API tools faq
paste
Advertisement
Guest User

Untitled

a guest
Nov 22nd, 2019
151
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 26.14 KB | None | 0 0
raw download clone embed print report
  1. {
  2. "cells": [
  3. {
  4. "cell_type": "markdown",
  5. "metadata": {},
  6. "source": [
  7. "# Trabalho (traindata)"
  8. ]
  9. },
  10. {
  11. "cell_type": "markdown",
  12. "metadata": {},
  13. "source": [
  14. "#Grupo 7 | 1110299 Marta Ferreira | 1180167 Mariana Oliveira | 1180149 Ana Lima"
  15. ]
  16. },
  17. {
  18. "cell_type": "code",
  19. "execution_count": 1,
  20. "metadata": {},
  21. "outputs": [],
  22. "source": [
  23. "#Passos que temos que fazer: (Para retirar daqui depois)\n",
  24. "#\n",
  25. "#Carregar dados - e tratar os dados! temos de usar mais do que o que estamos a dar nas aulas\n",
  26. "#Split (treino/Teste)\n",
  27. "#criar standardscaler (standardizar, média 0 e desvio padrao 1)\n",
  28. "#treinar standardscaler\n",
  29. "#Transformar dados treino\n",
  30. "#treinar KNN\n",
  31. "#Transformar testset\n",
  32. "#Score/predict\n",
  33. "#Trabalho da Iris fizemos uma batota, porque normalizamos dados entes de os dividir (confirmar)"
  34. ]
  35. },
  36. {
  37. "cell_type": "code",
  38. "execution_count": 2,
  39. "metadata": {},
  40. "outputs": [],
  41. "source": [
  42. "import numpy as np\n",
  43. "import matplotlib.pyplot as plt\n",
  44. "import pandas as pd\n",
  45. "\n",
  46. "from sklearn import preprocessing as pp\n",
  47. "from sklearn.model_selection import train_test_split\n",
  48. "from sklearn.preprocessing import (\n",
  49. " StandardScaler,\n",
  50. " MinMaxScaler,\n",
  51. " RobustScaler)\n",
  52. "\n",
  53. "from sklearn.neighbors import KNeighborsClassifier\n",
  54. "from sklearn.pipeline import make_pipeline\n",
  55. "from sklearn.pipeline import Pipeline"
  56. ]
  57. },
  58. {
  59. "cell_type": "code",
  60. "execution_count": 3,
  61. "metadata": {},
  62. "outputs": [
  63. {
  64. "data": {
  65. "text/html": [
  66. "<div>\n",
  67. "<style scoped>\n",
  68. " .dataframe tbody tr th:only-of-type {\n",
  69. " vertical-align: middle;\n",
  70. " }\n",
  71. "\n",
  72. " .dataframe tbody tr th {\n",
  73. " vertical-align: top;\n",
  74. " }\n",
  75. "\n",
  76. " .dataframe thead th {\n",
  77. " text-align: right;\n",
  78. " }\n",
  79. "</style>\n",
  80. "<table border=\"1\" class=\"dataframe\">\n",
  81. " <thead>\n",
  82. " <tr style=\"text-align: right;\">\n",
  83. " <th></th>\n",
  84. " <th>0</th>\n",
  85. " <th>1</th>\n",
  86. " <th>2</th>\n",
  87. " <th>3</th>\n",
  88. " <th>4</th>\n",
  89. " <th>5</th>\n",
  90. " <th>6</th>\n",
  91. " <th>7</th>\n",
  92. " <th>8</th>\n",
  93. " <th>9</th>\n",
  94. " <th>10</th>\n",
  95. " <th>11</th>\n",
  96. " <th>12</th>\n",
  97. " <th>13</th>\n",
  98. " </tr>\n",
  99. " <tr>\n",
  100. " <th>Id</th>\n",
  101. " <th></th>\n",
  102. " <th></th>\n",
  103. " <th></th>\n",
  104. " <th></th>\n",
