJAxgm5VGJN2pw6PBFp8b0zo53mHtQZOvHXipvYW15jn4S8sTcJ2ZLwueCBNK8fv6bZiO2LxmPGSA1XP1LA0WVpMWfMAPcHovdCmTsoDqoIKVQRUgxKiESsov9bucHE4HZ7F2RSid8CGaD/Es/iXeVO88IaWk1O/2Z9oZRLES7g1e1ifRtGFR3E5I2Ok/UJ9yiYIQFj3NoP8dElx8XtLEoq6/u9eSzeVU5ND12wTIuvnTca4VIJq4J3Nc13k5CVcinCfbhyWJKJf/vjcV6fI5ANS1Ri28IcrHBQIdqRxpYOnL7WDEOHGYYOMy7Wt+FOp+NwpUGAXXu2yeFsdhU5+pv11YYDlXDQFs3vbjT3T9Yn/KBghgQET2rCzPjPo9bdQrS0xE7BOyfx5I7THlcPtsyG+ZJwabmIC70DRSiN372u2TYL1RvF/UnfNihj+zvl8ndv6u7B7z+6ytn5uOBq1phwg4IZInqebl+uM3Ral0MokGHn68Z2lFY3enXfksr673rbeCfru9yGUG1OaVuTROhmob7qm8x74Hwtq/sdOH+Ds9mZ/zlxhZPzcI2PGShyFwUzRPT8XRMmEPlaAq27x7d9AH+v8L5qLQPg//2xHJdutfg0hkA1Ji3WrvaQJQ8pJVYh9NA8MjMz0S4B32gys05obmo3c7Ycs+eMeEsGhOrsoz8EZTsDElyoNLh7Qm9J7DQDSdEy3Gr1rk7KPy4E5wyaIlyCzm7Xv52/flfvxVJ7SJelRXcPg5t3Aqt6cml1IxbvKrcGNN5cuC3LMa52Jboj5tmPUJ199AcKZojo9K0hkxgl80vzPeKbOx3ivYgIpXTpDMz77cceLZMG8tJoSWU9Fu08g+QYBf70aY3X5/ClTotcKt4FB6qczR8KZohTl+ta8fDW4+jqAWRhwAfPT8Ow5GgA3hetc8dRl1sSGDrEsH1EZBbtPI3ylTq7juiJUeG43MhvDy0heZpE7kjefx7FJyvyvLrvoMRI0fYEY7tjjXiOghnSj77V1G8rc1cPkPfr4wAAdaQMepu1cNvpcV+mh2nnEQk2liVSdbQch5ZOBwCs2ncBlxvFm9chBjcMnV4vNw3VxuDk17d5GJVvpBJQET0eiXc+jghi3NoP3dZk0TtJ6vNl+6HQXW4JP2Lk3HVnDkSOuidT3gQ73u7+8XT2w5edeJ4wM6Dt2TyiYIZYjXj1AOtaLs54+wZEWf7BaV52us87rQKZpXuy0WS2dky/dSd4l5i4tub9Lzy+D9t6PRYlS6f5bTs81ZvhDwUzBACwcPsph/1dvGFbtI4t+rQanK41tePqxrlCD0MQmmg51NFyLN5VjlFFB7H71HWcuNSIi/X+7dgeyPaW12DShsMef0BiW6/HUhfHsh3e0tJjQU66t0N2i+rN8IOCGQKjyYyPLrpuE+AJ26J1bD+FUJZ/cLL8XgsnZAg8Ev+KCA/DyAEqTNlYSsunPrp5p9OrGjSWACU5Rubw+446wa+dNxq7n8rFhvljeJ2toZlo7lHXbIJV+y5g9yn+EhLZJAYL3eWW8KNipQ7qaDn9fgknvN1k4O3uS742JUwcqsaexRM5P28woq7ZhLX9LEuOe4vttGpaXASv4yD+l72uBIt3leNpyhNwSRrKiUUeKKmsx+W6VoxZfRBDl+/HmNUHcUPvPgdJKZdixZwsDEqIxLWmdqw/UMnqPWnbwhxelpwa7rhv5ko8QzMzIc5f26ELJ2Rg7bzRgo6BEBKcJAB+MiEDL+ZlYsvharsZGKPJjKlvHEFbV/86SELNGidFy3F6pY7TcwYrttdvqjMTwvy5HdpZgi8FMoQQXzHorXVlu1xuqX/lCpv2CZbdUZy+T9FMHOdomSmE+TMJzVGCL9WWIYQIjc0yONfdzGMUNI/ANQpmQpg/t0M7KmRFGf2EEDHQbT7m9HuWGkEdXT1Y8EAaFuSkY+JQtU+Pp1VRfiDXKDwMYYMSIv3S+8hSy6Evqi1DCBGDmuYOh8tNzpbBdVlan5aehmqjvbofcY5mZkLYi3mZHh2vifI89u2bYEeVUImYTRueIPQQQpIuS4trG+cKuqvLdrnJaDJjysYjToMVy+3eLj1Rw0nu0cxMiDKazJi08bBH94mNlKOBRbsDpSwMj45L61fLgZJ9iZhZAu8Rrx7grBo2ce3R7BSsnTfG+j7x49wMXmteuZO3+RiyUlSs3qdKKutRVZwPoHfJ/A+nroPNq8bZTDXxDc3MhCBLefWObvZv2GlxSjS2Om4w2ZcivLeSJgUyJFDkDorDtoU5WLyrnAIZPzpfc8fufULoGYsbzR0evU+tP1BprRx8deNcTBvmemYvLT4CWwuyfR0mcYCCmRDjbVDxz+emOOwA7Ejf42jXEhG709eaUbj9FL1OfVSxUocYBfvLSn2Lye5rpVwKdZTj9gNi1Dfv772fTkBVcT4eHZeCCAdrZjW3OzCq6CAef/eUv4YYMiiYCSG+BBUPrCuxdgB2p+9xtGuJBIITHPYnC0W6LC3U0XKPZnwdfUCaO3oAl8PilbOeckpZOFw9DccvNWHEqwd4GlVoomAmhBT/8wuv79sDQLflKDTRrmdnLJ2CbfHdLoEQIizbRP/wMPZZvI4+IHm6MUFIfZfFbDukd/e4DupMZgbj1h7ic3ghhYKZEHL0YoNP929q68a+n09xGtBoouUo71Oie/Gucujb2eXaEEICQ7JKganDE1E4IQNVxfl2Oxb/bXQyq3PIw9Dvgw8AbDlczdk4+aSOlPmcF9jU1gV9q8n9gcQt2s1EPDJ5Uyl0WVq88f2xKHjnJOpbTNDGyLF3yaR+b0yUK0NIcNJlJTnttVY8bwz+UnHT7Tk+X927E6hvR+vLDa2cjpUvc8fcXQ7z5b2u4J2TOLR0OkejCl0UzISQGZla7Cn/1ufz9P7Rfu72D5ByZQgJTq52HbHpZWTZntx3NsMfRTy5Yvsc+PJe1zcJmniHgpkQUvTIPZwEM8DdAlOu6iVQhV9CAkdqXARuNHe4PU4qQb+/fX2rqd9MLfC50+q5lm3wYpu5jZKFOeyu7eg425/fl/c6trtEiWsUzIQQrru/rj9QibXzRjt8I1NHy/3WLoGQeKUMt42hnZv1XwX34dNreuy/UAt9m/PnQpelxdaCbLulHcssw6iig24fx8wA2etKrDlyOetK0GCT99Fs7LJ+v6o4v9/jKOVSUSxBz8zUIDVeaTc2o8mM7HUlbu974pWH7L725b2O7S5R4pqEYZigrxBlMBgQGxuLO3fuQKVSCT0cwXH1iWjq8ER8VWuweyOzkIVJ8Oi4NM5mgghxZYQ2Cv94birWH6jEgfO1aAqxpPP/KrgP/3FfqvVrVz2F+vYfssX1bElClAxnV83qd/vgZftZVcvlWkKkDHPGDOhXndxW3+CsL0cbHYwmM6tAkM25iD22128KZkKUJenuq9o7KP/mjlfnUIRL0OlBTQlC+DJpiBr/u2Si1xeVQPeD7BS8+cP77W7rm1jr6gJui+uA5sHhCdj11ATr14XbT7Gu6TNpSDyGamNwuaEVZVf0Po1j2nA13ntqIqtjnQU0roIPT583CmTYoWDGRqgGMzf0RszeUgpLF4JB8Qr8/dkHHW6HFOP6NSFsFU7IwNp5o5G58gN0drvPefCGUhaGWKUMU4Yn4OTlJty808nL43jL3ayLK30DnxfzMjH9P0th6DBzMraESBlm3ZOEbnMPq51OFgU5adg4fywA39+jRmijPNo11Hf5fMcTuXj7o8sug0NnY4xVSMF8V34nWRXhcPcncYyCGRuhGMy4apaXEBWOs6tm97vdaDKj+P0vcbS6Hvo2k8seNWwT5Qjxh6rifNb5Dt6aOjwREbIwny6ouiwtzl1vdrmM4Qu2ZRMsjCYzpr95GLda+jeQlYYBZoH/xH+YnYJNNjNOvgQ0cUoZzr3Wf8mLDU+W7bydESOOUTBjI9SCGTZdf+VSCS6+PsflMa7+gMuv3kZziCdcEnGYNiwBGYlRrLsWeystLgI1LHb7OGO7rOAuL4Nrzgpain02NlmlwKkVeXa3vfznz/BnD2Z3LDydmbFw9zz5MiNG3GN7/abdTEHmht7Iquuvycxg4fZT2PXkBKfHbFuYY/2UcaW+FY2tnUiIViBZFYHEKBkFM0QUjl9uAi7z31fJl0AGABpaTTCaepdt8u9NRnXdHVTeMKC1i//Pkw2tJuSsK7EGNIEQyDjTbfZu6ct215Cr2RPb76XGRbh9nixlKoDeHZ6X61vR1NqJxGgFhmqjaWbGTyiYCTIPbz3O+tiPLja5rRWjlEtRZ+jAya+/S76rb0PZ174l4hESqvI2H2NVy4UPDa0m6FtNUMqlARPITM/U9LvtXxdueXwe255xjgr17T51HWnxEWjvNHvVfkW3+Thqmo12t12sb8PJr/XYfeo6zd74AfVmCjJtnZ59anFXuTKQP8ERIjZCBTIWU984At2WY4KOwROvPXJvv9tMHubx2C6xuXo/q7nd4XUfub6BTF8llfWYsrHUOoNDuEfBTJCJUng2nemqcqUYClsRQrjT1tWDmtvCBlRsWVoe9MW+JzfwaHaKNZAR+v2sptmIUUUHsXhXuWBjCGYUzASZD56f5tHxgxIinX6PeisRQoTyxvfHOrx9cEIE63OsnTfG+v9ieT8rqayngIYHFMwEmVS1EnIp+88ujhrGGU1mrNp3Ae9/Xsvl0AghhLWCd046vD1dHcXq/vFKKWf9k7hmmzRMuEHBTBC6+PocVgGNo2ncxbvKMaroIHafuk67lUjIiKCtEKLjqJv04l3lOH6J3c6120azXcCQGsd+RscfxDJTFCwomAlSF1+fg09enun0F+wou56SfUmo6qYPyaLTt5u0NzkvtgGDxKNsG/6JaaYoGNDnkSCWqlbi641zWVWkFDo5jhAhedpiTC6VsKrnRLw3Ni0Whe+etr5neTOTYRswuNtx5G+u8hWJ5yiYCQFKuRRr5412eUwgbdckRGgUyPDP0sPJUgdGHSnz+By2AcOghEicuMTZ8HzmKF+ReI+CGQKjyRww2zUJ4Ut4GMBTj0rCAW9qwPzjsxr87+nriFJI8fenp2D3qes8jMxzzradE+9RMEMoEY0QUCATjAydvb9UQ4cZeb9mXx2dTzMzNVQNmAeUAExwpb5V6CEQQgghXqNgJsQt3lV+t+8SIYR4KTwMGJbIf1KrN7kzYlJa3YCcdSVCDyPoUDATwmgrNiGEC9OGJ+Dy+rk4/MsZ0GVpeX2suWMGoKo4Hwty0pGsUnhUJFQsGlpNeHLHaY/vZyloWvjuaazad4EK79mgnJkQRVuxCSHeUCvDoTd22912/FITctaV4KOXZyJZFQF1lAz6Ns8SdtWR4dC3d7s9bsWcLDy/tyLg379KqxthNJlZJQIbTWbk/X9HceNOp/U2yy6vyUPV+OPiiXwONSAIPjOzevVqSCQSu38jR460fr+jowPPPvssEhISEB0djfnz5+PWLc9bwBN7tBWbEOIpuVTSL5CxaGg1WauHexrIaKLlqCia7XZWR5elDYpAxsKy+cJoMmP5385jwvrDmLD+CJb97XPrrIulKrttIGPrkyt6DFq2329jFisJwzCCFkxYvXo1/vrXv+Lw4cPW28LDw5GYmAgAeOaZZ7B//37s3LkTsbGxeO655xAWFoZPPvmE9WMYDAbExsbizp07UKlUnP8MgcZoMmNU0UGhh0EICSDJMTLUtXDX4iQMwDBtFPYumQR19N1qv86Wv3VZWmwtyA6q966pwxMRIQtzGpxpouVoaO3f1sGZaxvncjU00WB7/RbFMlN4eDiSk5P73X7nzh28++67+N///V/MnDkTALBjxw6MGjUKp06dwoQJE/w9VL/St5pQ8M5J1Bl6a8CMHBCDzCSVwwq+nqCt2ISEFkWYBJ09vn1ujY6QARwGMz0ABiZG2QUyALBtYY7TquWr9l3g7PHF4GpDK2qandf48iSQAYDHtn2CPy6e7OuwApIogplLly4hJSUFERERmDhxIjZs2ICMjAycPXsWXV1dyMvLsx47cuRIZGRkoKyszGkw09nZic7Ou1NyBoOB95+BSzf0RkzeVNrv9jNXm3HmajN2n7rusLcSW9QThJDQ4msgo8vSovzqbY5Gc1