0.97598_in_private-score.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
    "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.feature_selection import VarianceThreshold\n",
    "from sklearn.discriminant_analysis import QuadraticDiscriminantAnalysis\n",
    "from sklearn.mixture import GaussianMixture\n",
    "from sklearn.datasets import make_classification\n",
    "from tqdm import tqdm_notebook\n",
    "import pickle\n",
    "import warnings\n",
    "import multiprocessing\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "add_sample = False\n",
    "sample_size = 128\n",
    "max_iter = 100\n",
    "loop1 = 5\n",
    "loop2 = 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calibration(pred,true_rate):\n",
    "    curr_rate = pred.mean()\n",
    "    change_weight = np.log(curr_rate/(1-curr_rate)) - np.log(true_rate/(1-true_rate))\n",
    "    score_adj = np.exp(np.log(pred/(1-pred)) - change_weight)/(np.exp(np.log(pred/(1-pred)) - change_weight)+1)\n",
    "    return score_adj "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data(data):\n",
    "    return pd.read_csv(data)\n",
    "\n",
    "with multiprocessing.Pool() as pool:\n",
    "    trn, tst, sub = pool.map(load_data, ['../input/instant-gratification/train.csv', '../input/instant-gratification/test.csv','../input/instant-gratification/sample_submission.csv'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('../input/instant/comp_result140.pkl','rb') as f:\n",
    "    out_info = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "637baca9ec7d4ff7bfe209a8984c18a7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=512), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 0.01149102745780295\n",
      "1 0.015852118748460153\n",
      "2 0.011031092821216282\n",
      "3 0.015841145436670145\n",
      "4 0.019547742468635\n",
      "5 0.022362875416310324\n",
      "6 0.009754642404355307\n",
      "7 0.01796650418697654\n",
      "8 0.007850481016376643\n",
      "9 0.016686753902662832\n",
      "10 0.005500709769002521\n",
      "11 0.008594827586206999\n",
      "12 0.010553703553134341\n",
      "13 0.00783829787234036\n",
      "14 0.01163640690874479\n",
      "15 0.014536849222447579\n",
      "16 0.009990114729053645\n",
      "17 0.015290168306815066\n",
      "18 0.0\n",
      "19 0.0057543555266391655\n",
      "20 0.019591662082034333\n",
      "21 0.015048063330506123\n",
      "22 0.007643710080533328\n",
      "23 0.009178218984962294\n",
      "24 0.0038659371621400096\n",
      "25 0.00865023122733466\n",
      "26 0.008541568968982172\n",
      "27 0.013655462184874012\n",
      "28 0.009543361260611394\n",
      "29 0.008798387096774274\n",
      "30 0.0109861879287938\n",
      "31 0.012639365020537974\n",
      "32 0.01544640098856953\n",
      "33 0.013561997703788742\n",
      "34 0.010070302109063234\n",
      "35 0.014920145699075404\n",
      "36 0.005706848217861382\n",
      "37 0.010610021146616577\n",
      "38 0.0033948744618763627\n",
      "39 0.008016562373524994\n",
      "40 0.01602344336696382\n",
      "41 0.011221258433154246\n",
      "42 0.007068412131703172\n",
      "43 0.02242009740788453\n",
      "44 0.00966947960618847\n",
      "45 0.014014058453570177\n",
      "46 0.01958431334901367\n",
      "47 0.009684494435305835\n",
      "48 0.008407432740538123\n",
      "49 0.006822344322344298\n",
      "50 0.008908803373747887\n",
      "51 0.009319905446665944\n",
      "52 0.018842337024155276\n",
      "53 0.008277981560424252\n",
      "54 0.00888051251027655\n",
      "55 0.003283316000594172\n",
      "56 0.011962546816479458\n",
      "57 0.009557025092852611\n",
      "58 0.008757513148008989\n",
      "59 0.01281548960067902\n",
      "60 0.005163409329644542\n",
      "61 0.018308406395048937\n",
      "62 0.015218220774265379\n",
      "63 0.00816678354371192\n",
      "64 0.0\n",
      "65 0.014119433198380715\n",
      "66 0.007432700706741624\n",
      "67 0.022357723577235755\n",
      "68 0.013340973657024913\n",
      "69 0.008827024274316697\n",
      "70 0.009478956417731998\n",
      "71 0.0032556418793933117\n",
      "72 0.016641153523282992\n",
      "73 0.008528941148838065\n",
      "74 0.027190384350461505\n",
      "75 0.012707190784031375\n",
      "76 0.013582516339869288\n",
      "77 0.012523540489641993\n",
      "78 0.008779388948562006\n",
      "79 0.031669085631349736\n",
      "80 0.016050883163182594\n",
      "81 0.018269899200555995\n",
      "82 0.011682600968315193\n",
      "83 0.010582360284937975\n",
      "84 0.010207474878912715\n",
      "85 0.007792887029288709\n",
      "86 0.011615223012281906\n",
      "87 0.011749287225236715\n",
      "88 0.0038946347769877043\n",
      "89 0.018648399655264658\n",
      "90 0.0\n",
      "91 0.006875981161695521\n",
      "92 0.015472931873479312\n",
      "93 0.012945009535918595\n",
      "94 0.008263465830267158\n",
      "95 0.009228710159362441\n",
      "96 0.014581988252523437\n",
      "97 0.018095128676470562\n",
      "98 0.0028290282902831976\n",
      "99 0.02026704816404401\n",
      "100 0.021645770283902577\n",
      "101 0.012950652772787175\n",
      "102 0.018754120961457788\n",
      "103 0.011334486237209895\n",
      "104 0.018377361027088868\n",
      "105 0.018473368342085572\n",
      "106 0.010436018804219738\n",
      "107 0.016506211673498083\n",
      "108 0.00024239159709127378\n",
      "109 0.0034705652683180155\n",
      "110 0.021271126534284424\n",
      "111 0.013321175682799224\n",
      "112 0.009149543632302137\n",
      "113 0.006619177725735059\n",
      "114 0.01905945236309059\n",
      "115 0.007771315852124094\n",
      "116 0.009967910459329987\n",
      "117 0.013973947700981815\n",
      "118 0.014854372167422025\n",
      "119 0.011895161290322753\n",
      "120 0.024348544643747583\n",
      "121 0.010094431781178859\n",
      "122 0.015519075847393426\n",
      "123 0.01007265267087265\n",
      "124 0.008789196612497219\n",
      "125 0.01512495218666332\n",
      "126 0.01102970883723442\n",
      "127 0.01652038462690586\n",
      "128 0.0039187227866472885\n",
      "129 0.013851720945888935\n",
      "130 0.018307064200783985\n",
      "131 0.023094562131469654\n",
      "132 0.01126407080697378\n",
      "133 0.012273537273537305\n",
      "134 0.01349043715846987\n",
      "135 0.008142855280142802\n",
      "136 0.0080099413103365\n",
      "137 0.012332856329532738\n",
      "138 0.006577322134387331\n",
      "139 0.017264559805131197\n",
      "140 0.012143121987393313\n",
      "141 0.01155303030303012\n",
      "142 0.013600998238402773\n",
      "143 0.0\n",
      "144 0.022000697512206346\n",
      "145 0.0\n",
      "146 0.016163252085189095\n",
      "147 0.018131070836294727\n",
      "148 0.013818227498272773\n",
      "149 0.01012223071046603\n",
      "150 0.019754674189622845\n",
      "151 0.007338855082262841\n",
      "152 0.028982293877292453\n",
      "153 0.013078694512824507\n",
      "154 0.011552609977806938\n",
      "155 0.008649648538660681\n",
      "156 0.0076616193671229516\n",
      "157 0.004390495867768629\n",
      "158 0.010624923341101544\n",
      "159 0.022376422891002723\n",
      "160 0.00979450739389287\n",
      "161 0.016915535444947394\n",
      "162 0.015371428571428702\n",
      "163 0.01013063183151175\n",
      "164 0.00781886792452835\n",
      "165 0.014911080711354496\n",
      "166 0.030178494755566665\n",
      "167 0.009829933452220496\n",
      "168 0.010311245439697081\n",
      "169 0.021039568032049205\n",
      "170 0.01375047515163541\n",
      "171 0.009368878903099342\n",
      "172 0.029643196309862896\n",
      "173 0.009382070135746656\n",
      "174 0.020848423770982394\n",
      "175 0.021229404309252242\n",
      "176 0.012978623443739679\n",
      "177 0.012210544064748197\n",
      "178 0.008459032982322978\n",
      "179 0.01792940935402898\n",
      "180 0.009416800428036387\n",
      "181 0.012202334630350142\n",
      "182 0.0146520146520146\n",
      "183 0.007328428959250566\n",
      "184 0.011757575757575744\n",
      "185 0.01621721235948137\n",
      "186 0.012990056818181772\n",
      "187 0.010355029585798814\n",
      "188 0.009978130125751772\n",
      "189 0.011139945566175036\n",
      "190 0.01872698422202923\n",
      "191 0.011216750347185167\n",
      "192 0.010558663619888109\n",
      "193 0.01729417443509207\n",
      "194 0.013468586811975047\n",
      "195 0.014933333333333354\n",
      "196 0.012951065658552086\n",
      "197 0.017764743692752716\n",
      "198 0.011307467226529533\n",
      "199 0.005983318290747164\n",
      "200 0.011178014512891687\n",
      "201 0.017442859597868954\n",
      "202 0.004409836065573769\n",
      "203 0.005002670185027869\n",
      "204 0.010260650444916974\n",
      "205 0.017744673091220076\n",
      "206 0.015181461653386519\n",
      "207 0.00839113517171386\n",
      "208 0.01655982905982889\n",
      "209 0.017065073041168666\n",
      "210 0.006528735632183702\n",
      "211 0.012465863453815396\n",
      "212 0.013993298368298435\n",
      "213 0.01631111803176566\n",
      "214 0.010377096266098396\n",
      "215 0.008363914917875248\n",
      "216 0.01865949331762229\n",
      "217 0.02096654275092924\n",
      "218 0.01705288175876407\n",
      "219 0.01431990118882187\n",
      "220 0.00620229007633577\n",
      "221 0.002720315223156744\n",
      "222 0.010408333333333353\n",
      "223 0.01496397408778427\n",
      "224 0.016459579690136672\n",
      "225 0.01457449816228451\n",
      "226 0.014667137900012839\n",
      "227 0.0181217715301889\n",
      "228 0.018572310100386735\n",
      "229 0.01575099541606706\n",
      "230 0.011812130177514701\n",
      "231 0.02004170528266913\n",
      "232 0.013666395148882238\n",
      "233 0.012701637408622957\n",
      "234 0.02189239943333432\n",
      "235 0.012825207904173297\n",
      "236 0.014758622969243351\n",
      "237 0.026365269711277395\n",
      "238 0.004695105316811943\n",
      "239 0.015018269647612437\n",
      "240 0.014630663988954407\n",
      "241 0.008495890074837198\n",
      "242 0.011315477804255014\n",
      "243 0.01090181413604474\n",
      "244 0.0027944787873654953\n",
      "245 0.020323129251700767\n",
      "246 0.016342830296318778\n",
      "247 0.01939008933815589\n",
      "248 0.005093215452623423\n",
      "249 0.02081314208179874\n",
      "250 0.008505705386946727\n",
      "251 0.026615969581749055\n",
      "252 0.018922125772809495\n",
      "253 0.00898097886049698\n",
      "254 0.013612167300380285\n",
      "255 0.014725568942436484\n",
      "256 0.003927729772191579\n",
      "257 0.007706226885287548\n",
      "258 0.015280594996619379\n",
      "259 0.005255060728745042\n",
      "260 0.008300462555691168\n",
      "261 0.012245783730158721\n",
      "262 0.02216135458167323\n",
      "263 0.016267690454682238\n",
      "264 0.016407323385628225\n",
      "265 0.010520833333333202\n",
      "266 0.012568140520896498\n",
      "267 0.009028535786683278\n",
      "268 0.024212924606462294\n",
      "269 0.013157078135666911\n",
      "270 0.01006877659264438\n",
      "271 0.007791915686652717\n",
      "272 0.01569993665731606\n",
      "273 0.02009336950350138\n",
      "274 0.005706243211406337\n",
      "275 0.0141239343427918\n",
      "276 0.009262900006049124\n",
      "277 0.006163028890301514\n",
      "278 0.011986940298507442\n",
      "279 0.012381521646315341\n",
      "280 0.015030044275774923\n",
      "281 0.011730926989335422\n",
      "282 0.014328808446455565\n",
      "283 0.011907766071541093\n",
      "284 0.013495720868992755\n",
      "285 0.009299609786664331\n",
      "286 0.005583842237228631\n",
      "287 0.013236287697315885\n",
      "288 0.011023622047244164\n",
      "289 0.017258142926592113\n",
      "290 0.011159239409734845\n",
      "291 0.013204043738394788\n",
      "292 0.010772743351448621\n",
      "293 0.0220598050994526\n",
      "294 0.01562827460255667\n",
      "295 0.012449168342704486\n",
      "296 0.02182722416651317\n",
      "297 0.010949416342412421\n",
      "298 0.0069082377476537005\n",
      "299 0.008545963424906988\n",
      "300 0.016399032798065538\n",
      "301 0.011702849389416459\n",
      "302 0.0126751501286817\n",
      "303 0.011491253044826122\n",
      "304 0.009885799247097316\n",
      "305 0.01807331628303488\n",
      "306 0.01806161395856043\n",
      "307 0.011598793363499338\n",
      "308 0.012043052603215587\n",
      "309 0.017373642684165413\n",
      "310 0.009678841884370848\n",
      "311 0.016616396293578806\n",
      "312 0.01059965051043854\n",
      "313 0.019256729929325922\n",
      "314 0.009074071415239304\n",
      "315 0.008809323132780045\n",
      "316 0.018000597399738916\n",
      "317 0.015983375598760063\n",
      "318 0.015518245044272239\n",
      "319 0.011372739757381511\n",
      "320 0.020414582272414217\n",
      "321 0.023664343786294895\n",
      "322 0.009886121887123012\n",
      "323 0.020187385298947258\n",
      "324 0.016271146616541388\n",
      "325 0.012050861361771914\n",
      "326 0.00392364046210214\n",
      "327 0.011865704286964096\n",
      "328 0.022589631782945818\n",
      "329 0.017907172995780374\n",
      "330 0.006384960623253444\n",
      "331 0.005012531328320691\n",
      "332 0.0017838373969296661\n",
      "333 0.016902254688043516\n",
      "334 0.008005071352753856\n",