  105. " <th></th>\n",
  106. " <th></th>\n",
  107. " <th></th>\n",
  108. " <th></th>\n",
  109. " <th></th>\n",
  110. " <th></th>\n",
  111. " <th></th>\n",
  112. " <th></th>\n",
  113. " <th></th>\n",
  114. " <th></th>\n",
  115. " </tr>\n",
  116. " </thead>\n",
  117. " <tbody>\n",
  118. " <tr>\n",
  119. " <th>0</th>\n",
  120. " <td>63.0</td>\n",
  121. " <td>1.0</td>\n",
  122. " <td>1.0</td>\n",
  123. " <td>145.0</td>\n",
  124. " <td>233.0</td>\n",
  125. " <td>1.0</td>\n",
  126. " <td>2.0</td>\n",
  127. " <td>150.0</td>\n",
  128. " <td>0.0</td>\n",
  129. " <td>2.3</td>\n",
  130. " <td>3.0</td>\n",
  131. " <td>0.0</td>\n",
  132. " <td>6.0</td>\n",
  133. " <td>0</td>\n",
  134. " </tr>\n",
  135. " <tr>\n",
  136. " <th>1</th>\n",
  137. " <td>67.0</td>\n",
  138. " <td>1.0</td>\n",
  139. " <td>4.0</td>\n",
  140. " <td>160.0</td>\n",
  141. " <td>286.0</td>\n",
  142. " <td>0.0</td>\n",
  143. " <td>2.0</td>\n",
  144. " <td>108.0</td>\n",
  145. " <td>1.0</td>\n",
  146. " <td>1.5</td>\n",
  147. " <td>2.0</td>\n",
  148. " <td>3.0</td>\n",
  149. " <td>3.0</td>\n",
  150. " <td>2</td>\n",
  151. " </tr>\n",
  152. " <tr>\n",
  153. " <th>2</th>\n",
  154. " <td>67.0</td>\n",
  155. " <td>1.0</td>\n",
  156. " <td>4.0</td>\n",
  157. " <td>120.0</td>\n",
  158. " <td>229.0</td>\n",
  159. " <td>0.0</td>\n",
  160. " <td>2.0</td>\n",
  161. " <td>129.0</td>\n",
  162. " <td>1.0</td>\n",
  163. " <td>2.6</td>\n",
  164. " <td>2.0</td>\n",
  165. " <td>2.0</td>\n",
  166. " <td>7.0</td>\n",
  167. " <td>1</td>\n",
  168. " </tr>\n",
  169. " <tr>\n",
  170. " <th>3</th>\n",
  171. " <td>37.0</td>\n",
  172. " <td>1.0</td>\n",
  173. " <td>3.0</td>\n",
  174. " <td>130.0</td>\n",
  175. " <td>250.0</td>\n",
  176. " <td>0.0</td>\n",
  177. " <td>0.0</td>\n",
  178. " <td>187.0</td>\n",
  179. " <td>0.0</td>\n",
  180. " <td>3.5</td>\n",
  181. " <td>3.0</td>\n",
  182. " <td>0.0</td>\n",
  183. " <td>3.0</td>\n",
  184. " <td>0</td>\n",
  185. " </tr>\n",
  186. " <tr>\n",
  187. " <th>4</th>\n",
  188. " <td>41.0</td>\n",
  189. " <td>0.0</td>\n",
  190. " <td>2.0</td>\n",
  191. " <td>130.0</td>\n",
  192. " <td>204.0</td>\n",
  193. " <td>0.0</td>\n",
  194. " <td>2.0</td>\n",
  195. " <td>172.0</td>\n",
  196. " <td>0.0</td>\n",
  197. " <td>1.4</td>\n",
  198. " <td>1.0</td>\n",
  199. " <td>0.0</td>\n",
  200. " <td>3.0</td>\n",
  201. " <td>0</td>\n",
  202. " </tr>\n",
  203. " </tbody>\n",
  204. "</table>\n",
  205. "</div>"
  206. ],
  207. "text/plain": [
  208. " 0 1 2 3 4 5 6 7 8 9 10 11 12 13\n",
  209. "Id \n",
  210. "0 63.0 1.0 1.0 145.0 233.0 1.0 2.0 150.0 0.0 2.3 3.0 0.0 6.0 0\n",
  211. "1 67.0 1.0 4.0 160.0 286.0 0.0 2.0 108.0 1.0 1.5 2.0 3.0 3.0 2\n",
  212. "2 67.0 1.0 4.0 120.0 229.0 0.0 2.0 129.0 1.0 2.6 2.0 2.0 7.0 1\n",
  213. "3 37.0 1.0 3.0 130.0 250.0 0.0 0.0 187.0 0.0 3.5 3.0 0.0 3.0 0\n",
  214. "4 41.0 0.0 2.0 130.0 204.0 0.0 2.0 172.0 0.0 1.4 1.0 0.0 3.0 0"
  215. ]
  216. },
  217. "execution_count": 3,