dJZT0W7yrv917mrGq5v967pg1PYN3Q0heuAhlvfHKlmXUeTrARPGcmNzcXO3fuxMGDB/G73/0OV69exdSpU9HS0oK6ujrI5XLExcXZ3ScpKQl1dXVOz7lhwwbExsZa/6Wnp/P8U3BnxKsHHAYyfVneBIDeGZxZm4/hvjWHMGvzMejdRPPUE4QQwpblg1Pfxo9cKamsd/ueZcH3e5ciXIKq4ny899QE3ndlRcn4ufw6mnkPhV1QgufM9NXc3IyBAwdi8+bNUCqVWLRokd0sCwCMHz8eM2bMwBtvvOHwHI5mZtLT00WfMzPi1QMe93xRK6XQG/u/MBXhEpwrmu2woWTx+19iT/m3Po2VEBIaLMGMvtWEbB7ro7CZbeY63y8MgFQqwcD4CPz56Sn9lrz4LF8Rp5Sh2cjdsp3F1OGJ2P1UrvVrZz/DtOFqZCREu2xALAYBlTNjKy4uDiNGjMDly5eh0+lgMpnQ3NxsNztz69Ythzk2FgqFAgqFwg+j5c4NvdGr5nWOAhkA6OxmMKrooN0bBNWVIYR4ynYpSAKAr0+/zpacbCnlUuiytJy9j31ZnO/yAm6bv7P/fK1X/aGc6WH46Z+RFheBVfsu4FpTOy7fMqDW4HjW6/glPXCpt2CqZZu3L+kLQhN8mamv1tZWXLlyBQMGDMC4ceMgk8lw5MgR6/erq6tx/fp1TJwYXPvqH97KT98QyxsEBTKEEG9ZloKSVPx+SCyprHe5BKJvNeFqQyu4KJPHttmjJX+nomgWqorzUTghA1OHJyItPsKnxzd08LPUs6e8BrtPXceJS41OAxlnbNMXAo3gy0y//OUv8cgjj2DgwIG4efMmXnvtNZw7dw6VlZXQaDR45plncODAAezcuRMqlQq/+MUvAAAnT55k/RiBsDV76PL98GJihgSAhMhwNLEoBkaI2KXEKnDTSb0TrhROyHCY/Dtu7SE0eVi/xhmuZiAsy/alVbdg6OyGsSs4upVWuZmx8qeAWWaqqanBggUL0NTUBI1GgylTpuDUqVPQaDQAgC1btiAsLAzz589HZ2cnZs+ejd/+9rcCj5p7UQopb5E6EZhEwunUOCFC4TuQARzvWPImn9AdZ9u/PRFMBfxsrT9Q6TCgFDPBZ2b8IRBmZpxtxybBYdqwBMjl0qB84yOES31nZh5/95RftkkD7GdsjCYzdFuOoea2663VUknvB9VRyTG43d6Fi/VtXA3VTmqswmmFYG/0TSIWEtvrt+hyZkJVqlqJ8MDrl0ZYOn65CcerG4QeBiGciODxzWrFnCzr/xtNZr8FMgC7nBFLewF3gQwAmBng9Aod/vT0ZCTFKrkaph1dlhZDtDGcnjMQy3dQMCMilzcEXylqchfX0+SECCVncAJmZmo4P2/fpFwhesiVVNbjht7o8HvebKQofv9LAPwFCFsLspEWx22gZBtQBgoKZkTm2sa5nP9S1JEyjs8YWn5wf4rQQyBEVAYlRGL7ovF2u3sKJ2TgDwvHe33OacMTrFuhV+27gHHFh1jNfvBh8qZS5PSpqWM0mb1aJj5a3XsfvgKE9QcqAQ6zRdju8hIbwROASX9fb5yLG3ojHt563GlSsARAHoukUl2WFsmqCOw+dZ2HkYaG8PDA+8MmhE+WC3PftgO+9E46fqkJOetKPO5HxJeGVhNy1pWgfKUOgG897YwmM3Sb+Sm/wWWLh0CuM0PBjEilqpU4vzofAHC5rhUPbz2Orh5AFgZ88Pw0DEuOBmCfkZ8WGwFIJKhpNtpl548r/lDIHyWg6bK0qGl2POVMSChy9cnd1wurWAIZi4ZWE27ojUhVK73+2fRtnbx2+rYsX5245P05wgBMGKJGsioiYHs70W6mICemTzqBxvIpZdW+CzSzRQiAGSM12PGE86WkYP1bEfMMd1Vx74deLgMmXZYWWwuyfd66zgXazUSgbzVRIOOlVJUCWwuyAQRmMhwhXJuZmegykAGC92+lpLJelDO0mmg5lHKptc0DV0oq6zGq6KC1kvDuU9cxquigqKsDUzATxAreYV8lmdi7YeidGn5yxxko5VLEKWlFloS2CzcMbo/h+qIqJke/auBlB5cvGlpN1vYP2xbm8P7ci7ndAS0zBbH71hzipStrqNFEy5F/b7Iop5iJa7ERUtwJosraQxKUyB4Uj79X3ESPAO/civAwjB+sdrvsEKy94B7NTsGpr2+LapYmWaWALivJ+vvoW9m429yDPeU1nD6mP9sdsL1+UzATZGxfyCcvN1K/J45MG57g1+JdhBtpcRGoaRZmey9fUuMi0GI0wdApfB8gZ7tfensWfYG/Vdyg+kp+5Oj3YTSZOU9AdtY/iw+UMxOCLJUpLeuc9B7CneOXmvDg8AShh0E8FGyBDADcaO4QRSADOF52sLwP7SmvoUDGzxz9PvhY+rtS38rp+bhAwUyQCNZpXTEZmBAFTbRc6GEQlhRelNxPiKTcKE+VVNZb8zbofUh4tr8Pi20Lczhtl3Pya73ocmcomAkC3lamJJ651tSO8pU60SUBEsc6uz2fFWhq78bMTA2k1CfNI+sPVNL7kIj0LfD3+Ltl8OLPwaWSynqMKz6EVfsu9AuehEDBTBDwpTIlYc9SnGrrgmw8lJmIWGU41FEy/CgnDVXF+f1Kuw9J4KexXCCThQE5A2OFHoZLpdUNgiTXBrLL9a30PiQi+8/XWv//yR1ncPySnpfHaWrvEs22bZpTDQJclrMmzv0wOwODlu3vd/ufymvwp/IayKQSDFQr8eefTYY6Wo7hr/Y/NtR19QDl39wRehhuUSzjmabWToRL6bOxWOjbu2A0mfH83gqUVjf45TEt+TpCtUOgYCYIDEqI9KmUNWHnkd9+7PL7XWYGlxvakb2uBIpwCbqFn3klxC8SoxX0PiQya97/wu/LfpZ8HSEqBVMoHQSCtepmIOvsZujTPQkZja2deHzCYKGHQWz849xNQR5XqOVGCmb8yNLavvDd05wmTQVz1U0hpcVH4P2fTxF6GD4Lo2RWgt6+SpacrgU56Zye+2J9G/J+zU9XaOIdY5cw2/eFSnugonl+YDSZkbf5GG44qHnBZcv14SsOoIsyFzkVLgHnuwAIEcqCnHQUPXIPlHIpstcegr6NKoR7K1IWhnaBAgYxK8hJw8b5Yzk7H1UAtiFkMPPkjtMorW50eYxUAowbFIfMJJXXnUn5qPJICBGHipU6vHmwCns+5aYsvZi7QJPAx+WHdKoALAI560rcBjIAYGaAM1ebfdriRtsiiVhQjRbubTlcjfBw7t6uSyrrcaTqFmfnI8SWEA0pKZjhyaKdZ9DQavLqvt68EGh7Nj/+fTTlInkqWuH/nQzB7lpTO+d/4zfvdHJ6PkJsOapEzCcKZnhgNJlx9Cvf9vZ7+kKwFHQj3Gps6xZ6CAEnSRVBCekcGxAjp79xEnD8uWJAwQwPit//kpPzePJCoO3Z/Dj5NT+VM4PZjidysW1hDgU0HPrX+Vr6GycBx58rBhTM8OBoNTeFivZfqHV/0Hdoe7Y4yUJwX/TkTaUYtGw/SqvqoVJI8dCwOKGHFPDauxk8+d4ZpMVHCD0UQljz52wiBTMipm/r8mipiT4Ni8vMTA0KxnNbzyOQmBnA0GnGkcvNQg8lKJRd0aPmdv/yDoSIlT9nEymY4cEMDrsqe7rmuG1hjl3Dw7Q4+iQnlM9rmtFlpjoUhAiJ0tGFkRoX4de2BhTM8KDokXs5O5c3a45KuRRr543GO4UPoMZBob5QIRf41d3U1oW95dzUBSGEeIdapAnj8NLpfn08CmZ4oJRLOZsR8WXNMZRrz8zM1MBEkyKEEOJ3uiyt35tNUjDDk8GaaE7O48uaY6jWninISUNqvFLoYRBCSMjhsvqvJ8L9/oghYlBCJE5c8u0cvka3XIwhEDnqgUUIIYQ/C3LSUPTIvX6fkbGgmRme+JrFnRqrQEKU3NphW99q8rjjdqjWpRiUEEkFxgghxE90WVpsmD9WsEAGoEaTvFq8qxwlldzUnHGEzXQe32MQo6rifACgxpuEkKAWJgF6BL6C872sRI0m/cxoMvebOeG77gubHk7BVnsmRaVw+X3L0pxSLsVMDrfIE0KI2AgZyIzQRqGqOF+Q/BhHKGeGA31nP05cAnafum6NWI0mM3Sbj/GyTdrSw8nV9J5lDOsPVOJaUzsGJURixZws5G0+FnD5JbdaOhElC0NbV/+tSrosLbYWZGPVvgvYf74W+vYuAUZIHEmIDEdTO/W5IiRYJMUqBV1W6ouCGR+5WsaxzJxsLcjmtd7L+gOVWDtvNABA32pCwTsnUWfofbyRA2KQmaTCijlZ1mMsDi+dHnBLMWYG1kAmIjwMOYPV1uDs+b0VAffzhIJpwxJw/HKT0MMghHBIbHmJFMz4wGgyu81HKamsR9G+87yOw7IFO2ddCRpaTXbfO3O1GWeuNtvNFFko5VKoI2UBO4PR0d2DL769jVt3jBi9+kN0C714TBw6dY2adRISbJ5+cJjQQ7BDOTM+YFuU7i8VN3kdx6CESIeBTF+OcmzmjhnA59BY02VpMTMz0eP73e4w42J9GwUyItbZTb8bQoLN49vLhB6CHQpmfHClvkXoIQAAHp8w2G0gY2HJsbHgYvt2pMzzztBpcRGYOjwRhRMyULFSh6qbLSitbvR5LIQQQvj3jcianlIw44NGlgEE377/9sceHW87o6SUS33e7TRn9ADEKT1bsaxp7kCELAx1hg5krytBTbPRpzEQQkiwGKGNwsShaqGH4VJ4mOcfYvlEwYwPEmJcbxP2l7ZOz1qp9W1z4Gr7tlzq/gX714qbaDZ6vlOlpLI+5GrgEEKIO+MGqnG1oU3oYbj0b6OThR6CHUoA9sEwTTTKrgif3BilkMLQwT6gcZSF7mz7tlIuZZWPQwghhBt7yr8VeghuFc8bI/QQ7FAFYB8YTWZRbAX+5OWZmLyplPXxFSt12HK4ul/Q4oq+1YTsdSW+DpUQQkiA82czSbbXb5qZ8YEl30TIpRJdlhapaiU00XJWsyeaaLldUNK3wJ8zbHpBEUIICW4JkTLRVP21RTkzPtq2MAeaaLkgj20bgJSv1LkdR0KkzGnA4641wsNbj3s/UEIIIX4VztPVfY5Iynn0RTMzPjKazH7PJ/n30Vq88YPsfktD5St1TisAX29sc1uF1bJt22gyo+Cdk6hvMUEbI8feJZM8TjIOZLIwCZJUEbTDihASkORSCUxmfjJIuCjnwQfKmfGSJVm2pPIW6gydnJyTjarifI/7YXjSOVsRHobO7v59j4JZpCwM8nCpNXBTfzfDlfefR3G5sd3NvQkhRDhJMXLMzNSi5k4HBiVE4vEJg5H3a35m0v2ZK2PB9vpNwYwXPAkOuFY4IaNfjyVXxJKkLHYFOWm43tSOxtZOJMQokKSSY99ndUIPixBCWFFHyRARHoabd/j5cC1EIANQAjBvhAxkgP41Ytxh23Ih1O0tr7n7Rb246zsQQkhf+jZ+euylxUWgZOl0UXXIdoSCGQ+waSzJN087lXoa/BBCCCGAcLMx3qDdTB4QwyyHJfnKaDJj1b4LKHz3NFbtu+B06zTXbdrjlDJ88vJMqCKkYFEcmBBCSACaOFQdMIEMQDMzHhF6lmNmZiKUcmm/pS7bWjFvfH+s3U6kHU/kYvep627PrQyXwMiiu7E2Ro5UtRLnV+cDYLfsFiUPQ2dXD6h5MiGEBIZhmmihh+CRgJmZeeuttzBo0CBEREQgNzcXZ86c8fsYuJ7l8IQmWo7ti3JdBg8llfXIXleCi/VtaDZ24WJ9GyZvKnXbX2lmpgafLMtjNY69SyZZ/59t/lCbiQIZQggJJGLdgu1MQAQzf/rTn7B06VK89tprqKiowNixYzF79mzU1/s3f0WoX+7MTA3KV+q8ztkxmRmnAY0uS4vti8ZDHS13W3RPEy23blsWQ/4QIYQQ7umytKJP+O0rILZm5+bmIicnB7/5zW8AAD09PUhPT8cvfvELLFu2zO39udya7a/dTLIwCQrGp9v1TVq17wKrJSNnvj92ABRyKWqaO5z2ZHLWVFITLUf5Sp31a1/HQgghRHzElvQbNFuzTSYTzp49i+XLl1tvCwsLQ15eHsrKyhzep7OzE52dd/faGwwGzsazbWGO04BGl6UFAK+CnShZGGQOCrfZOnC+1vMB2/j757XWcXpSqwYAepgerNp3wdqc8nJDq09jIYQQIh4jtFH4x3NTrR9wLYVhPWlILCTRBzONjY0wm81ISkqyuz0pKQlfffWVw/ts2LABa9as4W1M2xbmuPxFs5m9UUgliI+SY0amBkWP3Ov2RWI0mdHUzk0dAUsfpr7Rt7NZGQBoauu2zsScuMTJMAghhIiEbSDjbJNJalwEhmiiRRnciD6Y8cby5cuxdOlS69cGgwHp6emcPoZSLnU6u9E32EmLUwIMYy037c2L4P7iQ1wM28rSh8kyjht6o997TBFCCBGebY6Mqw/jN5o7cKO5w24HrViWpEQfzCQmJkIqleLWrVt2t9+6dQvJyckO76NQKKBQKPwxPKdcBTue0rea0MFDv6T1Byqxdt5owasaOyMBIPqELkIICWC2AYmnGzuczfILQfS7meRyOcaNG4cjR45Yb+vp6cGRI0cwceJEAUfmPwXvnOTlvNea2kUbyADAcG0UqorzUTghA1OHJ6JwQgYW5HA7w0YIIaFoxHfvr7aBiDeFYS2z/EIT/cwMACxduhSPP/44HnjgAYwfPx6//vWv0dbWhkWLFgk9NL+ob+Fn+SctTok95d/6fJ5wCXipI7N3yaR+M1xGk5mTMRNCSChLiFH0S/Z9/3PvNpnkbT6GT5Y9xOXwPBYQwcyPfvQjNDQ0oKioCHV1dbjvvvtw8ODBfknBwchoMqOH4X6JCQAYjhZxuhmgYqUOc7Z+hDoDdx1bZ/7nEZxb/TAA9MlBikBNcwdnj0MIIaGm7Ioei3eVA/BuB66tG80dgi83BUSdGV9xWWfGn/hcApqZqUFXD4MTlxo5OV/hhAwA4Lz2THgYkBwbgZrbFLwQQoit1LgIZCbFoLS6QeihAACqivM53+HE9vot+pyZUMVnINPbGmE8p+0ZrjW181IhubsHFMj4AfUMJSSw6LK0+GTZQ9i+aHy/3MIZIzWCjEnIZswBscwUavhsFZCiUuDkit4+TCvmZHE2kzIoIRJKuRS6LK1oE4oJISQY/CA7BW/+8H7r1452zwqxuUPIZsw0MyNCfEa3R345w/r/SrkUafERnJzXMiuzbWGOtRKyv8Uo6OXsraBfayYkiJy43OT2mG0LczBpSLwfRnOXkM2Y6d1fhPiKbh01Dxuc6Hub977n3bYwxzrtmTsoDqoIKdw07uZEbKTrRplipI2SYUFOOgpy0oQeCiEkyAzVxvj18YTstE3BjAjxFd2e/roJN/RGTh8rLT7CYQa7ZdrzT09PxvnV+biyYS7vMzaBmFszLVODDfPH4AbtziKEsBQtl7Kq7eLP4GJmZqKg7Q0omBGY0WTGqn0XUPjuaazadwFGk5m3F6Chw4zJm0ox4tUD1tt8faySF6ezPnbbwhzMzEz06fGCzV8rbmLVvgtIjeNmuY8QEvwuN7ZjVNFB69ZqZ5RyKeIj/ZMamxov3BITQFuzBeUsQWvK0HgoFTJek7fkUgkuvj7H5TjcUUfKMHfMALyYl4kth6txrakdqXERkECCmmZjvz5URpMZo4oOcvpzEEJIKHPVH8mfScBThydi91O5nJ+X7fWbdjP5maX42/4LtdC3Oe6C/fGV27yPw2RmcENvRKpaiW0Lc5y+6KUAnE1m6tu7sPvUdac7oizNyKJkYTjxykPYcriaux+AEEJIv6bBFv7ezSRk8i9AMzN+JbY+SKoIKc6vzrd+bVtl13ZWxRqAna+Fvt1xAMaGLEyCrp6gf7kRQojPZmZqUHO7HRfr29weWzghw7o122gyo/j9L7CnvIbvIdrho2AeQDMzoiO2QAYA2jrt51ycdfpWyqWc1KShQIYQQth567FxWLL7U1bBjGUHrFDXGUc7Zf2NEoD9gM8ieL6IUrB/8RW//yWPIyFEvFQe/J0QkqxSYOrwRKTFKX06T/H7X7JeukmLjcCUN44IFsgI2ZPJgoIZPxCyxLMrHzw/jfWxR6vFF4y5Eh4GyOnVTTgwJlU8mwaI+OmykvBO4QOoaTa6P9iFPeXfsj7Hnk9r/F6aIi0uAlXF+aIIZABaZvILIUs8OyOXSpCq9uSTA3dV78IlvZ22vRElk6Cty/2du/lpNE5C0Mdf85+QT4LHijlZnH2APfpVA+RSCUxm4Zfo0+IiMFgT3W+XqlhQMOMHgxIiceKS0KO4y3Zbti1nCcAAMD0zEXs5SiiLVEhxT2osmlo6ER8pw+lrzazv284ikCGEECGkxUVAKZdy+gFWDIEMAJQsnS66AMYWBTN+wGVDR1fCw9zPSEwZEoc/LJkMwD54udrQZjeladlWbVkPfe2RezkLZgwdZpRd0Xt1X3H8WRNCSH+DNb3tYcT2AZYL6w9UOtwgIhaUVeAHlm7SfEhWKVA4IQMzMxNZLa18/HUzjCYzFu8qx6iig9h96jpOXGp0ujZbUlmPxbvKef0ZCDsR4fTnSoiYWRJ2X8zLFHgk3BNjuoQtmpnxk20Lc/DkjtMorW7k9LxHv+uC7UllXd3m4x4lp5VU1kPfakKyKgLqKJnTYn+EPymxCty80yn0MAghLqyYk4XH3z2F45fcd7UONEIXxXOHghk/2r4oF5M3HMYNji5Klr39q/Zd8Oh+3mTZZ68r8fg+hDvdVKOHEFHTZWkxds2Hoslx4ZqQHbHZoHlrPzv80gxOzpOiUli3xIl9+o8PsjDudlcFgvoWk9BDIET0wsMkEOKtQZelRcU1fdAGMmIoiucOBTN+ppRLOekcPWl4gvX/xT79x4eC8emoWKmjFzAhBAAwY6QGC8anY/KwRJ/zy2Zmalgd92h2ChbkpMNgNKGpvdunxxSr1LgI0dSScYWuBQLYvigXmmi5T+f42GZN1pPpvxSVwqfHdUYmAbTR/lu1XDEnC+poOX6Yk+a3xySEiNfRrxqsGxo6vtsNESmTwNO45tFxKaisNbA69q8VN7Gn/Fucvtrs4WgDx5DvdmiJHQUzAilfqWMd/bvTO9vj/lwSAEd+yc0yV19dDFDf6p9PJpZdVav2XcD1Jvd9Swghoam9i/G4gOZfz95EnYGS7S0CZeafEoAFtH3ReBhNZug2H0NNs2elqGeMuBu8LN5VjtLqBrf3YQA8uKkUk4fE45MArWpqCWQ82b1FCCHEO2JP/LWgmRmBKeVSfLzsIVQV56NwQgYmDlWzul/Rv98LwPMuqQ2tJvz3jx/gsDkBv34wLgVThyeicEIGqorzAUCUTTsJISTYBELirwXNzIiEUi61Vld0F6BYXmDeduMueOckKovzRT+7UTghw/qcGE1mvLbvPAUyhBDiB2Lphs2WhGGY4NxLZsNgMCA2NhZ37tyBShUYHXCdBTS2L7BV+y541SaBTdsDMUiKkUMikeB2mwmdQbrlkRBCxGSENgr/eG6qaGZk2F6/KZgRsb6NH1/My8SWw9XWry/Xt6Lsa+96HBFCCCF9VRXniyaQAdhfv2mZScT6Lj3ZVuEVexOz3EFxHnXDJoQQIqxAypHpixKAA4CnSb5C02Vpka4OjO18hBASCqYNS0DhhAyoI2UOvx9oOTJ90cyMyHmb5OuILEyCLp57/Fj+IPL+s5TXxyGEEMKObaCydt7ofikMK+ZkBeyMjAUFMyK3/kAlJ+eZNjwBKXFK7Cmv4eR8jlSs1EH9XWXj+lbqJUQIIf4yLDESDwxW45umdjQYjDB0diNMEoYZmVoUPXKP3bG2KQzBgoIZkbvc0MrJefhuSR8RLrEGMoQQQvzrcmM7Lje2QxMtR4PNh8k95d9iT/m3Ab+M5A7lzIjUDb0RQ5fvR9mVwNit1NHNwGgyW78eNSBwdo0RQkiwaHAyK15SWY/Fu8r9PBr/oWBGhEa8egCTN5Ui0Eqr2C6JjUiK8fj+uiyttV0BIYQQbpVU1tt96AwmFMyIzIhXD8AUaFHMd641tVv/39N+HimxCiSrItDR1QOFNFCaLRBCSGDhKg9TbCiYEZEbemPABjIAcOuOEYXvnsaqfRcAwKNZlpt3OrH71HWcuNQYktV+Hx2XIvQQCCEhwPZDZzChYEZEHt56XOgh+ORifRtOXGrE7lPXMaroIMzmHlo2YunmbaPQQyCEhIBBCcFZA4yCGRFp6wyutczS6kacu96MquJ8pMZFCD0cUWukreyEED/wNAUgUNDWbBGJUkhh6AiugKah1YRn/3gWnyx7qF+hpu7uHuz5lL+6N4GkoaVT6CEQQkLA83srgnKLNgUzIvLB89MweVPwVc4trW6A0WTuV6ip8N3TAo5KXG4bu5EUJcWttuAKZgkh4mLZ0RToFX/7omUmEUlVKxEepL8RRxn0wbp26y0KZAgh/hCMO5qC9NIZuBaMzxB6CLxwlEEfrGu3hBAiZsG4o4mCGZEJxhcZ4HgWRimXQkMtEAghxK+SVMH3vkvBjMgE69LLH09dx5jVB3FDf3cLstFkdlp6mxBCCD/+evYmctaVCD0MTlEwIzI1zfzXG5mZqeH9MfrqAWDoMGPyplKMePUAgOBctyWEkEDQ0GoKqoCGghkRMZrMOPpVA6+PMWOkBtsXjUdVcT7S4oWp/WIyMxjx6gHWS2oTh6qpTg0hhHCsodUEfZDMjlMwIyL+mKlIi1MC6M1X+fiVh5AQJeP9MR0xmRloWD721/VtuNHcwfOICCEk9BS8c1LoIXCCghkR8Ufyr+1j5KwrQVNbF++P6cy+c7WsjrtFBeUIIYQX9S00M0M45o/kX8vMjL7VJHjybQ9Au5kIIURA2pjgeA+mCsAismJOFnafus7rY+wp/xaNbZ34prGN18dhq6HVhJmZGpRW85srRAghpL+9SyYJPQRO0MyMiCjlUr90mS6prMfFenEEMwCQGq9EVXE+CidkIFmlEHo4hBASEjTRcqiDZHacghmR2bYwB+pI9km504Yn+CUA4tO1pnZr36bhSTFCD4cQQoKeJlqO8pU6oYfBGUGDmUGDBkEikdj927hxo90x58+fx9SpUxEREYH09HRs2rRJoNH6z9wxA1gdJw8D3ntqArYtzLHObHgSCImFba5QsBYNJIQQoUklwAhtFCpW6oIqkAFEMDNTXFyM2tpa679f/OIX1u8ZDAbMmjULAwcOxNmzZ/Hmm29i9erVeOeddwQcMf/Y9iz6fHW+9f+VcinqDB3Qtwu3O8lbtj8v9WsihBB+/Dg3A4eWTg+apSVbgicAx8TEIDk52eH3/vjHP8JkMmH79u2Qy+W45557cO7cOWzevBlLlizx80j9x5I7U1JZ7/QYXZbWroW70WR2ebxY9f052PzsfApD7y4rQgjhiyJcgs5uxu+PG8wfFgWfmdm4cSMSEhJw//33480330R3d7f1e2VlZXjwwQchl9+NImfPno3q6mrcvn3b6Tk7OzthMBjs/gWabQtznObC6LK02LYwx/62zcf8MCrudff0/4N29bPzrQdAQqTgMT4hJIh1djN+byvT94NjsBH0Xfv5559HdnY21Go1Tp48ieXLl6O2thabN28GANTV1WHw4MF290lKSrJ+Lz4+3uF5N2zYgDVr1vA7eD/YtjAHRpMZ6w9U4lpTOwYlRGLFnKx+L8jFu8pRE6AVco9+1YCcdSX91m+3LczB8r9+jj2f1vh9TE3t3e4PIoQQH1h2cfZ9fwdgve1qYytqbvv+3u7oA3CwkTAMw+lc17Jly/DGG2+4PKaqqgojR47sd/v27dvxs5/9DK2trVAoFJg1axYGDx6M3//+99ZjKisrcc8996CyshKjRo1yeP7Ozk50dt6tGmswGJCeno47d+5ApVJ5+ZOJk9Fkxqiig0IPw2czM3t7RtkKlp+NEBK44pRSNBvNnJ936vBE7H4q1+1xi3eVe73snhYXgZKl0wN6RsZgMCA2Ntbt9ZvzmZmXXnoJTzzxhMtjhgwZ4vD23NxcdHd349q1a8jMzERycjJu3bpld4zla2d5NgCgUCigUIRGvZJg6TxdWt0Ao8kc0H90hJDgoomWY/qIRPyl4ibn52a7c9N2hn7/+VrWmzxCYTbGFufBjEajgUbj3VrguXPnEBYWBq22N19i4sSJePXVV9HV1QWZrHfLcUlJCTIzM50uMYUaf/Rz8pf1Byqxdt5ou68JIUQI4RLgvow4XgIZAOg297D+AGepw7V23miXMzXqSBnmjhngMB0h2AmWM1NWVobTp09jxowZiImJQVlZGV588UX85Cc/sQYqP/7xj7FmzRo89dRTeOWVV/DFF1/gv/7rv7Blyxahhi06gxIiceKS0KPgRt/ALJgCNUJIYEmKjeB1V+We8hrsKa9BWpwSgzVRTnMi+7LM1BS//yWOVveOb3qmBq89cm/IBTC2OM+ZYauiogI///nP8dVXX6GzsxODBw9GYWEhli5dardEdP78eTz77LMoLy9HYmIifvGLX+CVV17x6LHYrrkFomDKKymckGE3M7Nq3wXee1URbkkACPKGQkiQCLXlIXfYXr8FC2b8KZiDGcC3BDExqSrO71c7J1gCtVAwLj0GZ79tEXoYhAQ8CmjuYnv9FrzODPHdtoU5iJIF9q/SUQ0EfzXeJNygQIaEOg1HlXVLKuthNHG/gyqYBfYVkFjJwgN3rdTVpxB/F9BTBO7TSAgRgDpShsIJGagqzkf5Sp21T56vF1faAOEZKnUaJLQxcjQbA68v0ycvz0SqWunymK0F2X5bbuqkD0OEEJYqVur69Tmy7Dz6x7kbMHR4/4ZCGyA8QzMzQWLvkklCD8FjEsBtIAOI6xNKWlyE0EMIeBE0+0WCgC5L67Jh46gBvuVnsq1DQ3pRMBMk1NFyztZr/SVWKWN1nJg+oRi7aOrGVz58WCVEFNgk6I5IivHpMYK5KSQfKJgJIuUrdQEV0Ghj2I1VTJ9Qmtq6IJdKhB4GIUQAI7RRqCrOZ7XTyJdgJNibQvKBgpkgU75Sh4qVOozQRiFOKYOYL7uOlsaMJjNW7buAwndPY9W+CzCazKL7hGIyM5g2PEHoYRBCWNBlaTnbRPCP56ayDjK83Y1J27K9Q3Vmgtyszcdwsb5N6GH0o4mW9+uU7axeji5Li6qbBlF1BlfKwpAaq8DlRqPQQ/FYRHgYOrp7hB4GIX5hCQ70rSZkryvx6VwKqQTnXpvt0awJmzpg6igZ5o4OzTYE7lDRPBuhHMxw8QfMNU8CGYuZmYkorW706vGoKi0hoa2qOB/rD1RyVlE8NS4CA9VKNLaakBCjwDBNdL9AxNIc8lpTO1LjIiCBBNf0bWgwdKDF2A1JmAQzMjUoCvE2BO4I1jWbiIslMbih1ST0UAA43optNJndfnIprW7EjJEaHP2qwekxuiwtthZkW99AbHud3NAb8fDW4z5tlSSEBKbX93+Jb/TczaLeaO7ADctMcX0byq7osfvUdcwYqUFSjAJ/r7gBk7n/RyhdlhZ7XprB2TjIXZQzEwLElBj89keX+93Gdut11c0WpDrZGm2ZSrbUeNj9VC7Wzhtt/cSTqlYidwjluRASis59e8cvGwmOftWAveU1DgMZoLey7+Jd5byPIxRRMBMiylfq8MnLMz26j6vEuZmZGq/G4WibNdut13WGu5+G0uIiMHV4orXypruEuWDpX+WKKkKKH2SnCD0MQkSntbNbNBsJqFUBP2iZKYSkqpXQZWldXtTjlFI8MjbVbv3Xdu3XsnQDALotx1Bz27Ok3LTYCKzad8HuXGmxnheiq2nuwKgUlV2XbWfYLGMFg2O/nAmlXIq/VNwUeiiEiMpQTaR1d5EY3gvWH6hk9d5F2KME4BDkatcQmy2BT+44g9Jq57kr/tS307Yjq/Zd4CzxT8wKJ2SgztAhijdrQsTk1TkjsfjBoQDEMUubECnD2aJZgo4hUNBuJhsUzPTnaLaFTUZ9zroS0SQTA0CKSoGTK/JcHlP47mmcuOTdTqhAoo6UQd8eeP25COGTBED1uochD7+bVWE0mVG07zz+VnETQhUpoHoy7NBuJuKSJVHWE0/uOC2qQAYAbho6YTSZXQZigxIiceKSHwclEApkCOlvyYOD7QIZoPf9780f3o+6FpNgH3QsuTO0LZsblABMWDGazKzrvCzISUfhhAxrgu6CnHRex+ZuN5RYEv8IIf4jAfCzBwdjuYu/f6FbpYipiW6go5kZwkrx+1+yPvbDyjpUrLq7Hlz47mk+hmRl2Q1l7mFw8lIj/vZZDdpNZuQMUuPxSYNElfjHF4VUgk4n20HdCQMEm2onhA/LZmfiyalD+s3I9LViTpag+XRiaqIb6CiYIayUfnWL9bH6ti676VO+l3kGJUTi4Be1WPrnz9Fus+XxUOUtrP+gCkumDsa2hTmYsvGIqFoicCkhSo6bhk6v7kuBDAk2N+4Y3QYyAAT/oCP0zFAwoWUmwsqdds9yZWynT/le5snJiMfTf6iwC2QsGAb4/UdXseFAJUqWTud1HELRRHsfyBASjDyZ8di2MAdRMmEuhbQEzh0KZgg7Es/6b9u+mXjbPZaNOGU4Xv+gyu1xv//oKu60d2HGSO+K/YnVzEyN6JKyCRGapzMeJ155iKeROKfL0lLyL4comCGsxCo9a4fQ981k28IcpDlpReCLnEFq3GphdzGfsPEIjrvo7RRoqorzkRqvdH8g8YhnYTsRI09nPCw97PzlweEJtC2bYxTMEFZmjvRsZsXRmwkfyzyz7kn26PhAzw8JDwMOvzAN1zbOhVIupQRCHgR94S0B+SNQ9HbGw1897BKiZNj11ATeHyfUUDBDWCl65B7Wxzp7M1HKpZiZmcjZmHRZWqTFh1YC3YLxGRiWHG39mhIISSDhO1BUR8l8mvEoX6lDxUodRmijEKeUQaXw/BIZJXd+H020HGdXUeVfPlAwQ1hhm/firqrl9kW5Hn36cXasLkuLrQXZ2PdZDetzBYP952sB9Nb9WbXvAv71eWj0YdJEy6GKoPwC4trc0QN8Poc6Wo5DS6fj3GuzcPpVzwIPTbQcXxY/jGsb59oFRSO0UahYqUP5Sp3P4yOOUTsD4hFnfU3S4pQoWTqN9fSus/5OaXFKDNZE2bVYcNR64fm9FUFdN8aVhCgZmtpCr9rv4RemIe/Xx4UeBhExNr3aPOVpLydqU8At6s1kg4IZbnnb14mr84ihURzxv8IJGfhT+bcweVkckCtxynA0G7sFHQPpj88gwtP3HD6CqlBFwYwNCmaCh9Fkxqiig0IPgwhg6vBE7H4qFyNePSB4QEOEEREeho7u/mn8aXERKFk6ndcAwmgy4z9+cwIX69vcHls4IcPj3nfEMWo0SYIS9TIJXZZk54uvz8ENvRGztpQiBFfbQpJcKsH8+1NR9O/3AgDyNh/DDZtq3jXNHRhVdBBp8REoeXE6AHAye2xLKZciKVbJKpihXYb+R8EMCSj0JhG6btw2Wv8/Va3El2vnYvCy/bSVOgSYzAz2fFqDPZ/WQBMtd1oosuZ2R7+Z2xOXgN2nrnOyDMW2NQvtMvQ/2s1EAgq9SYSu0uoGLN5VbnfblOHcbfUngcHbitcllfWY8sYRGB20PWGLbTG+F/MyvX4M4h0KZkhAoV4moa2kst7uYhQqwa2CRdNE4p5l5qZvUMwW2xIV2etKfA6ciGfoL4QEFKVciihZ8OwSSIrxXwn1YGHJmzKazOgyB0ZN50fHpXh9X12WFg+OoBkoLpVU1nsV0BhNZiSrIqCOlLk91hI4Za89hFX7LlBgwzPazUQCjr7VhOx1JUIPw2fqSBk+WfYQ7c7y0NThiYiQhQXU9vxklQLhYRLU2CStupMaF4HD37UAodcIPzzZQs1FSQiqQeM52pptg4KZ4JO58gA6uwP7pSsLA3q++xFopzF7aXERHgUFgUyXpUXVTUPI/Lz+lhYfgY9ZdMzmsrYVBTSeYXv9pmUmwit9qwmzNh/DfWsOYdbmY9B7mbzX17mi2ZycR0hdPb1BDAUyngmGCzvbN96Syvqg+HnFquZ2h9vlJqPJzOksYEllPWfvg+QuCmYIb3LWlSB7XQku1reh2diFi/VtyF5XghwOlojYJuKR4JIaF8H62AU56Zg4VM3jaLwXGJk+oaFvUnlffNS2yl5X4nUSMnGMghnCi5x1JU63UDa0mhwGNEaTGb/682cYteoAhizbj8wV+zFj0xEseKfMYQLdtoU5mJmp4WX8RHxSYhUYool2f+B3wqUS7Fk8kbeg94EMFdLi2QdXoWpwgvifI1cBC1+1rbxNQiaOUdE8wjl9q8ltLYiGVhNu6I1IVSsBOF6T7uwBruo7cFXfgbKv9Q4LX21fNB6Ldp7B0a/6N63kkiI8DJ0OyqizkRKrwFBtDE5daUQXfST32lBtDOuiZQDw/ue16DYzSIiSQx0pg76d23LBF2+1wdBJO1Rc0UTLkaaOxtUmcS+VuQpY0uKUvD2uZVaI+jj5jmZmCOcK3jnJ6rjJm0ox4fUSDF62n/WatKNPMzueGI8ouWcv5YQo91srbX3//lRIPLrHXTfvdCJCFmZN9iXeSVLJPaoz1Gzswp7yb7G3vMYukJFJAaUsDNoYOQpy0lCxUufVeCiQcS0hSobylTpegwGuuKxXxPMeGWrRwg2amSGcq29hn9xW58GxFo4+zcikUrDNRLDM7li6du8/X+v2U/ue8m89Hqetksp6qBRhMHTS1Iy3zl1vhlIuxczMRJRWN3p9ni4zMD0z0W6GT5el9TjJM0wC3gNUyxJZIG1Dt2hq63K53CwmroLkmjv8zipRixZu0MwM4ZzWD4Xg+n6aYfuYwxKV1ouYUi7F2nmjUVE0yy/JxHkjKWHZF42tvQHn9kW50ET79hrrm/TpzVZZGc/vnpage9vCHEi9nRYUWCAEMgDw/N4Kp9/ju8p0qFSx5hsFM4Rze5dM4v0x+n6aYfuYf356isPbty3MQVVxPgonZGDq8EQU5KT5PMa+Gtq7IQ/Uq5II2Aas5St1SIj0bWL5ka3H7QIaT4MTPleZCnLS7AKsBwbF8fdgxOWOJr5bqFCLFm5QMEM4p46W+/zJ2Z2+n2bYPKYmWg61i2MsMzW7n8qFTMr9n8aghEhcfH0OBTResg1YjSYzmtq7fTrf5UajXZ+eD56f5tP5uHSsusHu4joiiYp98s1Z7gqfZSB0WVpK/uUIBTOEF+VeJlWyVdNsdPiYzgIaRbgEIweoWPdIYbuOHaeUsZ7FsXwCu/j6HHx/bDKr+5BefQNRLpMmSyrrcU/RB1BHy71O8uZanaHTLtCiT+/8c/U3v21hDhJY9GPyBFUC5hYFM4Q3I7RRvJ376FcNDms0lK/UoWKlDiO0UYhTyhD13dpBZzeDE5casfvUdYwqOojJG4+g8N3TToMbtuvYj4wdgI3zx7r95Nb3E1iDj7MKoUQTLe8XHHOdNNlm6kH2uhLEKaWiCWiAu7v3qEgk/9z9zc8ZM4Czx6pYqaNAhmMUzBDe8J0742id22gyY8vhaiTFKhEdIUWbk8IuN5o77IKbvp1t2X4Sthy3bWGO04uNo09gviT98b2EJwbhkt5guGKlzuEsH19Jk7eNvb//wy9Ms+bQyMJ6v546XJjO1ZbXuavXGPGdu795rmbHdFlal8vdxDsUzBDe+CN3ZtW+8zCazFi17wLGFR/CqKKD2H3qOk5cakTNbfZbKvVtXdbAxvJJ2F114b6zLbZJxJOGqDEsMRJJMXJcqDFg2d8+twZKRpMZ3V42ZAoP6519WpCT7tX9A0U3AwxMjHL6ps/nsgsD4EfvfIJL6+fi2sa5uLR+LoYlRwu66+Q/fnPCGtBYXmNibdUQiNjkrnAxO0ZLS/yhrtmEd3zWmggDP31uNNFyl2OeNjwBA+KUOFbdCIDBjEwtih65B0q5FE/uOIPSascVid2d1x1VhBRj0+Nx6VYL6gydXp8nUEwcqsYwTTRWzMnqd7HhspOxIxUrdXbBlNFkxqiig7w9Hhu2F8NV+y5g96nrgo4nGHgaYDh73c3M1CA1XolrTe0YlBCJF/MyseVwtfVrR69h4h7b6zcFM8Qv9K0mFLxzEvUtJmhj5EiOjcBHl5qEHhbnwsMAL7seEDccXXT4DGhGaKNwaOl0vz0eW5bnofDd0zhxyfvigaEqLT4CgxOjfQowLAU3KVDhHwUzNiiYEScxXBhIYHEU0NheWLi8uMcpZTj32qx+t4vhdVuxUoeCd07iYn2boOMINLTME3gomLFBwYx42V6I0uKUPrcNIMGvqjjf6afgMasPwtDBTTU7RzMzFk/uOO11SwWpBBicEAlDZzfutHeh08v8KeKZcekx+NuzDwo9DOIhttdv3hKAX3/9dUyaNAmRkZGIi4tzeMz169cxd+5cREZGQqvV4le/+hW6u+23rB47dgzZ2dlQKBQYNmwYdu7cydeQiQBsC9VtmD+GdmsQt1zVmOGy8J2z3XiLd5X71Bvqx7kZOPzLGTj+q5kUyPjR+RstQg+B8Ii3YMZkMuEHP/gBnnnmGYffN5vNmDt3LkwmE06ePIn33nsPO3fuRFFRkfWYq1evYu7cuZgxYwbOnTuHF154AT/96U/x4Ycf8jVsIjC2208jwmkjXqhyVWMmVa3kpMKys2rRRpPZ5yUmy06sUOyWHBEuwaPZKV53KvcFxY3Bjfdlpp07d+KFF15Ac3Oz3e0ffPAB/u3f/g03b95EUlISAODtt9/GK6+8goaGBsjlcrzyyivYv38/vvjiC+v9CgoK0NzcjIMH2e8qoGWmwGO7/JQcI8dn3zajsbUL2hg59i6ZBKVcKvjOEiKMZJUCuqwkl0mXI149AJOXVy9HRfosfN1BZJuz8eN3ynDya73X5wpEhRMysHbeaAD+zz1SRUhxfnW+3x6PcIPt9du3Tm0+KCsrw+jRo62BDADMnj0bzzzzDL788kvcf//9KCsrQ15ent39Zs+ejRdeeMHluTs7O9HZeXfbqsFg4HTshH+W5SdXdFlawRMxxSJc0lubJRTUGTqx+9R17D51HTMzE7F9UW6/Yy6+Pgc39EZM3lTq9DzJKgUmDY7H+ZsGu0DZVUEzXyoP2wYyi3eVh1wgA9g/f9sW5vg1oBFT7y3CPcGCmbq6OrtABoD167q6OpfHGAwGGI1GKJVKh+fesGED1qxZw8OoiZhsW5iDKRuPoKaZfXG8YCSXSryehQh0pdWNGLf2EM6u6r/rKFWtxLWNc3FDb8TDW4+jrdOMKIUUHzw/Dalqx+8dlgKMzrbcDkqIxIlLno1xSIIS+//fNOt5/HkBV4QBnSIqFXDmqn05hm0Lc2A0mfEfvznB686scAmc/s5JcPAo8WDZsmWQSCQu/3311Vd8jZW15cuX486dO9Z/335LO2SCVYmT3SahYtowdcgGMhZNbV0YV+w8jy5VrcT51fm4smEuzq/Od3pRW7yr3K6CtG1FaAtvKg/bBjJc5Ny4M35QHAonZKCqOB8/HJ/B62N5qrObQc66ErvblHIp/vHcVF4f9/KGubyenwjPo2DmpZdeQlVVlct/Q4YMYXWu5ORk3Lp1y+42y9fJyckuj1GpVE5nZQBAoVBApVLZ/SPBKZQa8FWs1KFwQgamDk+0XqxqQ6AKMBtN7d0uAxp3XM2WWJo9AviuzQX7Hk19y+T7I+k3M1mFtfNGQymXYu8Z8X2Qa2g1Qd+nCjZff8dxEWG4tpECmVDgUTCj0WgwcuRIl//kcna9eCZOnIgLFy6gvv7uG0hJSQlUKhWysrKsxxw5csTufiUlJZg4caInwyZBLhQa8Fma01m2sVsuVvUt/LSJCERN7d14cscZj+/HZrakpLIe+lYTVu27gK4eQBHufseUowJtXHf7dsTyGPpWE7p6xDlrV/DOyX63uWvW+snLM+Fso9rkofGoWKnDCG0U4pQya5PSc6sf5nLYRMR4y5m5fv069Ho9rl+/DrPZjHPnzgEAhg0bhujoaMyaNQtZWVkoLCzEpk2bUFdXh5UrV+LZZ5+FQqEAADz99NP4zW9+g5dffhlPPvkkSktL8ec//xn79+/na9gkQFnW3nWbjwVdDs3MzEQkqyJQ+O5pax4H0Pspv9nYJfDoxKW0ugFGk9nhLqfefJjz2H+hDt09DAaqlfjzzyZjy+FqVufO7rM84kxytAxHX37I4Ri8ybnxlKUhpqOAQSycBeGWv2NnrQKubJjr8vvOihyS4Mfb1uwnnngC7733Xr/bjx49iunTpwMAvvnmGzzzzDM4duwYoqKi8Pjjj2Pjxo0ID78bYx07dgwvvvgiKisrkZaWhlWrVuGJJ57waCy0NTu0iKHcPBcW5KTjlqHDadNK4lhBTho2zh9rd9G72tDqNMiVhUl8msFIjYvAEA27Xj/+aFY5LDESjW1duGPsgjjnZVxXVybEFrUzsEHBTOjxZpZmZqZGNIGDZbo9GIKyvmRhQBePO2ySVRHIGhDj19+lqxYLfQVLsO2Lvh3JCXFG8HYGhAhJKZfi42UPsU7W7K1XMp517s3kofG4tnEuL7k6uiwtthZkB+UFL2dgLK+BDADUCTCb5UlibyjkeLnirLoyIb6gmRkS9J7cccblxW1mpgbbF423fu3sk7M6Soa5owf0W0rou4Z/47bR4eNpouVoaO2fK+BomcLXSrNiNSRBia+bjEIPg3NThydi91P9i/e50vd1U/JlLepagjsHylJd2VXeCyG2aJnJBgUzxFmA4mjHCdD/QuPpm62z+7M9b+G7p3HikvfNDIl/2Zbp9xbbfJoHhyfgo0tNbo8TUmxEOBKiZAAY6NvNdtWVPf1bJKGNghkbFMwQwPcAxRf6VhMK3jmJ+hYTq7L5wTozE6zc5cywfe25y6dxNrvHRhiA6AgpDB1m1vfRZWnx8aVGGD1cGyyckIEVc7L6/czP761w+fPNzEzEW489QLM2xIqCGRsUzBAh5awrcXgBctXQ0B+7Xgg33M0oPLbtJD65cpv1/ZwFNDMzE1Fa7f1s3QhtFHKHJLAKkkdoo/CP56Zag4gH1h5CYxv7JTB1pAz6du6WzGjWJnRRAjAhIuAskAF6K6H2Le1uoZRLoY6U8Tk0wgHbi6ylr1Phu6exat8FGE1mDF6232EgA9hXFra1bWEOqorzUTghAxOHqjFCG4VJQ+JRc9u3XKO9SyaxLtqXFKu0mw35dNUsVKzUIZJlZTIuAxnA+XNFiIVgjSYJCXb6VpPbJQFLaXdHS05zxwygpSYRW5CThg3zxwLoP5ty4hJY/e5KKusdFvlTyqWoM3Sg7EpvZ21fmzBadhCxLdpnKbxnYTSZMW5diaB1a5w9V4QANDNDCG/YVmB1dpw3TQ2J/xQ9ci8A3+vGONrWzXUtmtGpKhhNZnSzbEpq+9qzNOAUQz5C8ftfCD0EIlI0M0MIT9j2TXJ2nKX5nljqzfBd7C6QpMVFWHen+fr76bv0w0dn7dLqRtY5WLbNMT0NqnytpuzOUZEUtSTiQzMzhPBEG8OuMJir48RUYO3YL2eyOu7R7BSPOksHosGaaADcdMHuu6Tjj87azvTNAfI0qBJrY0sS/CiYIYQne5dM4uS4bQtzsCAnnYshAejNn/CULkuLVLUSMjfvGBIAMqkUSTERiFUEb26DJQDhogt23+VEf3TWdsZsswwlZFDlzIxMcQT2RHxomYkQnqij5W7rgrAt7V70yD3YU/6tx2OoWKnDlsPV/Wp2eLJ8YPm0/uSOM26XmRjAq3EGGksA4msX7MlD4/sltPqjs7YzpdUNePx/ypCRGI33P6/16hzxSiluG9nXsvHELQP7XmsktFCdGUJ45k2dGUc8zV9wV5vDUsht//lah1tpk6NlOPryQx4HP8HO9nm9oTdi8qZSr891bePcfrcFeo2hwgkZOPhFndfF/dzxpKknCXxUNM8GBTNEaJ5WAHaGbVDhaZExVxVqA/3iyqWESBnOFs2yfj1r8zGvt00vyEnHhvljHH7PVX0isbMEGwvfPcVL2wUuWkeQwMH2+k3LTIT4gTpajkNLp/t8nm0Lc/oFHi/mZTpcSvKEUi51eoHwV+6ELkuLto4unPzacZE5MZgzZoDd1183eF//hXGy2dloMgdsIGO7E2rXUxPcNnn1hpA5RUS8KJghJMA4Cjz4/KTqr4vH6n+7Fw9vPe6Xx/KWbbKu0WRGtw/z2jeaHed/iDHxlg1Hs4HbF43nPKDpu/uLEIB2MxFC3EiLU/rlcSZvKvWoCaK/2c46AL4HHc4uyoE285AWp0RVcb7TZc3ti8ZzulWfikkSRyiYIYS4FvxpdW45mnXwNehwdlEOpJkHXZYWHy+b6XZZc/uiXE7qJfUNKAmxoGUmQohLNXdCbztsSqwCQ7UxLnOQfNlCPTMz0elFecWcLNH25FqQk4aa5g6vcrP65nslxchx7ttmNLZ2ITFahtaOLtS1OG9QSZ2ziSsUzBBCXBKy7okQNNFynFye5/Y4b4OOhCgZti/Kdfp9sbWxsNBlaa2NNb3lKtEcuLur7nJDK5paOpEYLcdQbYxXSe0ktNDWbEKIS2LYmh0ugU/JtmzNGKnBjifGsz7e0/o7amU4Kl6bzcu5+USzIkQotDWbEMIJMcwUDFQrMSQphtMxpKoUeDBTi5pmo9db2rctzGEddCREhuNsEbtARkwWPJCGDY/6NiNDCN8oAZgQ4pbQDS9rmo3YtjAH04apOTvnDUMn9pR/iwhZGNbOG+31Msa2hTmoKs5H4YQMqCNlDo+ZmanxKJDho3O2t0IxZ4oEHpqZIYSwYpvAeeB8LZoctEDgS6e5t4ry8ct6zs9dUlmPxbvKfVpGseSCrJ032mU1ZbbEVGsmkHZXkdBFOTOEEK/4O6djhDbK69YBbIip50/hu6dx4lKj0MMAIK7nhYQettdvWmYihHhl28IczMzU+O3x6lv4LfE/Zs2HmLX5GPQiaCUgltkQqutCAgUFM4QQrxhNZs777riijfG8MacnuswMLta3IXtdCcatPcTrY7kjhiq3tIOJBBIKZgghXvF3XsfeJZP89lhNbV0Y8eoBvz1eX0q5FGlxEX59zLS4CEwdnojCCRku2xMQIkaUAEwI8Yo/ewjpsrRQR8v9ukXcZGbw+LtleO+piX55vL5Klk73qb6PXCqByew8JTItTonBmiivk5QJEROamSGEeMVfeR22yx3bFuYgLd5/MxbHL+lhNAnT/NJS38cVXZYWVcX5+EF2CpQyCWRSYFhiJCpW6nDx9TlO72/pqbT7qVyftqUTIha0m4kQ4hW+KwN/b2wy1s+/r9+F1t8ViVNUCpxc4b69AV+c7Rpjm9PCxVZxQoRCFYAJIbziszKwLkuLLQvG+f1xHblp6ITRZBYsAOjboNHTgMRdPyRCggHNzBBCfMJ1vRm2Mw7+rHNTOCGDAgJCBEAzM4QQv+g7c5AWpwTAoKa5A4MSInHg/E00tXf3u58mWo6PXp7p9YxD38e9dcfIW1E9fyY7E0I8RzMzhBDe6VtNKHjnJOpbTNDGyLF3ySSoo7mtG8Nn1VyamSFEGDQzQwgRDXW0HIeWTuf1MQYlROLEJX7OLYYidoQQ52hrNiEkKPAVcEwZEke7fwgROQpmCCFBgU1dFk9JAPxhyWROz0kI4R4FM4SQoLFtYQ5nAY0sDLi6cS4n5yKE8ItyZgghQcVRXZYX8zKx5XA1rjW1Y0CMHBXXm9HY1gVtjBw7nsjF//dhJfadr0MPA6gUUnzw/6YhVa0U+kchhLBEu5kIIYQQIkpsr9+0zEQIIYSQgEbBDCGEEEICGgUzhBBCCAloFMwQQgghJKBRMEMIIYSQgEbBDCGEEEICGgUzhBBCCAloFMwQQgghJKBRMEMIIYSQgBYS7QwsRY4NBoPAIyGEEEIIW5brtrtmBSERzLS0tAAA0tPTBR4JIYQQQjzV0tKC2NhYp98Pid5MPT09qK6uRlZWFr799lvqz8QBg8GA9PR0ej45Qs8nt+j55BY9n9yi55M9hmHQ0tKClJQUhIU5z4wJiZmZsLAwpKamAgBUKhW9eDhEzye36PnkFj2f3KLnk1v0fLLjakbGghKACSGEEBLQKJghhBBCSEALmWBGoVDgtddeg0KhEHooQYGeT27R88ktej65Rc8nt+j55F5IJAATQgghJHiFzMwMIYQQQoITBTOEEEIICWgUzBBCCCEkoFEwQwghhJCAFnTBzOuvv45JkyYhMjIScXFxDo+RSCT9/u3du9fumGPHjiE7OxsKhQLDhg3Dzp07+R+8CLF5Pq9fv465c+ciMjISWq0Wv/rVr9Dd3W13DD2fzg0aNKjf63Hjxo12x5w/fx5Tp05FREQE0tPTsWnTJoFGK35vvfUWBg0ahIiICOTm5uLMmTNCDykgrF69ut/rcOTIkdbvd3R04Nlnn0VCQgKio6Mxf/583Lp1S8ARi8tHH32ERx55BCkpKZBIJNi3b5/d9xmGQVFREQYMGAClUom8vDxcunTJ7hi9