      "335 0.01159126259552301\n",
      "336 0.014649539649539634\n",
      "337 0.002483863336825376\n",
      "338 0.01239316239316235\n",
      "339 0.008550688976378007\n",
      "340 0.0019304686557388706\n",
      "341 0.009044820831892952\n",
      "342 0.010439449640901821\n",
      "343 0.01806568473235126\n",
      "344 0.013116016047134749\n",
      "345 0.012987114043824688\n",
      "346 0.018220086074199893\n",
      "347 0.0038492645957522997\n",
      "348 0.015247805765938316\n",
      "349 0.019300144300144262\n",
      "350 0.009102693360219072\n",
      "351 0.011877176875209616\n",
      "352 0.011797169299455845\n",
      "353 0.006729990747166403\n",
      "354 0.01639150943396228\n",
      "355 0.0124983735488442\n",
      "356 0.03304942965779456\n",
      "357 0.0\n",
      "358 0.021082621082621045\n",
      "359 0.009749957330602466\n",
      "360 0.01410273996090683\n",
      "361 0.00888065422949147\n",
      "362 0.01017414046054732\n",
      "363 0.010462393912789025\n",
      "364 0.014954780361757014\n",
      "365 0.00833053785083293\n",
      "366 0.01400907930202866\n",
      "367 0.008939636566177023\n",
      "368 0.018790513833992106\n",
      "369 0.02297340340818599\n",
      "370 0.02238805970149249\n",
      "371 0.017738461538461592\n",
      "372 0.024803767660910414\n",
      "373 0.01868762816131231\n",
      "374 0.024531249999999893\n",
      "375 0.016601762656282104\n",
      "376 0.011608964404250743\n",
      "377 0.0194852996409417\n",
      "378 0.010471367841960455\n",
      "379 0.008965517241379395\n",
      "380 0.016646681766369187\n",
      "381 0.014309028384915767\n",
      "382 0.018660427025445903\n",
      "383 0.015211864406779796\n",
      "384 0.006842518402227604\n",
      "385 0.006821779044001164\n",
      "386 0.010162106161253015\n",
      "387 0.01711304889442533\n",
      "388 0.01626403899131157\n",
      "389 0.010957001812054279\n",
      "390 0.015799023901889586\n",
      "391 0.01383019592777568\n",
      "392 0.007632504823516073\n",
      "393 0.012460913070668989\n",
      "394 0.02086855854620795\n",
      "395 0.005092893134254539\n",
      "396 0.0112650152307604\n",
      "397 0.015158529741862936\n",
      "398 0.007559041030534397\n",
      "399 0.009390024038461453\n",
      "400 0.01296113770768803\n",
      "401 0.008007473642065777\n",
      "402 0.015056265629341481\n",
      "403 0.01254270814601699\n",
      "404 0.011035852834438864\n",
      "405 0.01247322458395117\n",
      "406 0.016267598732564448\n",
      "407 0.014887648876488857\n",
      "408 0.00588400112076215\n",
      "409 0.01626051388408789\n",
      "410 0.019585959383753404\n",
      "411 0.011673151750972721\n",
      "412 0.012261070675464114\n",
      "413 0.008582117471006456\n",
      "414 0.014842475210242179\n",
      "415 0.012759709272446962\n",
      "416 0.01068116454683865\n",
      "417 0.013337574529153406\n",
      "418 0.016929940032177737\n",
      "419 0.006298873663820981\n",
      "420 0.013365255376344143\n",
      "421 0.015325670498084087\n",
      "422 0.008788515406162434\n",
      "423 0.007081572140558823\n",
      "424 0.02330892196235912\n",
      "425 0.010032910502604131\n",
      "426 0.016665047128558985\n",
      "427 0.019779990252732582\n",
      "428 0.01346177984184771\n",
      "429 0.014021887824897328\n",
      "430 0.01582403550907474\n",
      "431 0.01364125248057213\n",
      "432 0.006658333333333322\n",
      "433 0.013498229740361856\n",
      "434 0.017443399512364843\n",
      "435 0.015548406862745279\n",
      "436 0.01784776902887153\n",
      "437 0.018764249886001005\n",
      "438 0.013374353715335685\n",
      "439 0.013225371120107798\n",
      "440 0.013260610693756392\n",
      "441 0.01382026789635471\n",
      "442 0.009908600497526399\n",
      "443 0.012434389426421166\n",
      "444 0.0\n",
      "445 0.007450090215936234\n",
      "446 0.011233753510547895\n",
      "447 0.012806396422339938\n",
      "448 0.01374946374946362\n",
      "449 0.0175497679988027\n",
      "450 0.007945949915546868\n",
      "451 0.009132687410356222\n",
      "452 0.018193086351742105\n",
      "453 0.0074341085271316665\n",
      "454 0.015477654776547833\n",
      "455 0.01119486588566676\n",
      "456 0.01648327137546468\n",
      "457 0.015279678872979652\n",
      "458 0.031833501853724244\n",
      "459 0.011314285714285721\n",
      "460 0.00997084064620446\n",
      "461 0.00722763237979307\n",
      "462 0.016427478691629638\n",
      "463 0.01614497003697546\n",
      "464 0.01245928596274859\n",
      "465 0.012316613423367428\n",
      "466 0.011035789780596383\n",
      "467 0.008630307280377392\n",
      "468 0.012277862500389092\n",
      "469 0.018970927254098324\n",
      "470 0.007067839909168261\n",
      "471 0.011186770428015635\n",
      "472 0.0\n",
      "473 0.012432626661875612\n",
      "474 0.011943337891670014\n",
      "475 0.014535607569721165\n",
      "476 0.008621463852716671\n",
      "477 0.007892219387754973\n",
      "478 0.011606364428944915\n",
      "479 0.00933362274799221\n",
      "480 0.021582408301158318\n",
      "481 0.021392346703550058\n",
      "482 0.016319346662912015\n",
      "483 0.00989062499999993\n",
      "484 0.014206612877709301\n",
      "485 0.006766092938031321\n",
      "486 0.02227859778597796\n",
      "487 0.017937725631768986\n",
      "488 0.011920698667686636\n",
      "489 0.025330525330525266\n",
      "490 0.016048360669874695\n",
      "491 0.011718496922578647\n",
      "492 0.015744355796978637\n",
      "493 0.011273863136666229\n",
      "494 0.009448095540951162\n",
      "495 0.018344226579520728\n",
      "496 0.004728169035918661\n",
      "497 0.022271859819569806\n",
      "498 0.009638144268891646\n",
      "499 0.023929571828731344\n",
      "500 0.021531939308938042\n",
      "501 0.020267635590216493\n",
      "502 0.024756659089584465\n",
      "503 0.009081123564789406\n",
      "504 0.01590266875981161\n",
      "505 0.026291972124174223\n",
      "506 0.012221565148827107\n",
      "507 0.010817164623259212\n",
      "508 0.014650401491463683\n",
      "509 0.009851745576279214\n",
      "510 0.017472388194821664\n",
      "511 0.023128733002922952\n",
      "\n",
      "AUC: 0.97647\n"
     ]
    }
   ],
   "source": [
    "cols = np.array([c for c in trn.columns if c not in ['id', 'target', 'wheezy-copper-turtle-magic','pred']])\n",
    "oof = np.zeros(len(trn))\n",
    "preds = np.zeros(len(tst))\n",
    "\n",
    "oof_loop = np.zeros(len(trn))\n",
    "preds_loop = np.zeros(len(tst))\n",
    "\n",
    "pre_init = {}\n",
    "for i in tqdm_notebook(range(512)):\n",
    "    pre_init[i] = [None,None]\n",
    "    trn2 = trn[trn['wheezy-copper-turtle-magic']==i]\n",
    "    tst2 = tst[tst['wheezy-copper-turtle-magic']==i]\n",
    "    idx1 = trn2.index; idx2 = tst2.index\n",
    "    trn2.reset_index(drop=True,inplace=True)\n",
    "\n",
    "    data = pd.concat([pd.DataFrame(trn2[cols]), pd.DataFrame(tst2[cols])])\n",
    "    usecols = [var for var in cols if data[var].std()>2]\n",
    "    data2 = VarianceThreshold(threshold=2).fit_transform(data[cols])\n",
    "    \n",
    "    trn3 = data2[:trn2.shape[0]]; tst3 = data2[trn2.shape[0]:]\n",
    "    skf = StratifiedKFold(n_splits=11, random_state=42)\n",
    "    oof_tmp = np.zeros(trn2.shape[0])\n",
    "    for trn_index, tst_index in skf.split(trn2, trn2['target']):\n",
    "        clf = QuadraticDiscriminantAnalysis(0.6)\n",
    "        clf.fit(trn3[trn_index,:],trn2.loc[trn_index]['target'])\n",
    "        oof[idx1[tst_index]] = clf.predict_proba(trn3[tst_index,:])[:,1]\n",
    "        oof_tmp[tst_index] = clf.predict_proba(trn3[tst_index,:])[:,1]\n",
    "        preds[idx2] += clf.predict_proba(tst3)[:,1] / skf.n_splits\n",
    "    auc_bench = roc_auc_score(trn2['target'], oof_tmp)\n",
    "    auc_basic = auc_bench\n",
    "    \n",
    "    model_cnt = 0\n",
    "    for loop in range(loop1):\n",
    "        components = out_info[i][1].shape[0]\n",
    "        trn2['tgt'] = np.nan\n",
    "        gmm = GaussianMixture(n_components=components,precisions_init=out_info[i][1])\n",
    "        trn2.loc[trn2['target']==1,'tgt'] = gmm.fit_predict(trn2[trn2['target']==1][usecols].values)+components\n",
    "        \n",
    "        precisions_pos = gmm.precisions_\n",
    "        gmm = GaussianMixture(n_components=components,precisions_init=out_info[i][2])\n",
    "        trn2.loc[trn2['target']==0,'tgt'] = gmm.fit_predict(trn2[trn2['target']==0][usecols].values)\n",
    "        \n",
    "        precisions_neg = gmm.precisions_\n",
    "        trn3 = trn2[usecols].values\n",
    "\n",
    "        skf = StratifiedKFold(n_splits=11, random_state=42)\n",
    "        oof_tmp = np.zeros(trn2.shape[0])\n",
    "        preds_tmp = np.zeros(tst3.shape[0])\n",
    "        for trn_index, tst_index in skf.split(trn2, trn2['tgt']):\n",
    "            clf = QuadraticDiscriminantAnalysis(0.6)\n",
    "            try:\n",
    "                clf.fit(trn3[trn_index,:],trn2.loc[trn_index]['tgt'])\n",
    "            except:\n",
    "                break\n",
    "            oof_tmp[tst_index] = clf.predict_proba(trn3[tst_index,:])[:,components:].sum(axis=1)\n",
    "            preds_tmp += clf.predict_proba(tst3)[:,components:].sum(axis=1) / skf.n_splits\n",
    "        auc_tmp = roc_auc_score(trn2['target'],oof_tmp)\n",
    "        \n",
    "        if auc_basic < auc_tmp:\n",
    "            if auc_bench < auc_tmp:\n",
    "                auc_bench = auc_tmp\n",
    "            preds_loop[idx2] += preds_tmp\n",
    "            oof_loop[idx1] += oof_tmp\n",
    "            pre_init[i] = [precisions_pos,precisions_neg]\n",
    "            model_cnt += 1\n",
    "        if model_cnt>0:\n",
    "            preds[idx2] = preds_loop[idx2]/model_cnt\n",
    "            oof[idx1] = oof_loop[idx1]/model_cnt\n",
    "    '''preds[idx2] = preds[idx2]/model_cnt\n",
    "    oof[idx1] = oof[idx1]/model_cnt'''\n",
    "    print(i,auc_bench-auc_basic)\n",
    "auc = roc_auc_score(trn['target'], oof)\n",
    "print(f'AUC: {auc:.5}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "trn['pred'] = oof\n",
    "tst['pred'] = preds\n",
    "tst['target'] = np.nan\n",
    "tst.loc[tst['pred']>=0.99,'target']= 1\n",
    "tst.loc[tst['pred']<=0.01,'target']= 0\n",
    "trn_new = trn.copy()\n",
    "trn_new['target_old'] = trn_new['target']\n",
    "trn_new.loc[trn_new['pred']>=0.995,'target'] = 1\n",
    "trn_new.loc[trn_new['pred']<=0.005,'target'] = 0\n",
    "trn_new = pd.concat([trn_new,tst[tst['target'].notnull()]]).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = np.array([c for c in trn.columns if c not in ['id', 'target', 'wheezy-copper-turtle-magic','target_old','pred']])\n",
    "oof = np.zeros(len(trn_new))\n",
    "preds = np.zeros(len(tst))\n",
    "oof_qda = np.zeros(len(trn_new))\n",
    "preds_qda = np.zeros(len(tst))\n",
    "oof_nusvc = np.zeros(len(trn_new))\n",
    "preds_nusvc = np.zeros(len(tst))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.svm import NuSVC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cb62a4679cb548a9a5264218e22eb326",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=512), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "第0轮QDA的CV为：0.9711531716077617，basic的CV为：0.9662284455544176\n",
      "第0轮NuSVC的CV为：0.9643202895907286，basic的CV为：0.9581608743844092\n",
      "第1轮QDA的CV为：0.9720913402315841，basic的CV为：0.9668791574279378\n",
      "第1轮NuSVC的CV为：0.9609047795023404，basic的CV为：0.9515582655826558\n",
      "第2轮QDA的CV为：0.9779214840943236，basic的CV为：0.972663139329806\n",
      "第2轮NuSVC的CV为：0.9631589261218891，basic的CV为：0.9631589261218891\n",
      "第3轮QDA的CV为：0.9841311602531261，basic的CV为：0.9803093454449784\n",
      "第3轮NuSVC的CV为：0.9759197972776493，basic的CV为：0.97013168644502\n",
      "第4轮QDA的CV为：0.9819998497483284，basic的CV为：0.9704755465404552\n",
      "第4轮NuSVC的CV为：0.9718728870858689，basic的CV为：0.9681466456314327\n",
      "第5轮QDA的CV为：0.9799286745851632，basic的CV为：0.9649267588198887\n",
      "第5轮NuSVC的CV为：0.9670930472457191，basic的CV为：0.9554658256948333\n",
      "第6轮QDA的CV为：0.9890521291669774，basic的CV为：0.9852337981952419\n",
      "第6轮NuSVC的CV为：0.976672384219554，basic的CV为：0.9703259005145797\n",
      "第7轮QDA的CV为：0.9672306586676667，basic的CV为：0.9638795150606174\n",
      "第7轮NuSVC的CV为：0.9505686789151356，basic的CV为：0.9505686789151356\n",
      "第8轮QDA的CV为：0.9836968495263274，basic的CV为：0.9759858999779687\n",
      "第8轮NuSVC的CV为：0.9735183961224939，basic的CV为：0.9665565102445473\n",
      "第9轮QDA的CV为：0.9760387970615242，basic的CV为：0.9686710858585859\n",
      "第9轮NuSVC的CV为：0.9671717171717171，basic的CV为：0.9650195133149679\n",
      "第10轮QDA的CV为：0.9836833785004516，basic的CV为：0.9828687572590011\n",
      "第10轮NuSVC的CV为：0.9723996644728352，basic的CV为：0.9721415666537617\n",
      "第11轮QDA的CV为：0.985344827586207，basic的CV为：0.9767413793103448\n",
      "第11轮NuSVC的CV为：0.9743103448275863，basic的CV为：0.9723362068965518\n",
      "第12轮QDA的CV为：0.9804699569070655，basic的CV为：0.9771688755183349\n",
      "第12轮NuSVC的CV为：0.9768273843401902，basic的CV为：0.974160500853728\n",
      "第13轮QDA的CV为：0.9858808510638298，basic的CV为：0.9810723404255319\n",
      "第13轮NuSVC的CV为：0.9786893617021277，basic的CV为：0.9746127659574468\n",
      "第14轮QDA的CV为：0.9891540867673059，basic的CV为：0.9869271045831634\n",
      "第14轮NuSVC的CV为：0.9791751665986672，basic的CV为：0.9772286821705426\n",