  218. "metadata": {},
  219. "output_type": "execute_result"
  220. }
  221. ],
  222. "source": [
  223. "# Importação dos dados do ficheiro excel\n",
  224. "traindata = pd.read_csv(\"traindata.csv\", header=0, index_col=0) \n",
  225. "traindata.head()"
  226. ]
  227. },
  228. {
  229. "cell_type": "code",
  230. "execution_count": 4,
  231. "metadata": {},
  232. "outputs": [
  233. {
  234. "data": {
  235. "text/plain": [
  236. "(216, 14)"
  237. ]
  238. },
  239. "execution_count": 4,
  240. "metadata": {},
  241. "output_type": "execute_result"
  242. }
  243. ],
  244. "source": [
  245. "traindata.shape"
  246. ]
  247. },
  248. {
  249. "cell_type": "code",
  250. "execution_count": 5,
  251. "metadata": {},
  252. "outputs": [
  253. {
  254. "data": {
  255. "text/html": [
  256. "<div>\n",
  257. "<style scoped>\n",
  258. " .dataframe tbody tr th:only-of-type {\n",
  259. " vertical-align: middle;\n",
  260. " }\n",
  261. "\n",
  262. " .dataframe tbody tr th {\n",
  263. " vertical-align: top;\n",
  264. " }\n",
  265. "\n",
  266. " .dataframe thead th {\n",
  267. " text-align: right;\n",
  268. " }\n",
  269. "</style>\n",
  270. "<table border=\"1\" class=\"dataframe\">\n",
  271. " <thead>\n",
  272. " <tr style=\"text-align: right;\">\n",
  273. " <th></th>\n",
  274. " <th>0</th>\n",
  275. " <th>1</th>\n",
  276. " <th>2</th>\n",
  277. " <th>3</th>\n",
  278. " <th>4</th>\n",
  279. " <th>5</th>\n",
  280. " <th>6</th>\n",
  281. " <th>7</th>\n",
  282. " <th>8</th>\n",
  283. " <th>9</th>\n",
  284. " <th>10</th>\n",
  285. " <th>11</th>\n",
  286. " <th>12</th>\n",
  287. " <th>13</th>\n",
  288. " </tr>\n",
  289. " <tr>\n",
  290. " <th>Id</th>\n",
  291. " <th></th>\n",
  292. " <th></th>\n",
  293. " <th></th>\n",
  294. " <th></th>\n",
  295. " <th></th>\n",
  296. " <th></th>\n",
  297. " <th></th>\n",
  298. " <th></th>\n",
  299. " <th></th>\n",
  300. " <th></th>\n",
  301. " <th></th>\n",
  302. " <th></th>\n",
  303. " <th></th>\n",
  304. " <th></th>\n",
  305. " </tr>\n",
  306. " </thead>\n",
  307. " <tbody>\n",
  308. " <tr>\n",
  309. " <th>0</th>\n",
  310. " <td>63.0</td>\n",
  311. " <td>1.0</td>\n",
  312. " <td>1.0</td>\n",
  313. " <td>145.0</td>\n",
  314. " <td>233.0</td>\n",
  315. " <td>1.0</td>\n",
  316. " <td>2.0</td>\n",
  317. " <td>150.0</td>\n",
  318. " <td>0.0</td>\n",
  319. " <td>2.3</td>\n",
  320. " <td>3.0</td>\n",
  321. " <td>0.0</td>\n",
  322. " <td>6.0</td>\n",
  323. " <td>0</td>\n",
  324. " </tr>\n",
  325. " <tr>\n",
  326. " <th>1</th>\n",
  327. " <td>67.0</td>\n",
  328. " <td>1.0</td>\n",
  329. " <td>4.0</td>\n",
  330. " <td>160.0</td>\n",
  331. " <td>286.0</td>\n",
  332. " <td>0.0</td>\n",
  333. " <td>2.0</td>\n",
  334. " <td>108.0</td>\n",
  335. " <td>1.0</td>\n",
  336. " <td>1.5</td>\n",
  337. " <td>2.0</td>\n",
  338. " <td>3.0</td>\n",
  339. " <td>3.0</td>\n",
  340. " <td>2</td>\n",
  341. " </tr>\n",
  342. " <tr>\n",
  343. " <th>2</th>\n",