Xo/HHnsMKpUKcXFxeOqpp9Da2urHnyJwBV0wYzKZ8IMf/ADPPPOMy+N27NiB2tpa67958+ZZv3f16lXMnTsXM2bMwLlz5/DCCy/gpz/9KT788EOeRy8+7p5Ps9mMuXPnwmQy4eTJk3jvvfewc+dOFBUVWY+h59O94uJiu9fjL37xC+v3DAYDZs2ahYEDB+Ls2bN48803sXr1arzzzjsCjlic/vSnP2Hp0qV47bXXUFFRgbFjx2L27Nmor68XemgB4Z577rF7HX788cfW77344ot4//338Ze//AXHjx/HzZs38f3vf1/A0YpLW1sbxo4di7feesvh9zdt2oStW7fi7bffxunTpxEVFYXZs2ejo6PDesxjjz2GL7/8EiUlJfjXv/6Fjz76CEuWLPHXjxDYmCC1Y8cOJjY21uH3ADD/93//5/S+L7/8MnPPPffY3fajH/2ImT17NocjDCzOns8DBw4wYWFhTF1dnfW23/3ud4xKpWI6OzsZhqHn052BAwcyW7Zscfr93/72t0x8fLz1+WQYhnnllVeYzMxMP4wusIwfP5559tlnrV+bzWYmJSWF2bBhg4CjCgyvvfYaM3bsWIffa25uZmQyGfOXv/zFeltVVRUDgCkrK/PTCANH32tMT08Pk5yczLz55pvW25qbmxmFQsHs2bOHYRiGqaysZAAw5eXl1mM++OADRiKRMDdu3PDb2ANV0M3MsPXss88iMTER48ePx/bt2+3ai5eVlSEvL8/u+NmzZ6OsrMzfwxS9srIyjB49GklJSdbbZs+eDYPBgC+//NJ6DD2frm3cuBEJCQm4//778eabb9ot05WVleHBBx+EXC633jZ79mxUV1fj9u3bQgxXlEwmE86ePWv3WgsLC0NeXh691li6dOkSUlJSMGTIEDz22GO4fv06AODs2bPo6uqye25HjhyJjIwMem5ZuHr1Kurq6uyev9jYWOTm5lqfv7KyMsTFxeGBBx6wHpOXl4ewsDCcPn3a72MONCHRaLKv4uJizJw5E5GRkTh06BB+/vOfo7W1Fc8//zwAoK6uzu7iDABJSUkwGAwwGo1QKpVCDFuUnD1Xlu+5Ooaez17PP/88srOzoVarcfLkSSxfvhy1tbXYvHkzgN7nb/DgwXb3sX2O4+Pj/T5mMWpsbITZbHb4Wvvqq68EGlXgyM3Nxc6dO5GZmYna2lqsWbMGU6dOxRdffIG6ujrI5fJ+eXNJSUnWv3PinOU5cvTatH2f1Gq1dt8PDw+HWq2m55iFgAhmli1bhjfeeMPlMVVVVXbJaq6sWrXK+v/3338/2tra8Oabb1qDmWDH9fNJ+vPkOV66dKn1tjFjxkAul+NnP/sZNmzYQOXOid88/PDD1v8fM2YMcnNzMXDgQPz5z38O+Q8cRPwCIph56aWX8MQTT7g8ZsiQIV6fPzc3F2vXrkVnZycUCgWSk5P7ZenfunULKpUqKP6ouXw+k5OT++0WsTx3ycnJ1v8G8/PpiC/PcW5uLrq7u3Ht2jVkZmY6ff6Au88xARITEyGVSh0+V/Q8eS4uLg4jRozA5cuXodPpYDKZ0NzcbDc7Q88tO5bn6NatWxgwYID19lu3buG+++6zHtM3Ub27uxt6vZ6eYxYCIpjRaDTQaDS8nf/cuXOIj4+3fgqeOHEiDhw4YHdMSUkJJk6cyNsY/InL53PixIl4/fXXUV9fb50iLSkpgUqlQlZWlvWYYH4+HfHlOT537hzCwsKsz+fEiRPx6quvoqurCzKZDEDv85eZmUlLTDbkcjnGjRuHI0eOWHcn9vT04MiRI3juueeEHVwAam1txZUrV1BYWIhx48ZBJpPhyJEjmD9/PgCguroa169fD+q/Y64MHjwYycnJOHLkiDV4MRgMOH36tHWn6MSJE9Hc3IyzZ89i3LhxAIDS0lL09PQgNzdXqKEHDqEzkLn2zTffMJ999hmzZs0aJjo6mvnss8+Yzz77jGlpaWEYhmH++c9/Mtu2bWMuXLjAXLp0ifntb3/LREZGMkVFRdZzfP3110xkZCTzq1/9iqmqqmLeeustRiqVMgcPHhTqxxKMu+ezu7ubuffee5lZs2Yx586dYw4ePMhoNBpm+fLl1nPQ8+ncyZMnmS1btjDnzp1jrly5wvzhD39gNBoNs3DhQusxzc3NTFJSElNYWMh88cUXzN69e5nIyEjm97//vYAjF6e9e/cyCoWC2blzJ1NZWcksWbKEiYuLs9ttRxx76aWXmGPHjjFXr15lPvnkEyYvL49JTExk6uvrGYZhmKeffprJyMhgSktLmU8//ZSZOHEiM3HiRIFHLR4tLS3W90cAzObNm5nPPvuM+eabbxiGYZiNGzcycXFxzD/+8Q/m/PnzzH/8x38wgwcPZoxGo/Uc+fn5zP3338+cPn2a+fjjj5nhw4czCxYsEOpHCihBF8w8/vjjDIB+/44ePcowTO9Wt/vuu4+Jjo5moqKimLFjxzJvv/02Yzab7c5z9OhR5r777mPkcjkzZMgQZseOHf7/YUTA3fPJMAxz7do15uGHH2aUSiWTmJjIvPTSS0xXV5fdeej5dOzs2bNMbm4uExsby0RERDCjRo1i1q9fz3R0dNgd9/nnnzNTpkxhFAoFk5qaymzcuFGgEYvff//3fzMZGRmMXC5nxo8fz5w6dUroIQWEH/3oR8yAAQMYuVzOpKamMj/60Y+Yy5cvW79vNBqZn//850x8fDwTGRnJfO9732Nqa2sFHLG4HD161OF75eOPP84wTO/27FWrVjFJSUmMQqFgHnroIaa6utruHE1NTcyCBQuY6OhoRqVSMYsWLbJ+cCSuSRjGZk8yIYQQQkiACdk6M4QQQggJDhTMEEIIISSgUTBDCCGEkIBGwQwhhBBCAhoFM4QQQggJaBTMEEIIISSgUTBDCCGEkIBGwQwhhBBCAhoFM4QQQggJaBTMEEIIISSgUTBDCCGEkIBGwQwhhBBCAtr/D6oU8gWQzXtgAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "all_embeddings = np.concatenate(all_embeddings, axis=0)\n",
    "\n",
    "tsne_embeddings = TSNE(n_components=2).fit_transform(all_embeddings)\n",
    "\n",
    "\n",
    "plt.scatter(tsne_embeddings[:, 0], tsne_embeddings[:, 1], label='Entity')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c98cecbc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cosine Similarity with 'fusion of lobes of lung': 0.5658900597281417\n",
      "Cosine Similarity with 'history of total lobectomy of lung': 0.2657808261321254\n",
      "Cosine Similarity with 'Accessory lobe of lung': 0.23197544447120108\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "\n",
    "# Sample data\n",
    "data = {\n",
    "    'text': [\"lobes of the lung\", \"fusion of lobes of lung\", \"history of total lobectomy of lung\", \"Accessory lobe of lung\"]\n",
    "}\n",
    "\n",
    "df = pd.DataFrame(data)\n",
    "\n",
    "# Initialize TfidfVectorizer\n",
    "vectorizer = TfidfVectorizer()\n",
    "\n",
    "# Fit and transform the data\n",
    "tfidf_matrix = vectorizer.fit_transform(df['text'])\n",
    "\n",
    "# Calculate cosine similarity\n",
    "entity_tfidf = tfidf_matrix[0]  # Assuming 'lobes of the lung' is the entity\n",
    "concept1_tfidf = tfidf_matrix[1]  # 'fusion of lobes of lung'\n",
    "Gold_concept_tfidf = tfidf_matrix[2]  # 'history of total lobectomy of lung'\n",
    "concpt2_tfidf = tfidf_matrix[3]  # 'Accessory lobe of lung'\n",
    "\n",
    "# Reshape the vectors for cosine similarity calculation\n",
    "entity_tfidf = entity_tfidf.reshape(1, -1)\n",
    "concept1_tfidf = concept1_tfidf.reshape(1, -1)\n",
    "Gold_concept_tfidf = Gold_concept_tfidf.reshape(1, -1)\n",
    "concpt2_tfidf = concpt2_tfidf.reshape(1, -1)\n",
    "\n",
    "# Calculate cosine similarity using the reshaped vectors\n",
    "similarity_entity_concept1 = cosine_similarity(entity_tfidf, concept1_tfidf)\n",
    "similarity_entity_Gold_concept = cosine_similarity(entity_tfidf, Gold_concept_tfidf)\n",
    "similarity_entity_concpt2 = cosine_similarity(entity_tfidf, concpt2_tfidf)\n",
    "\n",
    "# Print the results\n",
    "print(\"Cosine Similarity with 'fusion of lobes of lung':\", similarity_entity_concept1[0][0])\n",
    "print(\"Cosine Similarity with 'history of total lobectomy of lung':\", similarity_entity_Gold_concept[0][0])\n",
    "print(\"Cosine Similarity with 'Accessory lobe of lung':\", similarity_entity_concpt2[0][0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bb7cc5cc",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "42568cb3",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\annem\\anaconda3\\lib\\site-packages\\paramiko\\transport.py:219: CryptographyDeprecationWarning: Blowfish has been deprecated\n",
      "  \"class\": algorithms.Blowfish,\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cosine Similarity with 'fusion of lobes of lung': 0.81125736\n",
      "Cosine Similarity with 'history of total lobectomy of lung': 0.53987324\n",