      "第15轮QDA的CV为：0.9877394636015325，basic的CV为：0.978453910299752\n",
      "第15轮NuSVC的CV为：0.9723536924348283，basic的CV为：0.9684471489745322\n",
      "第16轮QDA的CV为：0.9785594464248271，basic的CV为：0.9755863762993141\n",
      "第16轮NuSVC的CV为：0.9721115537848607，basic的CV为：0.9685169098043914\n",
      "第17轮QDA的CV为：0.9879441429841503，basic的CV为：0.9812149360801986\n",
      "第17轮NuSVC的CV为：0.974416750973359，basic的CV为：0.9623685582022747\n",
      "第18轮QDA的CV为：0.9790588410768887，basic的CV为：0.9786561264822133\n",
      "第18轮NuSVC的CV为：0.9720933701245431，basic的CV为：0.9675591020956075\n",
      "第19轮QDA的CV为：0.9907395358924311，basic的CV为：0.9798886720161932\n",
      "第19轮NuSVC的CV为：0.9798019229379021，basic的CV为：0.9652714523241523\n",
      "第20轮QDA的CV为：0.9731111314872546，basic的CV为：0.9723546671556941\n",
      "第20轮NuSVC的CV为：0.9693288098294517，basic的CV为：0.9635598141695703\n",
      "第21轮QDA的CV为：0.9761450381679388，basic的CV为：0.9738903025162567\n",
      "第21轮NuSVC的CV为：0.9589977381962115，basic的CV为：0.9589977381962115\n",
      "第22轮QDA的CV为：0.9866634268258817，basic的CV为：0.9816162177173008\n",
      "第22轮NuSVC的CV为：0.9740072202166065，basic的CV为：0.9730213829491807\n",
      "第23轮QDA的CV为：0.9832148731203008，basic的CV为：0.977561090225564\n",
      "第23轮NuSVC的CV为：0.9749324483082706，basic的CV为：0.9728471569548871\n",
      "第24轮QDA的CV为：0.9701536546211054，basic的CV为：0.9665416243488517\n",
      "第24轮NuSVC的CV为：0.9574583101267897，basic的CV为：0.9553123873800085\n",
      "第25轮QDA的CV为：0.9760122434041987，basic的CV为：0.971237981169112\n",
      "第25轮NuSVC的CV为：0.9638187443856672，basic的CV为：0.9625045746415144\n",
      "第26轮QDA的CV为：0.9741240704763181，basic的CV为：0.9693675889328063\n",
      "第26轮NuSVC的CV为：0.9638909358879882，basic的CV为：0.9617136732096202\n",
      "第27轮QDA的CV为：0.9611581474695142，basic的CV为：0.9548556264611108\n",
      "第27轮NuSVC的CV为：0.9563878182852088，basic的CV为：0.9515701017249005\n",
      "第28轮QDA的CV为：0.9713699162181657，basic的CV为：0.9713699162181657\n",
      "第28轮NuSVC的CV为：0.9624507573655885，basic的CV为：0.9617294568051934\n",
      "第29轮QDA的CV为：0.9795161290322582，basic的CV为：0.9744032258064517\n",
      "第29轮NuSVC的CV为：0.9710967741935483，basic的CV为：0.9692580645161291\n",
      "第30轮QDA的CV为：0.9831958170882824，basic的CV为：0.9771058119503128\n",
      "第30轮NuSVC的CV为：0.9744612687762566，basic的CV为：0.971287816967389\n",
      "第31轮QDA的CV为：0.9759180332931839，basic的CV为：0.9678495321041416\n",
      "第31轮NuSVC的CV为：0.9672936162327433，basic的CV为：0.9613947311529079\n",
      "第32轮QDA的CV为：0.9857034977516906，basic的CV为：0.978237737273882\n",
      "第32轮NuSVC的CV为：0.9673737685785879，basic的CV为：0.9673737685785879\n",
      "第33轮QDA的CV为：0.9778559127439723，basic的CV为：0.9701636050516648\n",
      "第33轮NuSVC的CV为：0.9714982778415614，basic的CV为：0.9657290470723308\n",
      "第34轮QDA的CV为：0.9862026630029671，basic的CV为：0.9812040515061606\n",
      "第34轮NuSVC的CV为：0.9714260658588988，basic的CV为：0.9679182683172803\n",
      "第35轮QDA的CV为：0.9817097848759379，basic的CV为：0.9711403754553096\n",
      "第35轮NuSVC的CV为：0.9642445752000249，basic的CV为：0.9597615267270633\n",
      "第36轮QDA的CV为：0.9739187024429315，basic的CV为：0.9690628754505407\n",
      "第36轮NuSVC的CV为：0.9706314243759177，basic的CV为：0.9662678547590443\n",
      "第37轮QDA的CV为：0.9811516094924811，basic的CV为：0.9801016212406015\n",
      "第37轮NuSVC的CV为：0.9699688674812029，basic的CV为：0.968544407894737\n",
      "第38轮QDA的CV为：0.985207015110803，basic的CV为：0.9814509837912802\n",
      "第38轮NuSVC的CV为：0.9780416630550981，basic的CV为：0.9725520788188726\n",
      "第39轮QDA的CV为：0.9792659008125525，basic的CV为：0.9702686715855673\n",
      "第39轮NuSVC的CV为：0.9700818778991938，basic的CV为：0.967855919803244\n",
      "第40轮QDA的CV为：0.9806869755444543，basic的CV为：0.971136578835668\n",
      "第40轮NuSVC的CV为：0.9641819309759362，basic的CV为：0.9641819309759362\n",
      "第41轮QDA的CV为：0.9790100509431364，basic的CV为：0.9679608977006746\n",
      "第41轮NuSVC的CV为：0.9722979485061269，basic的CV为：0.9613933636238469\n",
      "第42轮QDA的CV为：0.9699772801038624，basic的CV为：0.9679577337805185\n",
      "第42轮NuSVC的CV为：0.9695986151682354，basic的CV为：0.9686609686609686\n",
      "第43轮QDA的CV为：0.9712661544610739，basic的CV为：0.9634498379002532\n",
      "第43轮NuSVC的CV为：0.9660848840135603，basic的CV为：0.9536350313096771\n",
      "第44轮QDA的CV为：0.9755066487661423，basic的CV为：0.9748993095512082\n",
      "第44轮NuSVC的CV为：0.9711513872906278，basic的CV为：0.9687779695691089\n",
      "第45轮QDA的CV为：0.9730817610062894，basic的CV为：0.9687606363300036\n",
      "第45轮NuSVC的CV为：0.9616130225675176，basic的CV为：0.9576248612652609\n",
      "第46轮QDA的CV为：0.982527720443527，basic的CV为：0.9688955023280372\n",
      "第46轮NuSVC的CV为：0.9666554664874638，basic的CV为：0.9587993407894526\n",
      "第47轮QDA的CV为：0.9783265904948205，basic的CV为：0.971754175373079\n",
      "第47轮NuSVC的CV为：0.9683901658293691，basic的CV为：0.9660486966315447\n",
      "第48轮QDA的CV为：0.9741222070223439，basic的CV为：0.9686787505699955\n",
      "第48轮NuSVC的CV为：0.9655865253077975，basic的CV为：0.9599008207934335\n",
      "第49轮QDA的CV为：0.9709706959706959，basic的CV为：0.968452380952381\n",
      "第49轮NuSVC的CV为：0.9621031746031746，basic的CV为：0.9539224664224664\n",
      "第50轮QDA的CV为：0.979124934106484，basic的CV为：0.9743806009488666\n",
      "第50轮NuSVC的CV为：0.9730231945176595，basic的CV为：0.9668423827095414\n",
      "第51轮QDA的CV为：0.9809108145375751，basic的CV为：0.973517679503595\n",
      "第51轮NuSVC的CV为：0.9652935093075938，basic的CV为：0.9612922288978627\n",
      "第52轮QDA的CV为：0.9710664028845848，basic的CV为：0.969829924375379\n",
      "第52轮NuSVC的CV为：0.9663438527074892，basic的CV为：0.9663438527074892\n",
      "第53轮QDA的CV为：0.9820643732857474，basic的CV为：0.974141634446978\n",
      "第53轮NuSVC的CV为：0.9629225736095965，basic的CV为：0.9572717358976901\n",
      "第54轮QDA的CV为：0.9721097615834458，basic的CV为：0.9678517691221865\n",
      "第54轮NuSVC的CV为：0.9563187367180108，basic的CV为：0.9531155474894132\n",
      "第55轮QDA的CV为：0.9911825880255535，basic的CV为：0.9854702124498589\n",
      "第55轮NuSVC的CV为：0.9844451047392662，basic的CV为：0.9815034913088694\n",
      "第56轮QDA的CV为：0.9815955056179775，basic的CV为：0.9798202247191011\n",
      "第56轮NuSVC的CV为：0.96825468164794，basic的CV为：0.9610936329588015\n",
      "第57轮QDA的CV为：0.9818673188996587，basic的CV为：0.9786212519249933\n",
      "第57轮NuSVC的CV为：0.9751638131473261，basic的CV为：0.9703475556360781\n",
      "第58轮QDA的CV为：0.9872746055597296，basic的CV为：0.9816553969446532\n",
      "第58轮NuSVC的CV为：0.9760205359378914，basic的CV为：0.9717630853994489\n",
      "第59轮QDA的CV为：0.9784585496457374，basic的CV为：0.9721895712933033\n",
      "第59轮NuSVC的CV为：0.9620269696672871，basic的CV为：0.9586965749175564\n",
      "第60轮QDA的CV为：0.9799529523161369，basic的CV为：0.9739906200529657\n",
      "第60轮NuSVC的CV为：0.9647290319717123，basic的CV为：0.9600686481928067\n",
      "第61轮QDA的CV为：0.9686354876360577，basic的CV为：0.9671697293775955\n",
      "第61轮NuSVC的CV为：0.9664639939198175，basic的CV为：0.9611166906435764\n",
      "第62轮QDA的CV为：0.9756831322825864，basic的CV为：0.9691673270523319\n",
      "第62轮NuSVC的CV为：0.9656599454081184，basic的CV为：0.9628716503771535\n",
      "第63轮QDA的CV为：0.981299672744273，basic的CV为：0.9774719495091164\n",
      "第63轮NuSVC的CV为：0.968910705937354，basic的CV为：0.9616935483870968\n",
      "第64轮QDA的CV为：0.986449640584979，basic的CV为：0.9835577955878707\n",
      "第64轮NuSVC的CV为：0.979757085020243，basic的CV为：0.9795918367346939\n",
      "第65轮QDA的CV为：0.9804872758820128，basic的CV为：0.9719057258530942\n",
      "第65轮NuSVC的CV为：0.9703730480046269，basic的CV为：0.9653412377096586\n",
      "第66轮QDA的CV为：0.9781783530532835，basic的CV为：0.9753831493686969\n",
      "第66轮NuSVC的CV为：0.9651711268164853，basic的CV为：0.9639402842849202\n",
      "第67轮QDA的CV为：0.9730595901464631，basic的CV为：0.9664976705946834\n",
      "第67轮NuSVC的CV为：0.9677080478669955，basic的CV为：0.9580859291738985\n",
      "第68轮QDA的CV为：0.9795971074380165，basic的CV为：0.9701543130165289\n",
      "第68轮NuSVC的CV为：0.9642949380165289，basic的CV为：0.9636412060950412\n",
      "第69轮QDA的CV为：0.9726984665874497，basic的CV为：0.9706756068579186\n",
      "第69轮NuSVC的CV为：0.96818593334465，basic的CV为：0.96211028121994\n",
      "第70轮QDA的CV为：0.9893249607535322，basic的CV为：0.9851237198175974\n",
      "第70轮NuSVC的CV为：0.9713687672871346，basic的CV为：0.965522912461688\n",
      "第71轮QDA的CV为：0.9694265630780613，basic的CV为：0.968583055863855\n",
      "第71轮NuSVC的CV为：0.9680355160932298，basic的CV为：0.9642471328153902\n",
      "第72轮QDA的CV为：0.9755649926752589，basic的CV为：0.9681168008164206\n",
      "第72轮NuSVC的CV为：0.9538459006139614，basic的CV为：0.9458792158411929\n",
      "第73轮QDA的CV为：0.9779730185992573，basic的CV为：0.9742223396603104\n",
      "第73轮NuSVC的CV为：0.9798593678894908，basic的CV为：0.973304854596967\n",
      "第74轮QDA的CV为：0.9808599669375947，basic的CV为：0.9680396748863479\n",
      "第74轮NuSVC的CV为：0.9696583551453368，basic的CV为：0.9646473343435735\n",
      "第75轮QDA的CV为：0.9798247495327675，basic的CV为：0.9744171083672907\n",
      "第75轮NuSVC的CV为：0.9649116088115445，basic的CV为：0.9649116088115445\n",
      "第76轮QDA的CV为：0.9791258169934641，basic的CV为：0.9692674291938999\n",
      "第76轮NuSVC的CV为：0.9693763616557735，basic的CV为：0.9642020697167756\n",
      "第77轮QDA的CV为：0.9849026993094789，basic的CV为：0.9787507846829882\n",
      "第77轮NuSVC的CV为：0.9716101694915255，basic的CV为：0.9675533584431889\n",
      "第78轮QDA的CV为：0.9816430958348251，basic的CV为：0.9724758324382384\n",
      "第78轮NuSVC的CV为：0.9701336675020886，basic的CV为：0.9647108843537415\n",
      "第79轮QDA的CV为：0.9763497822931786，basic的CV为：0.9611030478955007\n",
      "第79轮NuSVC的CV为：0.9599854862119013，basic的CV为：0.942989840348331\n",
      "第80轮QDA的CV为：0.9869875222816399，basic的CV为：0.9773780586614812\n",
      "第80轮NuSVC的CV为：0.973731972127694，basic的CV为：0.9598768432993032\n",
      "第81轮QDA的CV为：0.9625622755184801，basic的CV为：0.956957478855289\n",
      "第81轮NuSVC的CV为：0.9582029892248871，basic的CV为：0.9545533541883906\n",
      "第82轮QDA的CV为：0.9710884353741497，basic的CV为：0.9665226450940737\n",
      "第82轮NuSVC的CV为：0.9602714959857817，basic的CV为：0.9569467426610284\n",
      "第83轮QDA的CV为：0.9864287287948887，basic的CV为：0.98197106557979\n",
      "第83轮NuSVC的CV为：0.9725930821766909，basic的CV为：0.9714915179554968\n",
      "第84轮QDA的CV为：0.9779801922937902，basic的CV为：0.9732668257066436\n",
      "第84轮NuSVC的CV为：0.9644473360803875，basic的CV为：0.9631677871755946\n",
      "第85轮QDA的CV为：0.9714522315202231，basic的CV为：0.9644002789400278\n",
      "第85轮NuSVC的CV为：0.9670327754532775，basic的CV为：0.9619769874476987\n",
      "第86轮QDA的CV为：0.9883780435251024，basic的CV为：0.9798332794656325\n",
      "第86轮NuSVC的CV为：0.9762578108166344，basic的CV为：0.9682584572290455\n",
      "第87轮QDA的CV为：0.9568426377264785，basic的CV为：0.9490021153315551\n",
      "第87轮NuSVC的CV为：0.9620466599221313，basic的CV为：0.9587050492044513\n",
      "第88轮QDA的CV为：0.9796218487394958，basic的CV为：0.9763897866839044\n",
      "第88轮NuSVC的CV为：0.9668552036199095，basic的CV为：0.9646250808015514\n",
      "第89轮QDA的CV为：0.9658607982406611，basic的CV为：0.9611281167345241\n",
      "第89轮NuSVC的CV为：0.9564623020000595，basic的CV为：0.9477993402478528\n",
      "第90轮QDA的CV为：0.9795460584934269，basic的CV为：0.9733403154455785\n",
      "第90轮NuSVC的CV为：0.9706282337861285，basic的CV为：0.9595219858377754\n",
      "第91轮QDA的CV为：0.9803767660910517，basic的CV为：0.9752119309262167\n",
      "第91轮NuSVC的CV为：0.9704395604395605，basic的CV为：0.9630455259026688\n",
      "第92轮QDA的CV为：0.97143400243309，basic的CV为：0.9636709245742092\n",
      "第92轮NuSVC的CV为：0.9509732360097323，basic的CV为：0.9506234793187348\n",
      "第93轮QDA的CV为：0.9705499046408136，basic的CV为：0.9628019707565163\n",
      "第93轮NuSVC的CV为：0.9533057851239669，basic的CV为：0.9486729179910999\n",
      "第94轮QDA的CV为：0.9637945593823511，basic的CV为：0.9616080583871162\n",
      "第94轮NuSVC的CV为：0.9556517280897522，basic的CV为：0.9541136377344834\n",
      "第95轮QDA的CV为：0.9682208665338645，basic的CV为：0.9592567231075698\n",
      "第95轮NuSVC的CV为：0.957015687250996，basic的CV为：0.9536852589641434\n",
      "第96轮QDA的CV为：0.9621120262034865，basic的CV为：0.9553406925657055\n",
      "第96轮NuSVC的CV为：0.9543013715887438，basic的CV为：0.9463017495236445\n",
      "第97轮QDA的CV为：0.9770630908613445，basic的CV为：0.969070049894958\n",
      "第97轮NuSVC的CV为：0.9638753939075629，basic的CV为：0.9575974921218487\n",
      "第98轮QDA的CV为：0.9796740467404674，basic的CV为：0.9768603936039361\n",
      "第98轮NuSVC的CV为：0.9747232472324724，basic的CV为：0.9709563345633456\n",
      "第99轮QDA的CV为：0.9690033381020505，basic的CV为：0.960340168494675\n",
      "第99轮NuSVC的CV为：0.9578286440947384，basic的CV为：0.9481004609759975\n",
      "第100轮QDA的CV为：0.9796944804726906，basic的CV为：0.9633808906182447\n",
      "第100轮NuSVC的CV为：0.9655281740884853，basic的CV为：0.9602248162559446\n",
      "第101轮QDA的CV为：0.9871691220505451，basic的CV为：0.9802221822972812\n",
      "第101轮NuSVC的CV为：0.977497305066475，basic的CV为：0.977497305066475\n",
      "第102轮QDA的CV为：0.9691002817239106，basic的CV为：0.9631211412815441\n",
      "第102轮NuSVC的CV为：0.9615776538991788，basic的CV为：0.9561005214889408\n",
      "第103轮QDA的CV为：0.9664288802421098，basic的CV为：0.9585386943363597\n",
      "第103轮NuSVC的CV为：0.9572560887735985，basic的CV为：0.9507421818705865\n",
      "第104轮QDA的CV为：0.9708408953418028，basic的CV为：0.9579485111245547\n",
      "第104轮NuSVC的CV为：0.9495866102036701，basic的CV为：0.947193654634671\n",
      "第105轮QDA的CV为：0.9712193673418356，basic的CV为：0.9624281070267567\n",
      "第105轮NuSVC的CV为：0.9606776694173543，basic的CV为：0.949456114028507\n",
      "第106轮QDA的CV为：0.985573738711814，basic的CV为：0.9812765544983119\n",
      "第106轮NuSVC的CV为：0.9776094085717518，basic的CV为：0.9710424710424709\n",
      "第107轮QDA的CV为：0.966463317764433，basic的CV为：0.9583611349410607\n",
      "第107轮NuSVC的CV为：0.9526101737997649，basic的CV为：0.9509500206526229\n",
      "第108轮QDA的CV为：0.9774306490708322，basic的CV为：0.9767842714785887\n",