  344. " <td>67.0</td>\n",
  345. " <td>1.0</td>\n",
  346. " <td>4.0</td>\n",
  347. " <td>120.0</td>\n",
  348. " <td>229.0</td>\n",
  349. " <td>0.0</td>\n",
  350. " <td>2.0</td>\n",
  351. " <td>129.0</td>\n",
  352. " <td>1.0</td>\n",
  353. " <td>2.6</td>\n",
  354. " <td>2.0</td>\n",
  355. " <td>2.0</td>\n",
  356. " <td>7.0</td>\n",
  357. " <td>1</td>\n",
  358. " </tr>\n",
  359. " <tr>\n",
  360. " <th>3</th>\n",
  361. " <td>37.0</td>\n",
  362. " <td>1.0</td>\n",
  363. " <td>3.0</td>\n",
  364. " <td>130.0</td>\n",
  365. " <td>250.0</td>\n",
  366. " <td>0.0</td>\n",
  367. " <td>0.0</td>\n",
  368. " <td>187.0</td>\n",
  369. " <td>0.0</td>\n",
  370. " <td>3.5</td>\n",
  371. " <td>3.0</td>\n",
  372. " <td>0.0</td>\n",
  373. " <td>3.0</td>\n",
  374. " <td>0</td>\n",
  375. " </tr>\n",
  376. " <tr>\n",
  377. " <th>4</th>\n",
  378. " <td>41.0</td>\n",
  379. " <td>0.0</td>\n",
  380. " <td>2.0</td>\n",
  381. " <td>130.0</td>\n",
  382. " <td>204.0</td>\n",
  383. " <td>0.0</td>\n",
  384. " <td>2.0</td>\n",
  385. " <td>172.0</td>\n",
  386. " <td>0.0</td>\n",
  387. " <td>1.4</td>\n",
  388. " <td>1.0</td>\n",
  389. " <td>0.0</td>\n",
  390. " <td>3.0</td>\n",
  391. " <td>0</td>\n",
  392. " </tr>\n",
  393. " </tbody>\n",
  394. "</table>\n",
  395. "</div>"
  396. ],
  397. "text/plain": [
  398. " 0 1 2 3 4 5 6 7 8 9 10 11 12 13\n",
  399. "Id \n",
  400. "0 63.0 1.0 1.0 145.0 233.0 1.0 2.0 150.0 0.0 2.3 3.0 0.0 6.0 0\n",
  401. "1 67.0 1.0 4.0 160.0 286.0 0.0 2.0 108.0 1.0 1.5 2.0 3.0 3.0 2\n",
  402. "2 67.0 1.0 4.0 120.0 229.0 0.0 2.0 129.0 1.0 2.6 2.0 2.0 7.0 1\n",
  403. "3 37.0 1.0 3.0 130.0 250.0 0.0 0.0 187.0 0.0 3.5 3.0 0.0 3.0 0\n",
  404. "4 41.0 0.0 2.0 130.0 204.0 0.0 2.0 172.0 0.0 1.4 1.0 0.0 3.0 0"
  405. ]
  406. },
  407. "execution_count": 5,
  408. "metadata": {},
  409. "output_type": "execute_result"
  410. }
  411. ],
  412. "source": [
  413. "df=pd.DataFrame(traindata)\n",
  414. "df.head()"
  415. ]
  416. },
  417. {
  418. "cell_type": "code",
  419. "execution_count": 6,
  420. "metadata": {},
  421. "outputs": [],
  422. "source": [
  423. "traindata.replace({'?':np.nan}, inplace=True)"
  424. ]
  425. },
  426. {
  427. "cell_type": "code",
  428. "execution_count": 7,
  429. "metadata": {},
  430. "outputs": [],
  431. "source": [
  432. "traindata.dropna(axis=0, how='any', thresh=None, subset=None, inplace=True)"
  433. ]
  434. },
  435. {
  436. "cell_type": "code",
  437. "execution_count": 8,
  438. "metadata": {},
  439. "outputs": [],
  440. "source": [
  441. "trainset, testset=train_test_split(traindata)"
  442. ]
  443. },
  444. {
  445. "cell_type": "code",
  446. "execution_count": 9,
  447. "metadata": {},
  448. "outputs": [
  449. {
  450. "data": {
  451. "text/plain": [
  452. "0 93\n",
  453. "1 26\n",
  454. "3 19\n",
  455. "2 19\n",
  456. "4 2\n",
  457. "Name: 13, dtype: int64"
  458. ]
  459. },
  460. "execution_count": 9,
  461. "metadata": {},