      "Cosine Similarity with 'Accessory lobe of lung': 0.4561095\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from gensim.models import Word2Vec\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "# Sample data\n",
    "data = {\n",
    "    'text': [\"lobes of the lung\", \"fusion of lobes of lung\", \"history of total lobectomy of lung\", \"Accessory lobe of lung\"]\n",
    "}\n",
    "\n",
    "df = pd.DataFrame(data)\n",
    "\n",
    "# Tokenize the text\n",
    "tokenized_text = [text.split() for text in df['text']]\n",
    "\n",
    "# Train Word2Vec model\n",
    "word2vec_model = Word2Vec(sentences=tokenized_text, vector_size=100, window=5, min_count=1, workers=4)\n",
    "\n",
    "# Calculate the average vector for each text\n",
    "def average_vector(tokens, model, vector_size):\n",
    "    vector_sum = sum([model.wv[token] for token in tokens if token in model.wv])\n",
    "    return vector_sum / len(tokens)\n",
    "\n",
    "df['average_vector'] = df['text'].apply(lambda x: average_vector(x.split(), word2vec_model, 100))\n",
    "\n",
    "# Calculate cosine similarity\n",
    "entity_vector = df['average_vector'][0]  # Assuming 'lobes of the lung' is the entity\n",
    "concept1_vector = df['average_vector'][1]  # 'fusion of lobes of lung'\n",
    "Gold_concept_vector = df['average_vector'][2]  # 'history of total lobectomy of lung'\n",
    "concpt2_vector = df['average_vector'][3]  # 'Accessory lobe of lung'\n",
    "\n",
    "# Reshape the vectors for cosine similarity calculation\n",
    "entity_vector = entity_vector.reshape(1, -1)\n",
    "concept1_vector = concept1_vector.reshape(1, -1)\n",
    "Gold_concept_vector = Gold_concept_vector.reshape(1, -1)\n",
    "concpt2_vector = concpt2_vector.reshape(1, -1)\n",
    "\n",
    "# Calculate cosine similarity using the reshaped vectors\n",
    "similarity_entity_concept1 = cosine_similarity(entity_vector, concept1_vector)\n",
    "similarity_entity_Gold_concept = cosine_similarity(entity_vector, Gold_concept_vector)\n",
    "similarity_entity_concpt2 = cosine_similarity(entity_vector, concpt2_vector)\n",
    "\n",
    "# Print the results\n",
    "print(\"Cosine Similarity with 'fusion of lobes of lung':\", similarity_entity_concept1[0][0])\n",
    "print(\"Cosine Similarity with 'history of total lobectomy of lung':\", similarity_entity_Gold_concept[0][0])\n",
    "print(\"Cosine Similarity with 'Accessory lobe of lung':\", similarity_entity_concpt2[0][0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "32fb3b1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "from sentence_transformers import SentenceTransformer, InputExample\n",
    "from sklearn.manifold import TSNE\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm\n",
    "excel_file_path = r\"C:\\Users\\annem\\OneDrive\\Desktop\\ohdsi concepts\\Usagi_eval.xlsx\"\n",
    "df = pd.read_excel(excel_file_path)\n",
    "pretrained_model = \"FremyCompany/BioLORD-2023\"\n",
    "model = SentenceTransformer(pretrained_model)\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "85f1520b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f79d86c9464f494182af0ce622d656f2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "modules.json:   0%|          | 0.00/229 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\annem\\anaconda3\\lib\\site-packages\\huggingface_hub\\file_download.py:149: UserWarning: `huggingface_hub` cache-system uses symlinks by default to efficiently store duplicated files but your machine does not support them in C:\\Users\\annem\\.cache\\huggingface\\hub\\models--FremyCompany--BioLORD-2023. Caching files will still work but in a degraded version that might require more space on your disk. This warning can be disabled by setting the `HF_HUB_DISABLE_SYMLINKS_WARNING` environment variable. For more details, see https://huggingface.co/docs/huggingface_hub/how-to-cache#limitations.\n",
      "To support symlinks on Windows, you either need to activate Developer Mode or to run Python as an administrator. In order to see activate developer mode, see this article: https://docs.microsoft.com/en-us/windows/apps/get-started/enable-your-device-for-development\n",
      "  warnings.warn(message)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4b1efd32bbab4db3a33aa09f3c06dc98",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "config_sentence_transformers.json:   0%|          | 0.00/124 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0b2c7bcb504f49f79cc2d4654647cdde",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "README.md:   0%|          | 0.00/6.65k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fe7edca5577640fa8be40c5ffe6f2227",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "sentence_bert_config.json:   0%|          | 0.00/53.0 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4e35fee49034656aeb5e7d366ae0217",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "config.json:   0%|          | 0.00/649 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a29eb1102a114635affc32869fadaad1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "pytorch_model.bin:   0%|          | 0.00/438M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a65bd527da1949898670eba5b9402420",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "tokenizer_config.json:   0%|          | 0.00/1.46k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b1c86250a6ad49a8b819d82560f517ea",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "vocab.txt:   0%|          | 0.00/232k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "240a15ecb70a42c0a0a29a761e98be61",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "tokenizer.json:   0%|          | 0.00/712k [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "146d9ddac41e46c083d9588d5d43d005",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "special_tokens_map.json:   0%|          | 0.00/280 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4917d2a8edae4fa9957e4cae6f10fd0c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "1_Pooling/config.json:   0%|          | 0.00/190 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epoch 1/3: 100%|██████████| 2727/2727 [19:34<00:00,  2.32it/s]\n",