      "第108轮NuSVC的CV为：0.9643953676272555，basic的CV为：0.9608402908699165\n",
      "第109轮QDA的CV为：0.9773545616242245，basic的CV为：0.974910705247784\n",
      "第109轮NuSVC的CV为：0.9718739606380056，basic的CV为：0.9643182508351047\n",
      "第110轮QDA的CV为：0.9881354026090867，basic的CV为：0.9773150825782404\n",
      "第110轮NuSVC的CV为：0.9805925069082964，basic的CV为：0.9713707345286293\n",
      "第111轮QDA的CV为：0.9850118964510108，basic的CV为：0.9812933650572027\n",
      "第111轮NuSVC的CV为：0.9757084443431309，basic的CV为：0.972303352377153\n",
      "第112轮QDA的CV为：0.9825219307977929，basic的CV为：0.9752806994186305\n",
      "第112轮NuSVC的CV为：0.970946315773902，basic的CV为：0.969748073196349\n",
      "第113轮QDA的CV为：0.9841269841269841，basic的CV为：0.9789715066354411\n",
      "第113轮NuSVC的CV为：0.9728564923237053，basic的CV为：0.9702706219099662\n",
      "第114轮QDA的CV为：0.9686327831957989，basic的CV为：0.9606307826956738\n",
      "第114轮NuSVC的CV为：0.9636002750687671，basic的CV为：0.9584114778694673\n",
      "第115轮QDA的CV为：0.9742628303234364，basic的CV为：0.9676138489269802\n",
      "第115轮NuSVC的CV为：0.9645529367751592，basic的CV为：0.9645529367751592\n",
      "第116轮QDA的CV为：0.9863503185700776，basic的CV为：0.9806532624366349\n",
      "第116轮NuSVC的CV为：0.9769637403692623，basic的CV为：0.9749794595935324\n",
      "第117轮QDA的CV为：0.9826155989760539，basic的CV为：0.975227633582839\n",
      "第117轮NuSVC的CV为：0.9736722724474256，basic的CV为：0.9736722724474256\n",
      "第118轮QDA的CV为：0.9808301119701414，basic的CV为：0.9755065315915756\n",
      "第118轮NuSVC的CV为：0.9688749666755532，basic的CV为：0.9652925886430287\n",
      "第119轮QDA的CV为：0.978，basic的CV为：0.9727741935483871\n",
      "第119轮NuSVC的CV为：0.9698064516129032，basic的CV为：0.9684032258064517\n",
      "第120轮QDA的CV为：0.9738491786462267，basic的CV为：0.9650230093034522\n",
      "第120轮NuSVC的CV为：0.9590676601746713，basic的CV为：0.9526991408910227\n",
      "第121轮QDA的CV为：0.9752058426757222，basic的CV为：0.9700888496069219\n",
      "第121轮NuSVC的CV为：0.9641500364391931，basic的CV为：0.9590640554495976\n",
      "第122轮QDA的CV为：0.9780589755281959，basic的CV为：0.9696306429548563\n",
      "第122轮NuSVC的CV为：0.9611187110503117，basic的CV为：0.9554643562851497\n",
      "第123轮QDA的CV为：0.9856931311604374，basic的CV为：0.9773558129849296\n",
      "第123轮NuSVC的CV为：0.9671289083589135，basic的CV为：0.9659565934998225\n",
      "第124轮QDA的CV为：0.9896849011978331，basic的CV为：0.9835660334172579\n",
      "第124轮NuSVC的CV为：0.9812314030670634，basic的CV为：0.9739986266880293\n",
      "第125轮QDA的CV为：0.9631040418207318，basic的CV为：0.9529198010965193\n",
      "第125轮NuSVC的CV为：0.9495888053040928，basic的CV为：0.9371573377534107\n",
      "第126轮QDA的CV为：0.9749643116869124，basic的CV为：0.9731866296765157\n",
      "第126轮NuSVC的CV为：0.9666011258652732，basic的CV为：0.9643049532685107\n",
      "第127轮QDA的CV为：0.9870592816703113，basic的CV为：0.9753916664170386\n",
      "第127轮NuSVC的CV为：0.9819518916813947，basic的CV为：0.9717820447533175\n",
      "第128轮QDA的CV为：0.9777503628447025，basic的CV为：0.9721044992743105\n",
      "第128轮NuSVC的CV为：0.9673004354136429，basic的CV为：0.9632075471698113\n",
      "第129轮QDA的CV为：0.9623495750231423，basic的CV为：0.9527055457376084\n",
      "第129轮NuSVC的CV为：0.9495918539089455，basic的CV为：0.9468147774131112\n",
      "第130轮QDA的CV为：0.9699816197400739，basic的CV为：0.9579043250540834\n",
      "第130轮NuSVC的CV为：0.9637111859334081，basic的CV为：0.9517640169814082\n",
      "第131轮QDA的CV为：0.972654665709918，basic的CV为：0.9566187252363273\n",
      "第131轮NuSVC的CV为：0.9582878357720026，basic的CV为：0.9523601534957726\n",
      "第132轮QDA的CV为：0.9761650182660606，basic的CV为：0.9697793222251924\n",
      "第132轮NuSVC的CV为：0.9688882948706524，basic的CV为：0.9641509994356826\n",
      "第133轮QDA的CV为：0.976076626076626，basic的CV为：0.9754826254826254\n",
      "第133轮NuSVC的CV为：0.9666765666765667，basic的CV为：0.9597564597564598\n",
      "第134轮QDA的CV为：0.9815806631892698，basic的CV为：0.9722351589667164\n",
      "第134轮NuSVC的CV为：0.9689052409339294，basic的CV为：0.9624316939890711\n",
      "第135轮QDA的CV为：0.9842162909261601，basic的CV为：0.979333185558004\n",
      "第135轮NuSVC的CV为：0.968471698851264，basic的CV为：0.9622260180198975\n",
      "第136轮QDA的CV为：0.974428075218589，basic的CV为：0.974428075218589\n",
      "第136轮NuSVC的CV为：0.974278356689424，basic的CV为：0.9707899149598755\n",
      "第137轮QDA的CV为：0.9783347951191739，basic的CV为：0.9759805505547092\n",
      "第137轮NuSVC的CV为：0.9670867377556202，basic的CV为：0.9632553201310992\n",
      "第138轮QDA的CV为：0.9884665266798419，basic的CV为：0.9837110918972332\n",
      "第138轮NuSVC的CV为：0.9832015810276681，basic的CV为：0.9775506422924901\n",
      "第139轮QDA的CV为：0.9794296289266394，basic的CV为：0.9694410173673721\n",
      "第139轮NuSVC的CV为：0.9675271266331341，basic的CV为：0.9595156749232863\n",
      "第140轮QDA的CV为：0.9854699666295884，basic的CV为：0.9744005685329378\n",
      "第140轮NuSVC的CV为：0.9726548016314424，basic的CV为：0.960496230379434\n",
      "第141轮QDA的CV为：0.9777753496503496，basic的CV为：0.9709644522144523\n",
      "第141轮NuSVC的CV为：0.9600233100233101，basic的CV为：0.956191724941725\n",
      "第142轮QDA的CV为：0.9772019964768056，basic的CV为：0.9704785672342925\n",
      "第142轮NuSVC的CV为：0.9681738109219025，basic的CV为：0.9681738109219025\n",
      "第143轮QDA的CV为：0.9719192937438905，basic的CV为：0.96690188172043\n",
      "第143轮NuSVC的CV为：0.9615255376344086，basic的CV为：0.9613498900293256\n",
      "第144轮QDA的CV为：0.9745916647291328，basic的CV为：0.9633370146477563\n",
      "第144轮NuSVC的CV为：0.9613607300627761，basic的CV为：0.9488200418507324\n",
      "第145轮QDA的CV为：0.9781992882562276，basic的CV为：0.9648683274021354\n",
      "第145轮NuSVC的CV为：0.9650391459074733，basic的CV为：0.9508612099644128\n",
      "第146轮QDA的CV为：0.9831484330715038，basic的CV为：0.9763381650336059\n",
      "第146轮NuSVC的CV为：0.9681755607741518，basic的CV为：0.9642076281480281\n",
      "第147轮QDA的CV为：0.9837938284766643，basic的CV为：0.9716447524283347\n",
      "第147轮NuSVC的CV为：0.961087420042644，basic的CV为：0.9515221511490168\n",
      "第148轮QDA的CV为：0.9829784561271278，basic的CV为：0.9731800766283525\n",
      "第148轮NuSVC的CV为：0.968736260285158，basic的CV为：0.960296149739338\n",
      "第149轮QDA的CV为：0.9679808682366227，basic的CV为：0.9588384761019032\n",
      "第149轮NuSVC的CV为：0.9631979273923008，basic的CV为：0.9602584116650613\n",
      "第150轮QDA的CV为：0.9745339639340636，basic的CV为：0.9669764354463989\n",
      "第150轮NuSVC的CV为：0.960587454364421，basic的CV为：0.9562382453811262\n",
      "第151轮QDA的CV为：0.9713277552400271，basic的CV为：0.9679118773946361\n",
      "第151轮NuSVC的CV为：0.9626155059725041，basic的CV为：0.9600518368266846\n",
      "第152轮QDA的CV为：0.9654033765164234，basic的CV为：0.9631623958696272\n",
      "第152轮NuSVC的CV为：0.9515219663615343，basic的CV为：0.9465886414747712\n",
      "第153轮QDA的CV为：0.9827999682363218，basic的CV为：0.97327086476614\n",
      "第153轮NuSVC的CV为：0.9718732629238466，basic的CV为：0.9645596760104821\n",
      "第154轮QDA的CV为：0.984860000608032，basic的CV为：0.9781108442525766\n",
      "第154轮NuSVC的CV为：0.9760283342960507，basic的CV为：0.974341045207187\n",
      "第155轮QDA的CV为：0.9887828338882723，basic的CV为：0.9796374398816131\n",
      "第155轮NuSVC的CV为：0.9813244543100258，basic的CV为：0.978645948945616\n",
      "第156轮QDA的CV为：0.9854021216792093，basic的CV为：0.9773627104300783\n",
      "第156轮NuSVC的CV为：0.9737510200380812，basic的CV为：0.9712802611297489\n",
      "第157轮QDA的CV为：0.9889290633608816，basic的CV为：0.9842630853994491\n",
      "第157轮NuSVC的CV为：0.9820247933884297，basic的CV为：0.9770144628099173\n",
      "第158轮QDA的CV为：0.9776615969581749，basic的CV为：0.9721881516006379\n",
      "第158轮NuSVC的CV为：0.9728780816877223，basic的CV为：0.9685621856985158\n",
      "第159轮QDA的CV为：0.9860595665055356，basic的CV为：0.9718774364571964\n",
      "第159轮NuSVC的CV为：0.9679245283018869，basic的CV为：0.9626851707469205\n",
      "第160轮QDA的CV为：0.9814352474233404，basic的CV为：0.9764419691441009\n",
      "第160轮NuSVC的CV为：0.9719608219704244，basic的CV为：0.9690480763075348\n",
      "第161轮QDA的CV为：0.9778808446455505，basic的CV为：0.970105580693816\n",
      "第161轮NuSVC的CV为：0.969291101055807，basic的CV为：0.9621417797888385\n",
      "第162轮QDA的CV为：0.9634204081632654，basic的CV为：0.9606040816326531\n",
      "第162轮NuSVC的CV为：0.948604081632653，basic的CV为：0.9439510204081633\n",
      "第163轮QDA的CV为：0.9895361237003466，basic的CV为：0.9786556918155159\n",
      "第163轮NuSVC的CV为：0.9793555051986137，basic的CV为：0.9768978272460677\n",
      "第164轮QDA的CV为：0.9881358490566039，basic的CV为：0.9816301886792453\n",
      "第164轮NuSVC的CV为：0.9756830188679246，basic的CV为：0.970754716981132\n",
      "第165轮QDA的CV为：0.9902112783097735，basic的CV为：0.9795105639154887\n",
      "第165轮NuSVC的CV为：0.9847393220854234，basic的CV为：0.9783249734002127\n",
      "第166轮QDA的CV为：0.9696374900325093，basic的CV为：0.9525854137275348\n",
      "第166轮NuSVC的CV为：0.9595779917806538，basic的CV为：0.9484067349567564\n",
      "第167轮QDA的CV为：0.9780348832151711，basic的CV为：0.9711336320009278\n",
      "第167轮NuSVC的CV为：0.9699737578472736，basic的CV为：0.9651965261769098\n",
      "第168轮QDA的CV为：0.9845064187693671，basic的CV为：0.9812703210147913\n",
      "第168轮NuSVC的CV为：0.9796370075254538，basic的CV为：0.9748897131779396\n",
      "第169轮QDA的CV为：0.9885113362556971，basic的CV为：0.9716811333352686\n",
      "第169轮NuSVC的CV为：0.9717101634394867，basic的CV为：0.9605480883676373\n",
      "第170轮QDA的CV为：0.9838035268646603，basic的CV为：0.9737551027153882\n",
      "第170轮NuSVC的CV为：0.9688300527211727，basic的CV为：0.9652850083461418\n",
      "第171轮QDA的CV为：0.9750619310980528，basic的CV为：0.9723326420094481\n",
      "第171轮NuSVC的CV为：0.9724190574951032，basic的CV为：0.9666580251181012\n",
      "第172轮QDA的CV为：0.9839159468789098，basic的CV为：0.9596014410829226\n",
      "第172轮NuSVC的CV为：0.9595712929046263，basic的CV为：0.9489817452780417\n",
      "第173轮QDA的CV为：0.9791784502262444，basic的CV为：0.9735576923076924\n",
      "第173轮NuSVC的CV为：0.9705882352941176，basic的CV为：0.9659502262443439\n",
      "第174轮QDA的CV为：0.9843873019872138，basic的CV为：0.9638459358170881\n",
      "第174轮NuSVC的CV为：0.9614209680975025，basic的CV为：0.9550908575567663\n",
      "第175轮QDA的CV为：0.9584222077962224，basic的CV为：0.9512890661226003\n",
      "第175轮NuSVC的CV为：0.943831339454079，basic的CV为：0.9385143281090607\n",
      "第176轮QDA的CV为：0.9828297509983557，basic的CV为：0.9756871035940803\n",
      "第176轮NuSVC的CV为：0.9728535353535355，basic的CV为：0.9659678177120038\n",
      "第177轮QDA的CV为：0.9720520458633094，basic的CV为：0.9720520458633094\n",
      "第177轮NuSVC的CV为：0.9657149280575539，basic的CV为：0.9584644784172662\n",
      "第178轮QDA的CV为：0.9808367327578028，basic的CV为：0.9766151219049427\n",
      "第178轮NuSVC的CV为：0.9670492995604465，basic的CV为：0.9650807956234386\n",
      "第179轮QDA的CV为：0.9806191622696925，basic的CV为：0.9680515000939159\n",
      "第179轮NuSVC的CV为：0.9602820893738367，basic的CV为：0.9543824599149634\n",
      "第180轮QDA的CV为：0.9809323167469234，basic的CV为：0.97956126270733\n",
      "第180轮NuSVC的CV为：0.9693151417870519，basic的CV为：0.964780631353665\n",
      "第181轮QDA的CV为：0.9729260700389104，basic的CV为：0.9719688715953306\n",
      "第181轮NuSVC的CV为：0.9625214007782101，basic的CV为：0.956124513618677\n",
      "第182轮QDA的CV为：0.9740206246230344，basic的CV为：0.970166379804934\n",
      "第182轮NuSVC的CV为：0.9637377348220721，basic的CV为：0.9612368889477323\n",
      "第183轮QDA的CV为：0.9854486376999141，basic的CV为：0.9825158595079685\n",
      "第183轮NuSVC的CV为：0.9754968843627361，basic的CV为：0.968766263907839\n",
      "第184轮QDA的CV为：0.9833257575757576，basic的CV为：0.9767424242424243\n",
      "第184轮NuSVC的CV为：0.9748030303030304，basic的CV为：0.9693181818181819\n",
      "第185轮QDA的CV为：0.9848680303867968，basic的CV为：0.975469418678461\n",
      "第185轮NuSVC的CV为：0.9692446300960336，basic的CV为：0.9692446300960336\n",
      "第186轮QDA的CV为：0.9825568181818181，basic的CV为：0.9754971590909092\n",
      "第186轮NuSVC的CV为：0.9744602272727272，basic的CV为：0.9728977272727272\n",
      "第187轮QDA的CV为：0.9710292743693553，basic的CV为：0.965875116786048\n",
      "第187轮NuSVC的CV为：0.96013702896294，basic的CV为：0.9562285892245408\n",
      "第188轮QDA的CV为：0.979823522264747，basic的CV为：0.9689493347913251\n",
      "第188轮NuSVC的CV为：0.9662687564546504，basic的CV为：0.962183342445781\n",
      "第189轮QDA的CV为：0.9801727957465662，basic的CV为：0.9740331666561175\n",
      "第189轮NuSVC的CV为：0.9747927083992659，basic的CV为：0.9728859421482372\n",
      "第190轮QDA的CV为：0.9746265081883776，basic的CV为：0.9625909403910062\n",
      "第190轮NuSVC的CV为：0.9626508188377593，basic的CV为：0.9615430675728271\n",
      "第191轮QDA的CV为：0.9827780914737436，basic的CV为：0.9759488455140629\n",
      "第191轮NuSVC的CV为：0.9751858012727578，basic的CV为：0.97314084270606\n",
      "第192轮QDA的CV为：0.9849263257426523，basic的CV为：0.9817350878575368\n",
      "第192轮NuSVC的CV为：0.9864943660862028，basic的CV为：0.9834528405956977\n",
      "第193轮QDA的CV为：0.9684511418831266，basic的CV为：0.9603584034754277\n",
      "第193轮NuSVC的CV为：0.9599360424774489，basic的CV为：0.9503047032913989\n",
      "第194轮QDA的CV为：0.9744088741850734，basic的CV为：0.9672326554441957\n",
      "第194轮NuSVC的CV为：0.9638999708086018，basic的CV为：0.9608673088774285\n",
      "第195轮QDA的CV为：0.965225，basic的CV为：0.9567000000000001\n",
      "第195轮NuSVC的CV为：0.95，basic的CV为：0.94695\n",
      "第196轮QDA的CV为：0.9818772489547368，basic的CV为：0.9789344871290591\n",
      "第196轮NuSVC的CV为：0.9715484463091649，basic的CV为：0.9708054717888204\n",
      "第197轮QDA的CV为：0.9643098184155552，basic的CV为：0.9537522095452354\n",
      "第197轮NuSVC的CV为：0.9523541700144627，basic的CV为：0.9493090149445604\n",
      "第198轮QDA的CV为：0.9678960506521631，basic的CV为：0.9635317650559587\n",
      "第198轮NuSVC的CV为：0.9579937511365327，basic的CV为：0.9559273280322692\n",
      "第199轮QDA的CV为：0.9825091033581277，basic的CV为：0.9782764308611621\n",
      "第199轮NuSVC的CV为：0.9696243503158944，basic的CV为：0.9696243503158944\n",
      "第200轮QDA的CV为：0.9802454840203799，basic的CV为：0.9731665894704338\n",
      "第200轮NuSVC的CV为：0.9667593021460553，basic的CV为：0.9557125212289639\n",
      "第201轮QDA的CV为：0.9778546766860909，basic的CV为：0.9765970410411037\n",