  462. "output_type": "execute_result"
  463. }
  464. ],
  465. "source": [
  466. "#Contar os valores trainset. Fazer para a coluna 13, neste caso.\n",
  467. "trainset[\"13\"].value_counts()"
  468. ]
  469. },
  470. {
  471. "cell_type": "code",
  472. "execution_count": 10,
  473. "metadata": {},
  474. "outputs": [
  475. {
  476. "data": {
  477. "text/plain": [
  478. "0 22\n",
  479. "1 12\n",
  480. "4 7\n",
  481. "3 6\n",
  482. "2 6\n",
  483. "Name: 13, dtype: int64"
  484. ]
  485. },
  486. "execution_count": 10,
  487. "metadata": {},
  488. "output_type": "execute_result"
  489. }
  490. ],
  491. "source": [
  492. "#(No testset vamos ter que ver se os valores estão balanceados)\n",
  493. "testset[\"13\"].value_counts()"
  494. ]
  495. },
  496. {
  497. "cell_type": "markdown",
  498. "metadata": {},
  499. "source": [
  500. "# Função Standardscaler (dados já divididos)"
  501. ]
  502. },
  503. {
  504. "cell_type": "code",
  505. "execution_count": 11,
  506. "metadata": {},
  507. "outputs": [],
  508. "source": [
  509. "# Função Standardscaler - Transformar os dados de forma a que sua distribuição tem um valor \n",
  510. "# médio igual a 0 e um desvio padrão de 1."
  511. ]
  512. },
  513. {
  514. "cell_type": "code",
  515. "execution_count": 12,
  516. "metadata": {},
  517. "outputs": [
  518. {
  519. "data": {
  520. "text/plain": [
  521. "array([[-1.51112929, 0.73735291, -0.15951158, -0.150565 , -0.56516362,\n",
  522. " -0.3683942 , 1.04842767, 0.77092891, -0.63802794, 0.80364581,\n",
  523. " 0.61441966, -0.67320658, -0.80632771],\n",
  524. " [-1.7414932 , -1.35620268, -0.15951158, 0.31769583, -0.44273717,\n",
  525. " -0.3683942 , -0.97217839, 0.06973365, -0.63802794, -0.90921859,\n",
  526. " 0.61441966, -0.67320658, -0.80632771],\n",
  527. " [ 1.94432934, -1.35620268, 0.89725263, -1.20415186, -1.89145018,\n",
  528. " -0.3683942 , -0.97217839, -1.11353336, -0.63802794, 0.46107293,\n",
  529. " 0.61441966, -0.67320658, -0.80632771],\n",
  530. " [ 0.33178198, 0.73735291, 0.89725263, 1.02008707, 0.69990971,\n",
  531. " -0.3683942 , 1.04842767, -1.68325451, 1.56732949, -0.39535927,\n",
  532. " 0.61441966, 0.50305546, 0.75716139],\n",
  533. " [-0.93521952, -1.35620268, 0.89725263, 0.31769583, 0.02656423,\n",
  534. " -0.3683942 , 1.04842767, 0.06973365, 1.56732949, -0.90921859,\n",
  535. " 0.61441966, -0.67320658, -0.80632771],\n",
  536. " [-0.4744917 , -1.35620268, -1.21627579, -0.73589103, 0.04696864,\n",
  537. " -0.3683942 , -0.97217839, 0.50798069, -0.63802794, 0.03285683,\n",
  538. " -0.91202919, -0.67320658, -0.80632771],\n",
  539. " [ 0.33178198, 0.73735291, -1.21627579, 1.25421748, -0.19788426,\n",
  540. " -0.3683942 , 1.04842767, 0.5956301 , -0.63802794, -0.90921859,\n",
  541. " -0.91202919, 0.50305546, -0.80632771],\n",
  542. " [-1.9718571 , 0.73735291, -0.15951158, -0.150565 , 0.16939509,\n",
  543. " -0.3683942 , -0.97217839, 1.60359829, -0.63802794, 2.08829412,\n",
  544. " 2.14086851, -0.67320658, -0.80632771],\n",