      "Few-Shot Learning Update: 100%|██████████| 622/622 [04:02<00:00,  2.56it/s]\n",
      "Epoch 2/3: 100%|██████████| 2727/2727 [19:37<00:00,  2.32it/s]\n",
      "Few-Shot Learning Update: 100%|██████████| 622/622 [04:14<00:00,  2.45it/s]\n",
      "Epoch 3/3: 100%|██████████| 2727/2727 [1:45:46<00:00,  2.33s/it]      \n",
      "Few-Shot Learning Update: 100%|██████████| 622/622 [01:33<00:00,  6.69it/s]\n"
     ]
    },
    {
     "ename": "TypeError",
     "evalue": "expected Tensor as element 0 in argument 0, but got numpy.ndarray",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[1;32mIn[29], line 77\u001b[0m\n\u001b[0;32m     74\u001b[0m         optimizer\u001b[38;5;241m.\u001b[39mstep()\n\u001b[0;32m     76\u001b[0m \u001b[38;5;66;03m# Visualize embeddings\u001b[39;00m\n\u001b[1;32m---> 77\u001b[0m all_embeddings \u001b[38;5;241m=\u001b[39m \u001b[43mtorch\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcat\u001b[49m\u001b[43m(\u001b[49m\u001b[43mall_embeddings\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     78\u001b[0m tsne_embeddings \u001b[38;5;241m=\u001b[39m TSNE(n_components\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m2\u001b[39m)\u001b[38;5;241m.\u001b[39mfit_transform(all_embeddings)\n\u001b[0;32m     80\u001b[0m plt\u001b[38;5;241m.\u001b[39mscatter(tsne_embeddings[:, \u001b[38;5;241m0\u001b[39m], tsne_embeddings[:, \u001b[38;5;241m1\u001b[39m])\n",
      "\u001b[1;31mTypeError\u001b[0m: expected Tensor as element 0 in argument 0, but got numpy.ndarray"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "class CustomDataset:\n",
    "    def __init__(self, entities, concepts):\n",
    "        self.examples = [InputExample(texts=[str(entity), str(concept)]) for entity, concept in zip(entities, concepts)]\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.examples)\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        example = self.examples[idx]\n",
    "        entity_text = str(example.texts[0])\n",
    "        concept_text = str(example.texts[1])\n",
    "        return entity_text, concept_text\n",
    "\n",
    "def custom_loss(entity_embeddings, concept_embeddings, margin=0.5):\n",
    "    entity_embeddings = entity_embeddings.requires_grad_()\n",
    "    concept_embeddings = concept_embeddings.requires_grad_()\n",
    "\n",
    "    distance = torch.nn.functional.pairwise_distance(entity_embeddings, concept_embeddings)\n",
    "    loss = torch.nn.functional.relu(margin - distance).mean()\n",
    "    return loss\n",
    "\n",
    "few_shot_data = pd.read_excel(r\"C:\\Users\\annem\\Downloads\\cleaned_data.xlsx\")  \n",
    "few_shot_dataset = CustomDataset(entities=few_shot_data['entity_source_text'].tolist(), concepts=few_shot_data['concept_name'].tolist())\n",
    "few_shot_dataloader = DataLoader(few_shot_dataset, batch_size=batch_size, shuffle=True)\n",
    "\n",
    "custom_dataset = CustomDataset(entities=df['Entity'].tolist(), concepts=df['Concept(gold)'].tolist())\n",
    "batch_size = 3\n",
    "dataloader = DataLoader(custom_dataset, batch_size=batch_size, shuffle=True)\n",
    "num_epochs = 3\n",
    "learning_rate = 0.01\n",
    "optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)\n",
    "\n",
    "\n",
    "model.train()\n",
    "all_embeddings = []\n",
    "\n",
    "for epoch in range(num_epochs):\n",
    "    for batch in tqdm(dataloader, desc=f\"Epoch {epoch + 1}/{num_epochs}\"):\n",
    "        entity_texts, concept_texts = batch\n",
    "        entity_embeddings = model.encode(entity_texts, convert_to_tensor=True, device=device)\n",
    "        concept_embeddings = model.encode(concept_texts, convert_to_tensor=True, device=device)\n",
    "\n",
    "        all_embeddings.append(entity_embeddings.detach().numpy())\n",
    "\n",
    "        loss = custom_loss(entity_embeddings, concept_embeddings)\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "    for batch in tqdm(few_shot_dataloader, desc=f\"Few-Shot Learning Update\"):\n",
    "        entity_texts, concept_texts = batch\n",
    "        entity_embeddings = model.encode(entity_texts, convert_to_tensor=True, device=device)\n",
    "        concept_embeddings = model.encode(concept_texts, convert_to_tensor=True, device=device)\n",
    "\n",
    "        loss = custom_loss(entity_embeddings, concept_embeddings)\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "78877fe8",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "d96b69a9",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "torch.save(model.state_dict(), \"trained_model.pth\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f2210020",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "loaded_model = SentenceTransformer(pretrained_model)\n",
    "loaded_model.load_state_dict(torch.load(\"trained_model.pth\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "9c7a215c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([[0.7091]])\n",
      "tensor([[0.7893]])\n",
      "tensor([[0.4589]])\n"
     ]
    }
   ],
   "source": [
    "pretrained_model = \"FremyCompany/BioLORD-2023\"\n",
    "model = SentenceTransformer(pretrained_model)\n",
    "entity=model.encode(\"lobes of the lung\")\n",
    "concept1=model.encode(\"fusion of lobes of lung\")\n",
    "Gold_concept=model.encode(\"history of total lobectomy of lung\")\n",
    "concpt2=model.encode(\"Accessory lobe of lung\")\n",
    "\n",
    "print(util.cos_sim(entity, concept1))\n",
    "print(util.cos_sim(entity, concpt2))\n",
    "print(util.cos_sim(entity, Gold_concept))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "edfcf11b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e46db3e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "tensor([[0.7091]])\n",
    "tensor([[0.7893]])\n",
    "tensor([[0.4589]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9fbca335",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9117ebc0",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a43019a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3d1f4ae4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}