      "第201轮NuSVC的CV为：0.9603493258760486，basic的CV为：0.9586620631473699\n",
      "第202轮QDA的CV为：0.9805327868852459，basic的CV为：0.9778196721311476\n",
      "第202轮NuSVC的CV为：0.9761967213114754，basic的CV为：0.974344262295082\n",
      "第203轮QDA的CV为：0.9871752583796689，basic的CV为：0.9858794332924953\n",
      "第203轮NuSVC的CV为：0.9762824741620331，basic的CV为：0.9739971099173813\n",
      "第204轮QDA的CV为：0.9825104338924325，basic的CV为：0.9807386408378612\n",
      "第204轮NuSVC的CV为：0.9725175210646508，basic的CV为：0.9703126230411843\n",
      "第205轮QDA的CV为：0.9779138436504929，basic的CV为：0.9674984920534725\n",
      "第205轮NuSVC的CV为：0.9671057245858407，basic的CV为：0.961340459257389\n",
      "第206轮QDA的CV为：0.9611242529880478，basic的CV为：0.9536696962151395\n",
      "第206轮NuSVC的CV为：0.9526425547808766，basic的CV为：0.946308515936255\n",
      "第207轮QDA的CV为：0.9687277956352562，basic的CV为：0.9611487058027407\n",
      "第207轮NuSVC的CV为：0.9567247504652344，basic的CV为：0.9567247504652344\n",
      "第208轮QDA的CV为：0.9713064713064713，basic的CV为：0.9609737484737485\n",
      "第208轮NuSVC的CV为：0.9601953601953602，basic的CV为：0.9558531746031746\n",
      "第209轮QDA的CV为：0.9827523240371847，basic的CV为：0.9697211155378486\n",
      "第209轮NuSVC的CV为：0.9718791500664011，basic的CV为：0.9700365205843294\n",
      "第210轮QDA的CV为：0.9826666666666667，basic的CV为：0.9802758620689657\n",
      "第210轮NuSVC的CV为：0.970727969348659，basic的CV为：0.9627126436781609\n",
      "第211轮QDA的CV为：0.9649236947791165，basic的CV为：0.9634538152610441\n",
      "第211轮NuSVC的CV为：0.9539518072289156，basic的CV为：0.9539518072289156\n",
      "第212轮QDA的CV为：0.9842001748251747，basic的CV为：0.9771416083916085\n",
      "第212轮NuSVC的CV为：0.9689102564102565，basic的CV为：0.9682109557109557\n",
      "第213轮QDA的CV为：0.9728199937714109，basic的CV为：0.9641388975397073\n",
      "第213轮NuSVC的CV为：0.9608377452507008，basic的CV为：0.9562285892245408\n",
      "第214轮QDA的CV为：0.9863182927205905，basic的CV为：0.9795700917261188\n",
      "第214轮NuSVC的CV为：0.976991259767133，basic的CV为：0.976991259767133\n",
      "第215轮QDA的CV为：0.9857751095592688，basic的CV为：0.9827199201909715\n",
      "第215轮NuSVC的CV为：0.9746677592902697，basic的CV为：0.9730733601738697\n",
      "第216轮QDA的CV为：0.9803172191385779，basic的CV为：0.974820924679835\n",
      "第216轮NuSVC的CV为：0.978123352662098，basic的CV为：0.9703246612298055\n",
      "第217轮QDA的CV为：0.981635687732342，basic的CV为：0.9679256505576208\n",
      "第217轮NuSVC的CV为：0.9729516728624535，basic的CV为：0.9677472118959107\n",
      "第218轮QDA的CV为：0.9863636363636363，basic的CV为：0.9767825311942959\n",
      "第218轮NuSVC的CV为：0.9709595959595959，basic的CV为：0.9671717171717172\n",
      "第219轮QDA的CV为：0.9807472595337348，basic的CV为：0.9748185888528641\n",
      "第219轮NuSVC的CV为：0.966790180639185，basic的CV为：0.9591014358499305\n",
      "第220轮QDA的CV为：0.9748046165030897，basic的CV为：0.9732672967405791\n",
      "第220轮NuSVC的CV为：0.9630134496546711，basic的CV为：0.9579546831455228\n",
      "第221轮QDA的CV为：0.9649693112070925，basic的CV为：0.9602182314162309\n",
      "第221轮NuSVC的CV为：0.952761991361673，basic的CV为：0.9518223838751231\n",
      "第222轮QDA的CV为：0.965，basic的CV为：0.9601500000000001\n",
      "第222轮NuSVC的CV为：0.9432166666666667，basic的CV为：0.9411166666666666\n",
      "第223轮QDA的CV为：0.9821853516657854，basic的CV为：0.9778225806451613\n",
      "第223轮NuSVC的CV为：0.9721708090957166，basic的CV为：0.9666181914331042\n",
      "第224轮QDA的CV为：0.9870532290228563，basic的CV为：0.976069949378739\n",
      "第224轮NuSVC的CV为：0.9713299585826046，basic的CV为：0.9652093879429361\n",
      "第225轮QDA的CV为：0.977848459146169，basic的CV为：0.9719677692960135\n",
      "第225轮NuSVC的CV为：0.9696635566864575，basic的CV为：0.9624045801526717\n",
      "第226轮QDA的CV为：0.9794210255751189，basic的CV为：0.9716456753630638\n",
      "第226轮NuSVC的CV为：0.9644807865312942，basic的CV为：0.9638703251510088\n",
      "第227轮QDA的CV为：0.9824921669912778，basic的CV为：0.9725491856493634\n",
      "第227轮NuSVC的CV为：0.9685691704067521，basic的CV为：0.9573983120219041\n",
      "第228轮QDA的CV为：0.9771006727571001，basic的CV为：0.9610041689869513\n",
      "第228轮NuSVC的CV为：0.9652935598934426，basic的CV为：0.9539605376036602\n",
      "第229轮QDA的CV为：0.9757670559105966，basic的CV为：0.967812092214006\n",
      "第229轮NuSVC的CV为：0.9664904473516913，basic的CV为：0.961371164720447\n",
      "第230轮QDA的CV为：0.969896449704142，basic的CV为：0.9619304733727811\n",
      "第230轮NuSVC的CV为：0.9665828402366864，basic的CV为：0.9637278106508876\n",
      "第231轮QDA的CV为：0.9811708371949336，basic的CV为：0.9699799196787149\n",
      "第231轮NuSVC的CV为：0.9673154155081867，basic的CV为：0.95102718566574\n",
      "第232轮QDA的CV为：0.9855793887832986，basic的CV为：0.9707062479259058\n",
      "第232轮NuSVC的CV为：0.9742812320873684，basic的CV为：0.9649817479711588\n",
      "第233轮QDA的CV为：0.9718098313972671，basic的CV为：0.9617040857078362\n",
      "第233轮NuSVC的CV为：0.955357607959577，basic的CV为：0.9495754544980988\n",
      "第234轮QDA的CV为：0.9707195334892761，basic的CV为：0.9631008467037854\n",
      "第234轮NuSVC的CV为：0.9589167462853754，basic的CV为：0.9568741145850492\n",
      "第235轮QDA的CV为：0.9694819562066072，basic的CV为：0.9644083314259556\n",
      "第235轮NuSVC的CV为：0.9613946746013581，basic的CV为：0.9540550850690471\n",
      "第236轮QDA的CV为：0.9798700976746492，basic的CV为：0.9719371312410136\n",
      "第236轮NuSVC的CV为：0.9681689721850364，basic的CV为：0.9622192473598097\n",
      "第237轮QDA的CV为：0.976503851028566，basic的CV为：0.9632028301229821\n",
      "第237轮NuSVC的CV为：0.9669772559506399，basic的CV为：0.9541079959330909\n",
      "第238轮QDA的CV为：0.9920958548106883，basic的CV为：0.9903800594426178\n",
      "第238轮NuSVC的CV为：0.9832153574454032，basic的CV为：0.9818944103837925\n",
      "第239轮QDA的CV为：0.981552300736979，basic的CV为：0.9736019074750728\n",
      "第239轮NuSVC的CV为：0.969994426209203，basic的CV为：0.9678732891558804\n",
      "第240轮QDA的CV为：0.9785364629095017，basic的CV为：0.9682126270867328\n",
      "第240轮NuSVC的CV为：0.9621093259696247，basic的CV为：0.9615209614660475\n",
      "第241轮QDA的CV为：0.9818350509139983，basic的CV为：0.975930867378236\n",
      "第241轮NuSVC的CV为：0.9772113851061219，basic的CV为：0.9742822966507176\n",
      "第242轮QDA的CV为：0.97417987757024，basic的CV为：0.9683794466403163\n",
      "第242轮NuSVC的CV为：0.9655969520982863，basic的CV为：0.9553231260969449\n",
      "第243轮QDA的CV为：0.9749281453781253，basic的CV为：0.9718453503105974\n",
      "第243轮NuSVC的CV为：0.9657415706029607，basic的CV为：0.9590737707451247\n",
      "第244轮QDA的CV为：0.9736048776357016，basic的CV为：0.9736048776357016\n",
      "第244轮NuSVC的CV为：0.9625878567194512，basic的CV为：0.9606232534507579\n",
      "第245轮QDA的CV为：0.9814364207221351，basic的CV为：0.9708137100994244\n",
      "第245轮NuSVC的CV为：0.9740319204604918，basic的CV为：0.9616365777080064\n",
      "第246轮QDA的CV为：0.9851337293197759，basic的CV为：0.9778851639316755\n",
      "第246轮NuSVC的CV为：0.9722977281116817，basic的CV为：0.9691432598409344\n",
      "第247轮QDA的CV为：0.9780905128442919，basic的CV为：0.9663513555287644\n",
      "第247轮NuSVC的CV为：0.9710810225972982，basic的CV为：0.9635537420012984\n",
      "第248轮QDA的CV为：0.9816051636876162，basic的CV为：0.9810846306617976\n",
      "第248轮NuSVC的CV为：0.9782337113203922，basic的CV为：0.977769235697354\n",
      "第249轮QDA的CV为：0.9824013051624991，basic的CV为：0.9695354396846934\n",
      "第249轮NuSVC的CV为：0.9722976028946179，basic的CV为：0.9688650901337468\n",
      "第250轮QDA的CV为：0.9792455201888294，basic的CV为：0.9739242168186707\n",
      "第250轮NuSVC的CV为：0.9648877777622592，basic的CV为：0.961689409069959\n",
      "第251轮QDA的CV为：0.9715062644143864，basic的CV为：0.9471576388456024\n",
      "第251轮NuSVC的CV为：0.9475160506139749，basic的CV为：0.9368727794053482\n",
      "第252轮QDA的CV为：0.9739274339599076，basic的CV为：0.9642321863485918\n",
      "第252轮NuSVC的CV为：0.963373509023918，basic的CV为：0.9577140448385686\n",
      "第253轮QDA的CV为：0.9798240581373112，basic的CV为：0.9763670064874884\n",
      "第253轮NuSVC的CV为：0.9730202862732984，basic的CV为：0.9692028186004091\n",
      "第254轮QDA的CV为：0.984106463878327，basic的CV为：0.9759543726235741\n",
      "第254轮NuSVC的CV为：0.9715285171102661，basic的CV为：0.9656273764258554\n",
      "第255轮QDA的CV为：0.9686152015469285，basic的CV为：0.9684813327383608\n",
      "第255轮NuSVC的CV为：0.9656105905101889，basic的CV为：0.9575487133720066\n",
      "第256轮QDA的CV为：0.9903770620581305，basic的CV为：0.9864656978266562\n",
      "第256轮NuSVC的CV为：0.98702212621105，basic的CV为：0.983683555904687\n",
      "第257轮QDA的CV为：0.9776590847764322，basic的CV为：0.970429040808876\n",
      "第257轮NuSVC的CV为：0.9670163965651338，basic的CV为：0.9595641339026365\n",
      "第258轮QDA的CV为：0.9642476147547141，basic的CV为：0.9555705807227106\n",
      "第258轮NuSVC的CV为：0.9539027871685072，basic的CV为：0.9523476823679663\n",
      "第259轮QDA的CV为：0.9803481781376518，basic的CV为：0.9749878542510121\n",
      "第259轮NuSVC的CV为：0.9725344129554656，basic的CV为：0.9700566801619432\n",
      "第260轮QDA的CV为：0.9798447742557961，basic的CV为：0.9795255256959619\n",
      "第260轮NuSVC的CV为：0.9789437838758194，basic的CV为：0.9789437838758194\n",
      "第261轮QDA的CV为：0.9807245163690477，basic的CV为：0.9651692708333335\n",
      "第261轮NuSVC的CV为：0.968781001984127，basic的CV为：0.958131820436508\n",
      "第262轮QDA的CV为：0.969812416998672，basic的CV为：0.9552788844621514\n",
      "第262轮NuSVC的CV为：0.9618193891102258，basic的CV为：0.9511786188579018\n",
      "第263轮QDA的CV为：0.9668925022583559，basic的CV为：0.9541930141523637\n",
      "第263轮NuSVC的CV为：0.9536284251731407，basic的CV为：0.9462436013249022\n",
      "第264轮QDA的CV为：0.9715148416781163，basic的CV为：0.9597565261846183\n",
      "第264轮NuSVC的CV为：0.9641019906061284，basic的CV为：0.9618733424928907\n",
      "第265轮QDA的CV为：0.9827483974358974，basic的CV为：0.9791826923076923\n",
      "第265轮NuSVC的CV为：0.9758974358974359，basic的CV为：0.9716826923076924\n",
      "第266轮QDA的CV为：0.9704800121138702，basic的CV为：0.9631738340399757\n",
      "第266轮NuSVC的CV为：0.9603952150211992，basic的CV为：0.9603952150211992\n",
      "第267轮QDA的CV为：0.9893497583034462，basic的CV为：0.9849524403555279\n",
      "第267轮NuSVC的CV为：0.9805239357554967，basic的CV为：0.9780913768906908\n",
      "第268轮QDA的CV为：0.9661695664181166，basic的CV为：0.9517950842308754\n",
      "第268轮NuSVC的CV为：0.9436343551505109，basic的CV为：0.9341411212372274\n",
      "第269轮QDA的CV为：0.9799928629057283，basic的CV为：0.9741978526655496\n",
      "第269轮NuSVC的CV为：0.975904549121827，basic的CV为：0.9733445044374108\n",
      "第270轮QDA的CV为：0.9741650638223466，basic的CV为：0.9665442676458589\n",
      "第270轮NuSVC的CV为：0.9662528414058401，basic的CV为：0.9627702978376173\n",
      "第271轮QDA的CV为：0.959634663582032，basic的CV为：0.9578593920699183\n",
      "第271轮NuSVC的CV为：0.9524130839920314，basic的CV为：0.9492641861062915\n",
      "第272轮QDA的CV为：0.9854839682683318，basic的CV为：0.9699122251379966\n",
      "第272轮NuSVC的CV为：0.9623563478418241，basic的CV为：0.9570928724398998\n",
      "第273轮QDA的CV为：0.958685571888209，basic的CV为：0.9541511576556684\n",
      "第273轮NuSVC的CV为：0.9488360355813513，basic的CV为：0.9402561352596052\n",
      "第274轮QDA的CV为：0.9870806370186791，basic的CV为：0.9828200774090903\n",
      "第274轮NuSVC的CV为：0.9852830939158902，basic的CV为：0.9837685682377958\n",
      "第275轮QDA的CV为：0.9813094682670964，basic的CV为：0.9745231935077971\n",
      "第275轮NuSVC的CV为：0.9714269556488668，basic的CV为：0.9687760670710156\n",
      "第276轮QDA的CV为：0.9840224426834432，basic的CV为：0.9803853366402517\n",
      "第276轮NuSVC的CV为：0.9687103018571169，basic的CV为：0.9668501603048818\n",
      "第277轮QDA的CV为：0.990084410538956，basic的CV为：0.985025075934167\n",
      "第277轮NuSVC的CV为：0.9788973652610016，basic的CV为：0.9788973652610016\n",
      "第278轮QDA的CV为：0.9771299751243782，basic的CV为：0.9721703980099503\n",
      "第278轮NuSVC的CV为：0.9678327114427862，basic的CV为：0.9678327114427862\n",
      "第279轮QDA的CV为：0.9865425668175218，basic的CV为：0.9728631201434549\n",
      "第279轮NuSVC的CV为：0.9661173255913245，basic的CV为：0.9627188113739219\n",
      "第280轮QDA的CV为：0.980629348513599，basic的CV为：0.969797596457938\n",
      "第280轮NuSVC的CV为：0.9679316888045542，basic的CV为：0.963353889943074\n",
      "第281轮QDA的CV为：0.9803117309269894，basic的CV为：0.9718547244388097\n",
      "第281轮NuSVC的CV为：0.9726601536281603，basic的CV为：0.9702886121261839\n",
      "第282轮QDA的CV为：0.9761010558069381，basic的CV为：0.9721417797888386\n",
      "第282轮NuSVC的CV为：0.9647058823529413，basic的CV为：0.9647058823529413\n",
      "第283轮QDA的CV为：0.9818781149454294，basic的CV为：0.9755614787710555\n",
      "第283轮NuSVC的CV为：0.9752539271970223，basic的CV为：0.9712557567345909\n",
      "第284轮QDA的CV为：0.9831303489137591，basic的CV为：0.9685813034891376\n",
      "第284轮NuSVC的CV为：0.976152073732719，basic的CV为：0.968614219881501\n",
      "第285轮QDA的CV为：0.9801119371101427，basic的CV为：0.9784385770030477\n",
      "第285轮NuSVC的CV为：0.9734398587256828，basic的CV为：0.9708692927739326\n",
      "第286轮QDA的CV为：0.9879401313664262，basic的CV为：0.9830023535202819\n",
      "第286轮NuSVC的CV为：0.9830792659477918，basic的CV为：0.9830792659477918\n",
      "第287轮QDA的CV为：0.9646827590442847，basic的CV为：0.9577267981082243\n",
      "第287轮NuSVC的CV为：0.9612354892205638，basic的CV为：0.9612354892205638\n",
      "第288轮QDA的CV为：0.9842828469970665，basic的CV为：0.9729658792650918\n",
      "第288轮NuSVC的CV为：0.9671916010498688，basic的CV为：0.9628531727651691\n",
      "第289轮QDA的CV为：0.9718441095446443，basic的CV为：0.9651920272241128\n",
      "第289轮NuSVC的CV为：0.9673310646572678，basic的CV为：0.964057689191379\n",
      "第290轮QDA的CV为：0.9832097496512737，basic的CV为：0.97810733426327\n",
      "第290轮NuSVC的CV为：0.9682989501505029，basic的CV为：0.9642317010498495\n",
      "第291轮QDA的CV为：0.9806507707271067，basic的CV为：0.974932948215391\n",
      "第291轮NuSVC的CV为：0.9710866810103452，basic的CV为：0.9654130684665037\n",
      "第292轮QDA的CV为：0.9853522847863139，basic的CV为：0.9853522847863139\n",
      "第292轮NuSVC的CV为：0.9781329388687012，basic的CV为：0.9735987394282407\n",
      "第293轮QDA的CV为：0.9704729008143106，basic的CV为：0.961353624349219\n",
      "第293轮NuSVC的CV为：0.9517754638900013，basic的CV为：0.9484965291683353\n",
      "第294轮QDA的CV为：0.9812131373312176，basic的CV为：0.9732942127481213\n",
      "第294轮NuSVC的CV为：0.9572168497949163，basic的CV为：0.9555851621208946\n",
      "第295轮QDA的CV为：0.9741400825051014，basic的CV为：0.9615220869973722\n",
      "第295轮NuSVC的CV为：0.9543432623280531，basic的CV为：0.9410866597178384\n",
      "第296轮QDA的CV为：0.9796847178731504，basic的CV为：0.9675508073923988\n",
      "第296轮NuSVC的CV为：0.9590010437772457，basic的CV为：0.9420550132007122\n",
      "第297轮QDA的CV为：0.9770505836575876，basic的CV为：0.9676731517509728\n",
      "第297轮NuSVC的CV为：0.9663035019455253，basic的CV为：0.9630739299610894\n",
      "第298轮QDA的CV为：0.9759254431699687，basic的CV为：0.9721454640250261\n",
      "第298轮NuSVC的CV为：0.97007299270073，basic的CV为：0.959397810218978\n",
      "第299轮QDA的CV为：0.9843482514062117，basic的CV为：0.979294040922801\n",
      "第299轮NuSVC的CV为：0.9749870927420451，basic的CV为：0.9739409255183282\n",
      "第300轮QDA的CV为：0.9784704569409138，basic的CV为：0.9708289416578834\n",
      "第300轮NuSVC的CV为：0.957235414470829，basic的CV为：0.9516709033418067\n",
      "第301轮QDA的CV为：0.9908200474898237，basic的CV为：0.9797602894617821\n",
      "第301轮NuSVC的CV为：0.9796048168249661，basic的CV为：0.9732799072817729\n",
      "第302轮QDA的CV为：0.9772447812410637，basic的CV为：0.9719187875321704\n",
      "第302轮NuSVC的CV为：0.964469545324564，basic的CV为：0.9610666285387475\n",
      "第303轮QDA的CV为：0.9803153965391795，basic的CV为：0.9742178355635694\n",
      "第303轮NuSVC的CV为：0.9744076429091139，basic的CV为：0.9671158773844548\n",
      "第304轮QDA的CV为：0.9847995950776628，basic的CV为：0.9804814779665306\n",
      "第304轮NuSVC的CV为：0.9754357660308122，basic的CV为：0.9749137958305653\n",
      "第305轮QDA的CV为：0.9813066728667752，basic的CV为：0.9700999767495931\n",
      "第305轮NuSVC的CV为：0.9631713554987212，basic的CV为：0.9565604898085718\n",
      "第306轮QDA的CV为：0.9788883587786259，basic的CV为：0.9693463740458016\n",
      "第306轮NuSVC的CV为：0.9635019083969465，basic的CV为：0.9548204062159213\n",
      "第307轮QDA的CV为：0.979291101055807，basic的CV为：0.9743288084464555\n",
      "第307轮NuSVC的CV为：0.9642986425339368，basic的CV为：0.955444947209653\n",
      "第308轮QDA的CV为：0.9753749887727913，basic的CV为：0.9699410197299482\n",
      "第308轮NuSVC的CV为：0.960487709948804，basic的CV为：0.960487709948804\n",
      "第309轮QDA的CV为：0.9803296617451762，basic的CV为：0.9675806577611125\n",
      "第309轮NuSVC的CV为：0.9728927427544722，basic的CV为：0.9713147410358566\n",
      "第310轮QDA的CV为：0.9709254584714577，basic的CV为：0.965664061312004\n",
      "第310轮NuSVC的CV为：0.9657248867126912，basic的CV为：0.9616115689912107\n",
      "第311轮QDA的CV为：0.981556207711747，basic的CV为：0.9709111165457459\n",
      "第311轮NuSVC的CV为：0.9706257982120051，basic的CV为：0.9577389201380397\n",
      "第312轮QDA的CV为：0.9717189368159661，basic的CV为：0.9587050492044513\n",
      "第312轮NuSVC的CV为：0.9481130629387781，basic的CV为：0.9400272847113645\n",
      "第313轮QDA的CV为：0.9681886762487095，basic的CV为：0.9628523783054078\n",
      "第313轮NuSVC的CV为：0.9606765663463829，basic的CV为：0.9555626141507186\n",
      "第314轮QDA的CV为：0.9774583985412568，basic的CV为：0.9749170842366724\n",
      "第314轮NuSVC的CV为：0.9650748754468836，basic的CV为：0.9650748754468836\n",
      "第315轮QDA的CV为：0.9722834543568465，basic的CV为：0.9688310425311202\n",
      "第315轮NuSVC的CV为：0.9663835580912863，basic的CV为：0.9661890560165975\n",
      "第316轮QDA的CV为：0.9843811410335016，basic的CV为：0.963566476442013\n",
      "第316轮NuSVC的CV为：0.957851876306812，basic的CV为：0.953418541401374\n",
      "第317轮QDA的CV为：0.9804522400676248，basic的CV为：0.9673851789236404\n",
      "第317轮NuSVC的CV为：0.9600309946463792，basic的CV为：0.9540786136939982\n",
      "第318轮QDA的CV为：0.9804528325009958，basic的CV为：0.97018903765434\n",
      "第318轮NuSVC的CV为：0.9678758540396458，basic的CV为：0.9604460920984099\n",
      "第319轮QDA的CV为：0.9802443350881208，basic的CV为：0.9722476539253834\n",
      "第319轮NuSVC的CV为：0.9739070725566491，basic的CV为：0.9652738040741589\n",
      "第320轮QDA的CV为：0.9661953689003384，basic的CV为：0.9563250756373003\n",
      "第320轮NuSVC的CV为：0.9486565018122997，basic的CV为：0.9429724710181829\n",
      "第321轮QDA的CV为：0.9638501742160279，basic的CV为：0.9538327526132404\n",
      "第321轮NuSVC的CV为：0.9542198993418506，basic的CV为：0.9519857400954962\n",
      "第322轮QDA的CV为：0.9861719434363658，basic的CV为：0.9803059692153673\n",
      "第322轮NuSVC的CV为：0.9792735577524715，basic的CV为：0.9765204605180828\n",
      "第323轮QDA的CV为：0.9815673395795099，basic的CV为：0.9720692874851089\n",
      "第323轮NuSVC的CV为：0.963279564699443，basic的CV为：0.9547796129946231\n",
      "第324轮QDA的CV为：0.9742422462406015，basic的CV为：0.9661213580827068\n",
      "第324轮NuSVC的CV为：0.9644619360902256，basic的CV为：0.9560767152255638\n",
      "第325轮QDA的CV为：0.9896390484003281，basic的CV为：0.9802953240360953\n",
      "第325轮NuSVC的CV为：0.9748646431501231，basic的CV为：0.9687366694011484\n",
      "第326轮QDA的CV为：0.9806846999154691，basic的CV为：0.9806846999154691\n",
      "第326轮NuSVC的CV为：0.972879684418146，basic的CV为：0.9675824175824176\n",
      "第327轮QDA的CV为：0.9803149606299212，basic的CV为：0.9720816147981503\n",
      "第327轮NuSVC的CV为：0.9668947631546057，basic的CV为：0.9653402699662543\n",
      "第328轮QDA的CV为：0.9776601865310077，basic的CV为：0.9657521802325582\n",
      "第328轮NuSVC的CV为：0.9675841812015504，basic的CV为：0.9481286337209301\n",
      "第329轮QDA的CV为：0.9611476793248945，basic的CV为：0.9518481012658228\n",
      "第329轮NuSVC的CV为：0.9521012658227848，basic的CV为：0.9477721518987342\n",
      "第330轮QDA的CV为：0.9815818443559997，basic的CV为：0.9787619612160217\n",
      "第330轮NuSVC的CV为：0.9702938436785502，basic的CV为：0.9658650182064528\n",
      "第331轮QDA的CV为：0.9786485444380182，basic的CV为：0.9735878156930788\n",
      "第331轮NuSVC的CV为：0.9663582032003084，basic的CV为：0.961779448621554\n",
      "第332轮QDA的CV为：0.9700300513685955，basic的CV为：0.9700300513685955\n",
      "第332轮NuSVC的CV为：0.9635978741987536，basic的CV为：0.9635978741987536\n",
      "第333轮QDA的CV为：0.9792714163741582，basic的CV为：0.9669069980710243\n",
      "第333轮NuSVC的CV为：0.961681371153907，basic的CV为：0.961681371153907\n",
      "第334轮QDA的CV为：0.9903511616935959，basic的CV为：0.982898926760231\n",
      "第334轮NuSVC的CV为：0.9836802688996344，basic的CV为：0.9783288123599482\n",
      "第335轮QDA的CV为：0.9765469669304119，basic的CV为：0.9710556569960099\n",
      "第335轮NuSVC的CV为：0.9756542909312234，basic的CV为：0.9703793872996551\n",
      "第336轮QDA的CV为：0.9776952776952776，basic的CV为：0.9690228690228689\n",
      "第336轮NuSVC的CV为：0.9672854172854173，basic的CV为：0.9564152064152064\n",
      "第337轮QDA的CV为：0.9867611773850157，basic的CV为：0.9862796120442026\n",
      "第337轮NuSVC的CV为：0.9746882498056841，basic的CV为：0.9746882498056841\n",
      "第338轮QDA的CV为：0.9836233211233212，basic的CV为：0.9740842490842491\n",
      "第338轮NuSVC的CV为：0.9693681318681319，basic的CV为：0.9652472527472528\n",
      "第339轮QDA的CV为：0.9871871484814398，basic的CV为：0.9836895388076491\n",
      "第339轮NuSVC的CV为：0.9785749437570304，basic的CV为：0.9747961192350956\n",
      "第340轮QDA的CV为：0.9794025888963579，basic的CV为：0.9770929210403847\n",
      "第340轮NuSVC的CV为：0.9800230966785598，basic的CV为：0.9724391126738715\n",
      "第341轮QDA的CV为：0.975241963359834，basic的CV为：0.975241963359834\n",
      "第341轮NuSVC的CV为：0.9746226523793063，basic的CV为：0.9724046549141606\n",
      "第342轮QDA的CV为：0.9714892890936448，basic的CV为：0.9632370049792917\n",
      "第342轮NuSVC的CV为：0.9595607054772208，basic的CV为：0.9595607054772208\n",
      "第343轮QDA的CV为：0.9732271954494177，basic的CV为：0.9649649649649651\n",
      "第343轮NuSVC的CV为：0.9597375152930708，basic的CV为：0.9487423931868375\n",
      "第344轮QDA的CV为：0.9707110183416974，basic的CV为：0.9619398676555491\n",
      "第344轮NuSVC的CV为：0.95850542553034，basic的CV为：0.9526357862947994\n",
      "第345轮QDA的CV为：0.9761734312749004，basic的CV为：0.9720026145418327\n",
      "第345轮NuSVC的CV为：0.9685943725099602，basic的CV为：0.9634197584661355\n",
      "第346轮QDA的CV为：0.961180221782427，basic的CV为：0.9538215059843558\n",
      "第346轮NuSVC的CV为：0.9420412779191406，basic的CV为：0.9388292024000251\n",
      "第347轮QDA的CV为：0.9887068268280264，basic的CV为：0.9853368480963364\n",
      "第347轮NuSVC的CV为：0.9787017344157205，basic的CV为：0.9670790522122038\n",
      "第348轮QDA的CV为：0.9814280646007189，basic的CV为：0.9733144117601691\n",
      "第348轮NuSVC的CV为：0.9716639158401333，basic的CV为：0.9504928349966836\n",
      "第349轮QDA的CV为：0.9628878066378067，basic的CV为：0.953838985088985\n",
      "第349轮NuSVC的CV为：0.9441438191438192，basic的CV为：0.9334265271765272\n",
      "第350轮QDA的CV为：0.9948149215036728，basic的CV为：0.9903211868068559\n",
      "第350轮NuSVC的CV为：0.9836814057323924，basic的CV为：0.9776465504825004\n",
      "第351轮QDA的CV为：0.9842900654337721，basic的CV为：0.9777175417887173\n",
      "第351轮NuSVC的CV为：0.9724128885585624，basic的CV为：0.9723035893848643\n",
      "第352轮QDA的CV为：0.9890166958624645，basic的CV为：0.9795836110809679\n",
      "第352轮NuSVC的CV为：0.967460895716744，basic的CV为：0.9604228998403275\n",
      "第353轮QDA的CV为：0.9907688526486237，basic的CV为：0.9855858200323849\n",
      "第353轮NuSVC的CV为：0.9827665972704142，basic的CV为：0.9804678464029608\n",
      "第354轮QDA的CV为：0.9772184551886793，basic的CV为：0.9677329009433964\n",
      "第354轮NuSVC的CV为：0.9640477594339624，basic的CV为：0.9585347877358491\n",
      "第355轮QDA的CV为：0.9732177709667625，basic的CV为：0.9636397808266709\n",
      "第355轮NuSVC的CV为：0.9569026587054893，basic的CV为：0.9436163599300265\n",
      "第356轮QDA的CV为：0.9688593155893537，basic的CV为：0.9516501901140685\n",
      "第356轮NuSVC的CV为：0.9504486692015209，basic的CV为：0.9405019011406844\n",
      "第357轮QDA的CV为：0.9691708893295146，basic的CV为：0.9590071234486304\n",
      "第357轮NuSVC的CV为：0.9663361974003085，basic的CV为：0.9545568039950063\n",
      "第358轮QDA的CV为：0.9802669065826961，basic的CV为：0.9682860998650472\n",
      "第358轮NuSVC的CV为：0.9735942420152947，basic的CV为：0.968825910931174\n",
      "第359轮QDA的CV为：0.9729191557148547，basic的CV为：0.9690077942766115\n",
      "第359轮NuSVC的CV为：0.9622375832053252，basic的CV为：0.9582551061045685\n",
      "第360轮QDA的CV为：0.984534138061109，basic的CV为：0.97736703130894\n",
      "第360轮NuSVC的CV为：0.967336511093584，basic的CV为：0.9643702204999829\n",
      "第361轮QDA的CV为：0.9920776897521084，basic的CV为：0.9879745577419996\n",
      "第361轮NuSVC的CV为：0.9861430559104978，basic的CV为：0.9794417468836073\n",
      "第362轮QDA的CV为：0.9759041908528417，basic的CV为：0.9726829112712891\n",
      "第362轮NuSVC的CV为：0.9714390508388084，basic的CV为：0.9714390508388084\n",
      "第363轮QDA的CV为：0.9720734562481711，basic的CV为：0.9658764998536729\n",
      "第363轮NuSVC的CV为：0.9653204565408252，basic的CV为：0.961076968100673\n",
      "第364轮QDA的CV为：0.9815729974160208，basic的CV为：0.9729489664082687\n",
      "第364轮NuSVC的CV为：0.9673530361757106，basic的CV为：0.9673530361757106\n",
      "第365轮QDA的CV为：0.9834088113608409，basic的CV为：0.9796488873979649\n",
      "第365轮NuSVC的CV为：0.9771329531477133，basic的CV为：0.9771329531477133\n",
      "第366轮QDA的CV为：0.9849340331961981，basic的CV为：0.9753014612001702\n",
      "第366轮NuSVC的CV为：0.9777982692580508，basic的CV为：0.9729252376223577\n",
      "第367轮QDA的CV为：0.9792174336837707，basic的CV为：0.9716493768711271\n",
      "第367轮NuSVC的CV为：0.9647705911021374，basic的CV为：0.961268537213674\n",
      "第368轮QDA的CV为：0.9706403162055336，basic的CV为：0.9596126482213438\n",
      "第368轮NuSVC的CV为：0.9516521739130435，basic的CV为：0.9444189723320159\n",
      "第369轮QDA的CV为：0.9744147157190635，basic的CV为：0.9604077082337952\n",
      "第369轮NuSVC的CV为：0.9623347666825928，basic的CV为：0.9582736104475236\n",
      "第370轮QDA的CV为：0.9849464509513501，basic的CV为：0.9771419619459951\n",
      "第370轮NuSVC的CV为：0.9789221829782386，basic的CV为：0.9722570354335194\n",
      "第371轮QDA的CV为：0.9742384615384616，basic的CV为：0.9624615384615384\n",
      "第371轮NuSVC的CV为：0.9635846153846154，basic的CV为：0.9602230769230768\n",
      "第372轮QDA的CV为：0.9719656283566058，basic的CV为：0.9572172188713541\n",
      "第372轮NuSVC的CV为：0.9519953730480046，basic的CV为：0.94679005205321\n",
      "第373轮QDA的CV为：0.9703827751196173，basic的CV为：0.9544634313055367\n",
      "第373轮NuSVC的CV为：0.9589747095010254，basic的CV为：0.9465481886534518\n",
      "第374轮QDA的CV为：0.958828125，basic的CV为：0.9480781250000001\n",
      "第374轮NuSVC的CV为：0.9479687499999999，basic的CV为：0.9441484375000001\n",
      "第375轮QDA的CV为：0.9635000341600055，basic的CV为：0.9635000341600055\n",
      "第375轮NuSVC的CV为：0.9607330737172918，basic的CV为：0.9584187333469972\n",
      "第376轮QDA的CV为：0.9724505952924697，basic的CV为：0.9714381413477056\n",
      "第376轮NuSVC的CV为：0.9708443713650619，basic的CV为：0.9685606406625864\n",
      "第377轮QDA的CV为：0.9718537325541217，basic的CV为：0.964935474663101\n",
      "第377轮NuSVC的CV为：0.9569881165211904，basic的CV为：0.9524811081231315\n",
      "第378轮QDA的CV为：0.9772286821705427，basic的CV为：0.9631782945736435\n",
      "第378轮NuSVC的CV为：0.9631157789447362，basic的CV为：0.9594429857464366\n",
      "第379轮QDA的CV为：0.9888199233716475，basic的CV为：0.9842758620689656\n",
      "第379轮NuSVC的CV为：0.9819463601532568，basic的CV为：0.9753563218390805\n",
      "第380轮QDA的CV为：0.9665734037986113，basic的CV为：0.9647081464375064\n",
      "第380轮NuSVC的CV为：0.9568326153572856，basic的CV为：0.9529392607067253\n",
      "第381轮QDA的CV为：0.966952264381885，basic的CV为：0.9619251617415633\n",
      "第381轮NuSVC的CV为：0.955076645101125，basic的CV为：0.9537725126770414\n",
      "第382轮QDA的CV为：0.9855074583211465，basic的CV为：0.9751096811933314\n",
      "第382轮NuSVC的CV为：0.9597835624451593，basic的CV为：0.9597835624451593\n",
      "第383轮QDA的CV为：0.9751765536723163，basic的CV为：0.9585028248587572\n",
      "第383轮NuSVC的CV为：0.9600282485875706，basic的CV为：0.9508757062146893\n",
      "第384轮QDA的CV为：0.9680830581021638，basic的CV为：0.9638398080540292\n",
      "第384轮NuSVC的CV为：0.9548571513203691，basic的CV为：0.9520949658614613\n",
      "第385轮QDA的CV为：0.9709480848369735，basic的CV为：0.9696897752453308\n",
      "第385轮NuSVC的CV为：0.9630262741373852，basic的CV为：0.9588239949351061\n",
      "第386轮QDA的CV为：0.9633128161374857，basic的CV为：0.9633128161374857\n",
      "第386轮NuSVC的CV为：0.9543847888353951，basic的CV为：0.9495170333353649\n",
      "第387轮QDA的CV为：0.9800996574275926，basic的CV为：0.9654313298037994\n",
      "第387轮NuSVC的CV为：0.9531765805045157，basic的CV为：0.9474151354718157\n",
      "第388轮QDA的CV为：0.9556277056277056，basic的CV为：0.949989404534859\n",
      "第388轮NuSVC的CV为：0.9510640874277239，basic的CV为：0.9394090757727122\n",
      "第389轮QDA的CV为：0.9805585191959221，basic的CV为：0.9692966858033158\n",
      "第389轮NuSVC的CV为：0.967653982285898，basic的CV为：0.9654016156073768\n",
      "第390轮QDA的CV为：0.9837395194593919，basic的CV为：0.9734545113252409\n",
      "第390轮NuSVC的CV为：0.9793830559379303，basic的CV为：0.9659617069202853\n",
      "第391轮QDA的CV为：0.9910062565172053，basic的CV为：0.9826024916305361\n",
      "第391轮NuSVC的CV为：0.9749190494484385，basic的CV为：0.9658018220734317\n",
      "第392轮QDA的CV为：0.9776132107592782，basic的CV为：0.9735841561684259\n",
      "第392轮NuSVC的CV为：0.9723357167177393，basic的CV为：0.9723357167177393\n",
      "第393轮QDA的CV为：0.9744606003752345，basic的CV为：0.9692464040025015\n",
      "第393轮NuSVC的CV为：0.963117573483427，basic的CV为：0.9514149468417762\n",
      "第394轮QDA的CV为：0.9720483175676705，basic的CV为：0.9568157930813874\n",
      "第394轮NuSVC的CV为：0.9657115873813767，basic的CV为：0.9531447546801932\n",
      "第395轮QDA的CV为：0.9754242789576185，basic的CV为：0.9719079415128752\n",
      "第395轮NuSVC的CV为：0.969705400476058，basic的CV为：0.963229157006399\n",
      "第396轮QDA的CV为：0.9778291855853785，basic的CV为：0.9778291855853785\n",
      "第396轮NuSVC的CV为：0.9728720041381689，basic的CV为：0.9720098856256107\n",
      "第397轮QDA的CV为：0.9796015712682379，basic的CV为：0.9737373737373738\n",
      "第397轮NuSVC的CV为：0.9677749719416386，basic的CV为：0.9624649270482604\n",
      "第398轮QDA的CV为：0.9741024570610687，basic的CV为：0.9700694775763359\n",
      "第398轮NuSVC的CV为：0.9712398616412214，basic的CV为：0.9712398616412214\n",
      "第399轮QDA的CV为：0.9772686298076922，basic的CV为：0.9676382211538462\n",
      "第399轮NuSVC的CV为：0.9695913461538461，basic的CV为：0.9683668870192307\n",
      "第400轮QDA的CV为：0.9743734159391721，basic的CV为：0.9639538158265277\n",
      "第400轮NuSVC的CV为：0.9636018023092087，basic的CV为：0.9557730216840327\n",
      "第401轮QDA的CV为：0.9811407313492593，basic的CV为：0.9771787001201122\n",
      "第401轮NuSVC的CV为：0.9723241692246096，basic的CV为：0.9678366475377019\n",
      "第402轮QDA的CV为：0.9745415393164768，basic的CV为：0.9677167268685747\n",
      "第402轮NuSVC的CV为：0.9704605445957212，basic的CV为：0.9628889969435954\n",
      "第403轮QDA的CV为：0.9805790325481029，basic的CV为：0.9673320146256669\n",
      "第403轮NuSVC的CV为：0.9601690343463405，basic的CV为：0.9556209914284002\n",
      "第404轮QDA的CV为：0.9763483919409335，basic的CV为：0.9696455387310724\n",
      "第404轮NuSVC的CV为：0.9664153422600426，basic的CV为：0.9584689025153298\n",
      "第405轮QDA的CV为：0.981059482616576，basic的CV为：0.9735870818915803\n",
      "第405轮NuSVC的CV为：0.9696490360850223，basic的CV为：0.9659746251441753\n",
      "第406轮QDA的CV为：0.978881520922484，basic的CV为：0.9707150556952928\n",
      "第406轮NuSVC的CV为：0.9596821611733576，basic的CV为：0.9596821611733576\n",
      "第407轮QDA的CV为：0.9839123391233913，basic的CV为：0.9742297422974229\n",
      "第407轮NuSVC的CV为：0.9718372183721837，basic的CV为：0.9718372183721837\n",
      "第408轮QDA的CV为：0.9709068833473429，basic的CV为：0.9709068833473429\n",
      "第408轮NuSVC的CV为：0.958127081971296，basic的CV为：0.957473304068989\n",
      "第409轮QDA的CV为：0.987419345546722，basic的CV为：0.9757460536928219\n",
      "第409轮NuSVC的CV为：0.969020048392672，basic的CV为：0.9644328263624842\n",
      "第410轮QDA的CV为：0.9698223039215687，basic的CV为：0.9645045518207283\n",
      "第410轮NuSVC的CV为：0.9673202614379085，basic的CV为：0.9542629551820728\n",
      "第411轮QDA的CV为：0.9745670834942228，basic的CV为：0.9596192116909733\n",
      "第411轮NuSVC的CV为：0.9641637211512757，basic的CV为：0.9604657379629905\n",
      "第412轮QDA的CV为：0.9662778042310047，basic的CV为：0.9575028909597987\n",
      "第412轮NuSVC的CV为：0.9596285966941023，basic的CV为：0.9540252363784778\n",
      "第413轮QDA的CV为：0.97890759446315，basic的CV为：0.9744257388701834\n",
      "第413轮NuSVC的CV为：0.9708193041526375，basic的CV为：0.969951365506921\n",
      "第414轮QDA的CV为：0.9883739174093135，basic的CV为：0.9817528555290574\n",
      "第414轮NuSVC的CV为：0.9797131919166562，basic的CV为：0.9768890422994855\n",
      "第415轮QDA的CV为：0.9729094780118934，basic的CV为：0.970442698773952\n",
      "第415轮NuSVC的CV为：0.9631891931576242，basic的CV为：0.9559356875412965\n",
      "第416轮QDA的CV为：0.9828055551283432，basic的CV为：0.9769920487227204\n",
      "第416轮NuSVC的CV为：0.9703941803417357，basic的CV为：0.9637424831977361\n",
      "第417轮QDA的CV为：0.9835185157787278，basic的CV为：0.9761062863398973\n",
      "第417轮NuSVC的CV为：0.9714754922992705，basic的CV为：0.9648473909248272\n",
      "第418轮QDA的CV为：0.9785870996050899，basic的CV为：0.9696211788796255\n",
      "第418轮NuSVC的CV为：0.9656867046950418，basic的CV为：0.9576934327921602\n",
      "第419轮QDA的CV为：0.9773871870291987，basic的CV为：0.9581605567113853\n",
      "第419轮NuSVC的CV为：0.9528517110266159，basic的CV为：0.9400961331515891\n",
      "第420轮QDA的CV为：0.9758568548387097，basic的CV为：0.9668010752688172\n",
      "第420轮NuSVC的CV为：0.9619791666666667，basic的CV为：0.9607442876344086\n",
      "第421轮QDA的CV为：0.9724975568895714，basic的CV为：0.9671614934772829\n",
      "第421轮NuSVC的CV为：0.9674096824732034，basic的CV为：0.9666030682364621\n",
      "第422轮QDA的CV为：0.9683298319327731，basic的CV为：0.9581057422969187\n",
      "第422轮NuSVC的CV为：0.9474702380952381，basic的CV为：0.9474702380952381\n",
      "第423轮QDA的CV为：0.9902529808844868，basic的CV为：0.9842438962841461\n",
      "第423轮NuSVC的CV为：0.9783136710617627，basic的CV为：0.9783136710617627\n",
      "第424轮QDA的CV为：0.9779903929443263，basic的CV为：0.9611465469722025\n",
      "第424轮NuSVC的CV为：0.9553823135679975，basic的CV为：0.9364516891093786\n",
      "第425轮QDA的CV为：0.9751653513116273，basic的CV为：0.9689906380803273\n",
      "第425轮NuSVC的CV为：0.9660510592069528，basic的CV为：0.960147937501997\n",
      "第426轮QDA的CV为：0.9747271078288473，basic的CV为：0.9646454831082176\n",
      "第426轮NuSVC的CV为：0.9708969002040618，basic的CV为：0.9661678489294853\n",
      "第427轮QDA的CV为：0.9654041634755971，basic的CV为：0.956958852607394\n",
      "第427轮NuSVC的CV为：0.9541460697625844，basic的CV为：0.9455615122188958\n",
      "第428轮QDA的CV为：0.9711073804443329，basic的CV为：0.9648236475461278\n",
      "第428轮NuSVC的CV为：0.9610894941634242，basic的CV为：0.9567277519769047\n",
      "第429轮QDA的CV为：0.9875740994072048，basic的CV为：0.9812898996808025\n",
      "第429轮NuSVC的CV为：0.9799504103967168，basic的CV为：0.9748917008663931\n",
      "第430轮QDA的CV为：0.9849664062262486，basic的CV为：0.9734365974523455\n",
      "第430轮NuSVC的CV为：0.9744626516280059，basic的CV为：0.9715592983309519\n",
      "第431轮QDA的CV为：0.9732779906988018，basic的CV为：0.966355111871904\n",
      "第431轮NuSVC的CV为：0.962787632739006，basic的CV为：0.9564706951661036\n",
      "第432轮QDA的CV为：0.9714，basic的CV为：0.9672083333333333\n",
      "第432轮NuSVC的CV为：0.9580833333333333，basic的CV为：0.9532666666666666\n",
      "第433轮QDA的CV为：0.974478756884343，basic的CV为：0.96897128245476\n",
      "第433轮NuSVC的CV为：0.9633326776816155，basic的CV为：0.9565630736952531\n",
      "第434轮QDA的CV为：0.9839359108324626，basic的CV为：0.9734169278996866\n",
      "第434轮NuSVC的CV为：0.9766213862765588，basic的CV为：0.966388018112156\n",
      "第435轮QDA的CV为：0.9673866421568627，basic的CV为：0.9640395220588236\n",
      "第435轮NuSVC的CV为：0.9554534313725491，basic的CV为：0.953063725490196\n",
      "第436轮QDA的CV为：0.9679481241315424，basic的CV为：0.9563841284545314\n",
      "第436轮NuSVC的CV为：0.9587000154392465，basic的CV为：0.9528253821213524\n",
      "第437轮QDA的CV为：0.9765085879312966，basic的CV为：0.9695166438668491\n",
      "第437轮NuSVC的CV为：0.9749050007599939，basic的CV为：0.9651618787049703\n",
      "第438轮QDA的CV为：0.9759859847028158，basic的CV为：0.968714837129143\n",
      "第438轮NuSVC的CV为：0.9630389266333376，basic的CV为：0.9607671381161959\n",
      "第439轮QDA的CV为：0.9864253393665158，basic的CV为：0.9805350480273081\n",
      "第439轮NuSVC的CV为：0.9712629991267762，basic的CV为：0.96873858855283\n",
      "第440轮QDA的CV为：0.9767277752458801，basic的CV为：0.9721703591537909\n",
      "第440轮NuSVC的CV为：0.9641581276370532，basic的CV为：0.9580056159127329\n",
      "第441轮QDA的CV为：0.9897754721124287，basic的CV为：0.9794548748353097\n",
      "第441轮NuSVC的CV为：0.9791941150636803，basic的CV为：0.9728672595520422\n",
      "第442轮QDA的CV为：0.9873801604382704，basic的CV为：0.9791905414092853\n",
      "第442轮NuSVC的CV为：0.9762556949995808，basic的CV为：0.9702043212119518\n",
      "第443轮QDA的CV为：0.976285334851072，basic的CV为：0.9652817302219693\n",
      "第443轮NuSVC的CV为：0.9615664326819705，basic的CV为：0.9588471510782267\n",
      "第444轮QDA的CV为：0.9933279698163459，basic的CV为：0.987169172723742\n",
      "第444轮NuSVC的CV为：0.9771958053598181，basic的CV为：0.9653706375187261\n",
      "第445轮QDA的CV为：0.9771040684477039，basic的CV为：0.9747831907339503\n",
      "第445轮NuSVC的CV为：0.9745649263721553，basic的CV为：0.9694575403061522\n",
      "第446轮QDA的CV为：0.9751894063091894，basic的CV为：0.9671641303637908\n",
      "第446轮NuSVC的CV为：0.9681601463000457，basic的CV为：0.958502057344393\n",
      "第447轮QDA的CV为：0.9745566039333932，basic的CV为：0.9700506496366439\n",
      "第447轮NuSVC的CV为：0.9674927582877372，basic的CV为：0.9674758186099301\n",
      "第448轮QDA的CV为：0.9867724867724867，basic的CV为：0.98011583011583\n",
      "第448轮NuSVC的CV为：0.9812669812669814，basic的CV为：0.9787501787501788\n",
      "第449轮QDA的CV为：0.964301751234845，basic的CV为：0.9572519083969465\n",
      "第449轮NuSVC的CV为：0.960499925160904，basic的CV为：0.9537344708875916\n",
      "第450轮QDA的CV为：0.9739516780494969，basic的CV为：0.9685833884115443\n",
      "第450轮NuSVC的CV为：0.9664977601527504，basic的CV为：0.9627230667547917\n",
      "第451轮QDA的CV为：0.975111963235079，basic的CV为：0.9726751163539502\n",
      "第451轮NuSVC的CV为：0.9726751163539502，basic的CV为：0.970992008898516\n",
      "第452轮QDA的CV为：0.9865445530918606，basic的CV为：0.9747693155212781\n",
      "第452轮NuSVC的CV为：0.9737777165679659，basic的CV为：0.9709819584079328\n",
      "第453轮QDA的CV为：0.9749302325581395，basic的CV为：0.968279069767442\n",
      "第453轮NuSVC的CV为：0.9691317829457363，basic的CV为：0.9650542635658916\n",
      "第454轮QDA的CV为：0.9750483219117906，basic的CV为：0.9629385579570082\n",
      "第454轮NuSVC的CV为：0.9532302465881803，basic的CV为：0.9504773619164764\n",
      "第455轮QDA的CV为：0.9843399910377056，basic的CV为：0.9769380961526151\n",
      "第455轮NuSVC的CV为：0.9701843671980027，basic的CV为：0.968655975929838\n",
      "第456轮QDA的CV为：0.9887360594795539，basic的CV为：0.9768921933085501\n",
      "第456轮NuSVC的CV为：0.9806988847583642，basic的CV为：0.9697472118959107\n",
      "第457轮QDA的CV为：0.9681026130019532，basic的CV为：0.9590901401085804\n",
      "第457轮NuSVC的CV为：0.9597668445579122，basic的CV为：0.9583519170729456\n",
      "第458轮QDA的CV为：0.9654785979103472，basic的CV为：0.952258173238962\n",
      "第458轮NuSVC的CV为：0.9448938321536907，basic的CV为：0.937891809908999\n",
      "第459轮QDA的CV为：0.9794122448979591，basic的CV为：0.9752326530612245\n",
      "第459轮NuSVC的CV为：0.9654857142857143，basic的CV为：0.9593632653061225\n",
      "第460轮QDA的CV为：0.9773358577134093，basic的CV为：0.973148519229972\n",
      "第460轮NuSVC的CV为：0.9640690427163804，basic的CV为：0.9640621329169027\n",
      "第461轮QDA的CV为：0.9879793061472917，basic的CV为：0.9829732197200244\n",
      "第461轮NuSVC的CV为：0.9727936701156422，basic的CV为：0.9698569689592209\n",
      "第462轮QDA的CV为：0.9747942012092956，basic的CV为：0.9640708093538283\n",
      "第462轮NuSVC的CV为：0.9619727544255846，basic的CV为：0.9608873023967364\n",
      "第463轮QDA的CV为：0.9820381231671556，basic的CV为：0.9797590207828637\n",
      "第463轮NuSVC的CV为：0.9682041310722937，basic的CV为：0.9608727527731734\n",
      "第464轮QDA的CV为：0.9852480789298929，basic的CV为：0.9821965025456155\n",
      "第464轮NuSVC的CV为：0.9774689308414762，basic的CV为：0.9746070897764285\n",
      "第465轮QDA的CV为：0.9639964268951839，basic的CV为：0.9603249102938727\n",
      "第465轮NuSVC的CV为：0.9535799179397114，basic的CV为：0.9460173507547427\n",
      "第466轮QDA的CV为：0.9834536044901231，basic的CV为：0.9739193818791457\n",
      "第466轮NuSVC的CV为：0.9593847948101174，basic的CV为：0.9570668416065311\n",
      "第467轮QDA的CV为：0.9720770113476986，basic的CV为：0.9672000510008925\n",
      "第467轮NuSVC的CV为：0.9602193038378172，basic的CV为：0.9533023077903864\n",
      "第468轮QDA的CV为：0.9844698266471632，basic的CV为：0.9744950359465937\n",
      "第468轮NuSVC的CV为：0.9735146743020759，basic的CV为：0.9717562478603219\n",
      "第469轮QDA的CV为：0.9806368468237705，basic的CV为：0.9670370133196722\n",
      "第469轮NuSVC的CV为：0.973264600409836，basic的CV为：0.971983862704918\n",
      "第470轮QDA的CV为：0.9810105024127165，basic的CV为：0.9784132841328413\n",
      "第470轮NuSVC的CV为：0.9702952029520296，basic的CV为：0.9649872267953449\n",
      "第471轮QDA的CV为：0.9671844601167315，basic的CV为：0.9633086089494165\n",
      "第471轮NuSVC的CV为：0.9630198200389106，basic的CV为：0.9630198200389106\n",
      "第472轮QDA的CV为：0.9895728326612903，basic的CV为：0.9862651209677419\n",
      "第472轮NuSVC的CV为：0.98193359375，basic的CV为：0.9801222278225806\n",
      "第473轮QDA的CV为：0.9721739130434782，basic的CV为：0.9717570966582825\n",
      "第473轮NuSVC的CV为：0.9636363636363636，basic的CV为：0.9555587495508444\n",
      "第474轮QDA的CV为：0.9756153754682204，basic的CV为：0.9697068791247992\n",
      "第474轮NuSVC的CV为：0.9737499256792913，basic的CV为：0.9688001664783876\n",
      "第475轮QDA的CV为：0.9840014940239044，basic的CV为：0.9772161354581673\n",
      "第475轮NuSVC的CV为：0.9763134960159362，basic的CV为：0.9724072460159363\n",
      "第476轮QDA的CV为：0.9950157162101483，basic的CV为：0.9885645861397995\n",
      "第476轮NuSVC的CV为：0.9881155515641371，basic的CV为：0.9872923215087561\n",
      "第477轮QDA的CV为：0.9801498724489796，basic的CV为：0.9785235969387756\n",
      "第477轮NuSVC的CV为：0.9710299744897959，basic的CV为：0.9666135204081632\n",
      "第478轮QDA的CV为：0.9800644538547764，basic的CV为：0.9698592601818409\n",
      "第478轮NuSVC的CV为：0.9572642919417113，basic的CV为：0.950351849545398\n",
      "第479轮QDA的CV为：0.9819984082193765，basic的CV为：0.9798422690109254\n",
      "第479轮NuSVC的CV为：0.9686129802474495，basic的CV为：0.9629983358657116\n",
      "第480轮QDA的CV为：0.9751975748069498，basic的CV为：0.9644666988416989\n",
      "第480轮NuSVC的CV为：0.9600174951737451，basic的CV为：0.9584942084942085\n",
      "第481轮QDA的CV为：0.983579222375903，basic的CV为：0.9653757491931766\n",
      "第481轮NuSVC的CV为：0.9703396342400492，basic的CV为：0.9615337329030275\n",
      "第482轮QDA的CV为：0.9866375668825683，basic的CV为：0.9787243030132357\n",
      "第482轮NuSVC的CV为：0.9788791889608561，basic的CV为：0.976274288932695\n",
      "第483轮QDA的CV为：0.977640625，basic的CV为：0.97359375\n",
      "第483轮NuSVC的CV为：0.96665625，basic的CV为：0.9618828125\n",
      "第484轮QDA的CV为：0.9926119284923272，basic的CV为：0.9849390919158361\n",
      "第484轮NuSVC的CV为：0.9839265938933713，basic的CV为：0.9756209460528398\n",
      "第485轮QDA的CV为：0.9829082935327866，basic的CV为：0.9812486858310054\n",
      "第485轮NuSVC的CV为：0.977080892733772，basic的CV为：0.977080892733772\n",
      "第486轮QDA的CV为：0.9759686346863468，basic的CV为：0.9642681426814268\n",
      "第486轮NuSVC的CV为：0.9603628536285362，basic的CV为：0.9564114391143911\n",
      "第487轮QDA的CV为：0.9724376692238267，basic的CV为：0.9634476534296029\n",
      "第487轮NuSVC的CV为：0.9584414485559567，basic的CV为：0.9574402075812275\n",
      "第488轮QDA的CV为：0.9890992541594952，basic的CV为：0.9815771020590298\n",
      "第488轮NuSVC的CV为：0.9796328169822146，basic的CV为：0.9781347612672915\n",
      "第489轮QDA的CV为：0.9801554801554802，basic的CV为：0.9617279617279617\n",
      "第489轮NuSVC的CV为：0.9594139594139595，basic的CV为：0.9494689494689496\n",
      "第490轮QDA的CV为：0.9751384074093238，basic的CV为：0.9700074162225556\n",
      "第490轮NuSVC的CV为：0.9546920542936479，basic的CV为：0.9520015177385694\n",
      "第491轮QDA的CV为：0.9646987366375122，basic的CV为：0.9591026886945254\n",
      "第491轮NuSVC的CV为：0.9514739229024944，basic的CV为：0.9501457725947522\n",
      "第492轮QDA的CV为：0.9854297515984836，basic的CV为：0.9714111922141118\n",
      "第492轮NuSVC的CV为：0.9581706558026369，basic的CV为：0.9477522209019407\n",
      "第493轮QDA的CV为：0.9832370468088978，basic的CV为：0.9791353925002655\n",
      "第493轮NuSVC的CV为：0.9694157606635431，basic的CV为：0.9693475261186675\n",
      "第494轮QDA的CV为：0.9797752808988762，basic的CV为：0.9744894353755919\n",
      "第494轮NuSVC的CV为：0.9684969260122959，basic的CV为：0.9647657409370363\n",
      "第495轮QDA的CV为：0.9678721859114017，basic的CV为：0.9582135076252722\n",
      "第495轮NuSVC的CV为：0.954872912127814，basic的CV为：0.9463834422657953\n",
      "第496轮QDA的CV为：0.9815368874695769，basic的CV为：0.9782504224347746\n",
      "第496轮NuSVC的CV为：0.969259150169749，basic的CV为：0.9674066380392826\n",
      "第497轮QDA的CV为：0.9800196622715707，basic的CV为：0.9642753874624104\n",
      "第497轮NuSVC的CV为：0.9631332408049966，basic的CV为：0.9577261161230628\n",
      "第498轮QDA的CV为：0.967172613764128，basic的CV为：0.9630451506006272\n",
      "第498轮NuSVC的CV为：0.9620095863660572，basic的CV为：0.9620095863660572\n",
      "第499轮QDA的CV为：0.9700760304121648，basic的CV为：0.9625450180072029\n",
      "第499轮NuSVC的CV为：0.9594237695078032，basic的CV为：0.9525410164065626\n",
      "第500轮QDA的CV为：0.9717582557694759，basic的CV为：0.9609046283309958\n",
      "第500轮NuSVC的CV为：0.9610799438990182，basic的CV为：0.9563623613413235\n",
      "第501轮QDA的CV为：0.9747164126196387，basic的CV为：0.9603154909606524\n",
      "第501轮NuSVC的CV为：0.9546969159872386，basic的CV为：0.9468628146047501\n",
      "第502轮QDA的CV为：0.9798624468147111，basic的CV为：0.9651017077577665\n",
      "第502轮NuSVC的CV为：0.9704930931981115，basic的CV为：0.9599580346214374\n",
      "第503轮QDA的CV为：0.9819231820667031，basic的CV为：0.9769341170038272\n",
      "第503轮NuSVC的CV为：0.972132996172772，basic的CV为：0.9679469655549481\n",
      "第504轮QDA的CV为：0.9816562009419152，basic的CV为：0.970188383045526\n",
      "第504轮NuSVC的CV为：0.965557299843014，basic的CV为：0.965557299843014\n",
      "第505轮QDA的CV为：0.9854737985338039，basic的CV为：0.9729387274866503\n",
      "第505轮NuSVC的CV为：0.9716113072072888，basic的CV为：0.9583295622529941\n",
      "第506轮QDA的CV为：0.9729520430289206，basic的CV为：0.9677283095378896\n",
      "第506轮NuSVC的CV为：0.9708822618343612，basic的CV为：0.965609247838698\n",
      "第507轮QDA的CV为：0.984637245566004，basic的CV为：0.9771753362694917\n",
      "第507轮NuSVC的CV为：0.9730835614807761，basic的CV为：0.9709112010474943\n",
      "第508轮QDA的CV为：0.9542578646819629，basic的CV为：0.9467491669235625\n",
      "第508轮NuSVC的CV为：0.9412222369019833，basic的CV为：0.9350126979932767\n",
      "第509轮QDA的CV为：0.9729395823369998，basic的CV为：0.9681970349115256\n",
      "第509轮NuSVC的CV为：0.9644348796429141，basic的CV为：0.9635102821616451\n",
      "第510轮QDA的CV为：0.9618715673848754，basic的CV为：0.9514605588750075\n",
      "第510轮NuSVC的CV为：0.9497404792081597，basic的CV为：0.9423546985334056\n",
      "第511轮QDA的CV为：0.9771413140170286，basic的CV为：0.9628129368407675\n",
      "第511轮NuSVC的CV为：0.9604142838988435，basic的CV为：0.9535360274494853\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for i in tqdm_notebook(range(512)):\n",
    "    trn2 = trn_new[trn_new['wheezy-copper-turtle-magic']==i]\n",
    "    tst2 = tst[tst['wheezy-copper-turtle-magic']==i]\n",
    "    idx1 = trn2.index; idx2 = tst2.index\n",
    "    trn2.reset_index(drop=True,inplace=True)\n",
    "\n",
    "    data = pd.concat([pd.DataFrame(trn2[cols]), pd.DataFrame(tst2[cols])])\n",
    "    usecols = [var for var in cols if data[var].std()>2]\n",
    "    data2 = VarianceThreshold(threshold=2).fit_transform(data[cols])\n",
    "    #data2 = StandardScaler().fit_transform(data2)\n",
    "\n",
    "\n",
    "    trn3 = data2[:trn2.shape[0]]; tst3 = data2[trn2.shape[0]:]\n",
    "    skf = StratifiedKFold(n_splits=11, random_state=42)\n",
    "    oof_tmp_qda = np.zeros(trn2.shape[0])\n",
    "    oof_tmp_nusvc = np.zeros(trn2.shape[0])\n",
    "    for trn_index, tst_index in skf.split(trn2, trn2['target']):\n",
    "        clf_qda = QuadraticDiscriminantAnalysis(0.6)\n",
    "        clf_qda.fit(trn3[trn_index,:],trn2.loc[trn_index]['target'])\n",
    "        oof_qda[idx1[tst_index]] = clf_qda.predict_proba(trn3[tst_index,:])[:,1]\n",
    "        oof_tmp_qda[tst_index] = clf_qda.predict_proba(trn3[tst_index,:])[:,1]\n",
    "        preds_qda[idx2] += clf_qda.predict_proba(tst3)[:,1] / skf.n_splits\n",
    "        \n",
    "        clf_nusvc = NuSVC(probability=True, kernel='poly', degree=4, gamma='auto', random_state=4, nu=0.5, coef0=0.053)\n",
    "        clf_nusvc.fit(trn3[trn_index,:],trn2.loc[trn_index]['target'])\n",
    "        oof_nusvc[idx1[tst_index]] = clf_nusvc.predict_proba(trn3[tst_index,:])[:,1]\n",
    "        oof_tmp_nusvc[tst_index] = clf_nusvc.predict_proba(trn3[tst_index,:])[:,1]\n",
    "        preds_nusvc[idx2] += clf_nusvc.predict_proba(tst3)[:,1] / skf.n_splits\n",
    "    trn2['pred_qda'] = oof_tmp_qda\n",
    "    trn2['pred_nusvc'] = oof_tmp_nusvc\n",
    "    \n",
    "    auc_bench_qda = roc_auc_score(trn2.loc[trn2['target_old'].notnull()]['target_old'], trn2.loc[trn2['target_old'].notnull()]['pred_qda'])\n",
    "    auc_basic_qda = auc_bench_qda\n",
    "    \n",
    "    auc_bench_nusvc = roc_auc_score(trn2.loc[trn2['target_old'].notnull()]['target_old'], trn2.loc[trn2['target_old'].notnull()]['pred_nusvc'])\n",
    "    auc_basic_nusvc = auc_bench_nusvc\n",
    "\n",
    "    best_com = 0\n",
    "    for loop in range(6):\n",
    "        components = out_info[i][1].shape[0]\n",
    "        trn2['tgt'] = np.nan\n",
    "        gmm1 = GaussianMixture(n_components=components,max_iter = max_iter)#,precisions_init=pre_init[i][0])\n",
    "        trn2.loc[trn2['target']==1,'tgt'] = gmm1.fit_predict(trn2[trn2['target']==1][usecols].values)+components\n",
    "        gmm2 = GaussianMixture(n_components=components,max_iter = max_iter)#,precisions_init=pre_init[i][1])\n",
    "        trn2.loc[trn2['target']==0,'tgt'] = gmm2.fit_predict(trn2[trn2['target']==0][usecols].values)\n",
    "        trn3 = trn2[usecols].values          \n",
    "        if add_sample:\n",
    "            tmp = gmm1.sample(n_samples = sample_size) \n",
    "            trn2_pos = pd.DataFrame(tmp[0],columns=usecols)\n",
    "            trn2_pos['tgt'] = tmp[1]+components\n",
    "            trn2_pos['target'] = np.nan\n",
    "            tmp = gmm2.sample(n_samples = sample_size) \n",
    "            trn2_neg = pd.DataFrame(tmp[0],columns=usecols)\n",
    "            trn2_neg['tgt'] = tmp[1]\n",
    "            trn2_neg['target'] = np.nan\n",
    "            trn2_add = pd.concat([trn2,trn2_pos,trn2_neg]).sample(frac=1).reset_index(drop=True)\n",
    "        else:\n",
    "            trn2_add = trn2.copy()\n",
    "        trn3 = trn2_add[usecols].values\n",
    "        skf = StratifiedKFold(n_splits=11, random_state=42)\n",
    "        oof_tmp_qda = np.zeros(trn2_add.shape[0])\n",
    "        preds_tmp_qda = np.zeros(tst3.shape[0])\n",
    "        oof_tmp_nusvc = np.zeros(trn2_add.shape[0])\n",
    "        preds_tmp_nusvc = np.zeros(tst3.shape[0])\n",
    "        for trn_index, tst_index in skf.split(trn2_add, trn2_add['tgt']):\n",
    "            clf_qda = QuadraticDiscriminantAnalysis(0.6)\n",
    "            clf_nusvc = NuSVC(probability=True, kernel='poly', degree=4, gamma='auto', random_state=4, nu=0.5, coef0=0.053)\n",
    "            try:\n",
    "                clf_qda.fit(trn3[trn_index,:],trn2_add.loc[trn_index]['tgt'])\n",
    "                clf_nusvc.fit(trn3[trn_index,:],trn2_add.loc[trn_index]['tgt'])\n",
    "            except:\n",
    "                break\n",
    "            oof_tmp_qda[tst_index] = clf_qda.predict_proba(trn3[tst_index,:])[:,components:].sum(axis=1)\n",
    "            preds_tmp_qda += clf_qda.predict_proba(tst3)[:,components:].sum(axis=1) / skf.n_splits\n",
    "            oof_tmp_nusvc[tst_index] = clf_nusvc.predict_proba(trn3[tst_index,:])[:,components:].sum(axis=1)\n",
    "            preds_tmp_nusvc += clf_nusvc.predict_proba(tst3)[:,components:].sum(axis=1) / skf.n_splits\n",
    "        trn2_add['pred_qda'] = oof_tmp_qda\n",
    "        trn2_add['pred_nusvc'] = oof_tmp_nusvc\n",
    "        auc_tmp_qda = roc_auc_score(trn2_add[trn2_add['target_old'].notnull()]['target_old'],trn2_add[trn2_add['target_old'].notnull()]['pred_qda'])\n",
    "        auc_tmp_nusvc = roc_auc_score(trn2_add[trn2_add['target_old'].notnull()]['target_old'],trn2_add[trn2_add['target_old'].notnull()]['pred_nusvc'])\n",
    "        if auc_bench_qda < auc_tmp_qda:\n",
    "            del trn2['pred_qda']\n",
    "            auc_bench_qda = auc_tmp_qda\n",
    "            trn2 = trn2.merge(trn2_add[['id','pred_qda']],'left','id')\n",
    "            preds_qda[idx2] = preds_tmp_qda\n",
    "            oof_qda[idx1] = trn2['pred_qda'].values\n",
    "            #best_com = components\n",
    "        if auc_bench_nusvc < auc_tmp_nusvc:\n",
    "            del trn2['pred_nusvc']\n",
    "            auc_bench_nusvc = auc_tmp_nusvc\n",
    "            trn2 = trn2.merge(trn2_add[['id','pred_nusvc']],'left','id')\n",
    "            preds_nusvc[idx2] = preds_tmp_nusvc\n",
    "            oof_nusvc[idx1] = trn2['pred_nusvc'].values\n",
    "            #best_com = components\n",
    "    print(\"第{}轮QDA的CV为：{}，basic的CV为：{}\".format(i,auc_bench_qda,auc_basic_qda))\n",
    "    print(\"第{}轮NuSVC的CV为：{}，basic的CV为：{}\".format(i,auc_bench_nusvc,auc_basic_nusvc))\n",
    "trn_new['pred_qda'] = oof_qda\n",
    "trn_new['pred_nusvc'] = oof_nusvc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "auc_qda = roc_auc_score(trn_new.loc[trn_new['target_old'].notnull()]['target_old'], trn_new.loc[trn_new['target_old'].notnull()]['pred_qda'])\n",
    "auc_nusvc = roc_auc_score(trn_new.loc[trn_new['target_old'].notnull()]['target_old'], trn_new.loc[trn_new['target_old'].notnull()]['pred_nusvc'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9773307692427424\n",
      "0.9662022113829094\n"
     ]
    }
   ],
   "source": [
    "print(auc_qda)\n",
    "print(auc_nusvc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def find_the_threshold2(n,m,oof_qda,oof_nusvc,oof2,a,b):\n",
    "    for i in range(oof_qda.shape[0]):\n",
    "        if oof_qda[i]>n:\n",
    "            oof2[i] = oof_qda[i]\n",
    "        elif  oof_qda[i]<m:\n",
    "            oof2[i] = oof_qda[i]\n",
    "        else:\n",
    "            oof2[i] = oof_nusvc[i]*a+oof_qda[i]*b\n",
    "    return oof2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9774017865615675\n"
     ]
    }
   ],
   "source": [
    "oof2 = np.zeros(len(trn))\n",
    "print(roc_auc_score(trn_new.loc[trn_new['target_old'].notnull()]['target_old'],find_the_threshold2(0.99997,0.00007,\n",
    "                                                                                                   trn_new.loc[trn_new['target_old'].notnull()]['pred_qda'],\n",
    "                                                                                                   trn_new.loc[trn_new['target_old'].notnull()]['pred_nusvc'],\n",
    "                                                                                                   oof2,\n",
    "                                                                                                   0.5,0.5)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(preds_qda.shape[0]):\n",
    "    if preds_qda[i]>0.99997:\n",
    "        preds[i] = preds_qda[i]\n",
    "    elif preds_qda[i]<0.00007:\n",
    "        preds[i] = preds_qda[i]\n",
    "    else:\n",
    "        preds[i] = preds_nusvc[i]*0.5+preds_qda[i]*0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub['target'] = preds\n",
    "sub.to_csv('submission.csv',index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}