  545. " [-0.58967365, -1.35620268, -1.21627579, 0.08356541, 0.59788767,\n",
  546. " -0.3683942 , -0.97217839, 0.50798069, -0.63802794, -0.90921859,\n",
  547. " 0.61441966, -0.67320658, -0.80632771],\n",
  548. " [ 0.44696394, 0.73735291, 0.89725263, 1.02008707, 0.57748326,\n",
  549. " -0.3683942 , 1.04842767, -1.72707921, 1.56732949, -0.22407283,\n",
  550. " -0.91202919, -0.67320658, 1.27832442],\n",
  551. " [ 0.67732784, 0.73735291, 0.89725263, -0.150565 , 0.23060832,\n",
  552. " -0.3683942 , -0.97217839, -0.28086398, 1.56732949, 0.28978649,\n",
  553. " -0.91202919, 0.50305546, 1.27832442],\n",
  554. " [-1.05040147, 0.73735291, 0.89725263, -1.67241268, -0.68759007,\n",
  555. " -0.3683942 , 1.04842767, -0.10556517, 1.56732949, 1.66007802,\n",
  556. " 0.61441966, -0.67320658, -0.80632771]])"
  557. ]
  558. },
  559. "execution_count": 12,
  560. "metadata": {},
  561. "output_type": "execute_result"
  562. }
  563. ],
  564. "source": [
  565. "scaler=StandardScaler()\n",
  566. "scaler.fit(trainset.iloc[:,:-1])\n",
  567. "trainX=scaler.transform(trainset.iloc[:,:-1])#Fizemos o fit e o transform juntos, mas podia ser separados\n",
  568. "trainX[:12] #Queremos normalizar tudo menos a última coluna"
  569. ]
  570. },
  571. {
  572. "cell_type": "code",
  573. "execution_count": 13,
  574. "metadata": {},
  575. "outputs": [],
  576. "source": [
  577. "trainY=trainset[\"13\"] #isto é o que eu quero prever"
  578. ]
  579. },
  580. {
  581. "cell_type": "markdown",
  582. "metadata": {},
  583. "source": [
  584. "# KNN (Derterminação do vizinho mais próximo)"
  585. ]
  586. },
  587. {
  588. "cell_type": "code",
  589. "execution_count": 27,
  590. "metadata": {},
  591. "outputs": [
  592. {
  593. "data": {
  594. "text/plain": [
  595. "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n",
  596. " metric_params=None, n_jobs=None, n_neighbors=3, p=2,\n",
  597. " weights='uniform')"
  598. ]
  599. },
  600. "execution_count": 27,
  601. "metadata": {},
  602. "output_type": "execute_result"
  603. }
  604. ],
  605. "source": [
  606. "knnmodel=KNeighborsClassifier(n_neighbors=3)#ver os vizinhos porque são 3??\n",
  607. "knnmodel.fit(trainX, trainY) #temos de fazer o fit"
  608. ]
  609. },
  610. {
  611. "cell_type": "code",
  612. "execution_count": 15,
  613. "metadata": {},
  614. "outputs": [],
  615. "source": [
  616. "testX=scaler.transform(testset.iloc[:,:-1])\n",
  617. "testY=testset[\"13\"]"
  618. ]
  619. },
  620. {
  621. "cell_type": "code",
  622. "execution_count": 16,
  623. "metadata": {},
  624. "outputs": [
  625. {
  626. "data": {
  627. "text/plain": [
  628. "0.4716981132075472"
  629. ]
  630. },
  631. "execution_count": 16,
  632. "metadata": {},
  633. "output_type": "execute_result"
  634. }
  635. ],
  636. "source": [
  637. "knnmodel.score(testX,testY) #normalização feita com base nos dados de treino. \n",
  638. "#O testY é o que eu quero prever. \n",
  639. "#Está a analisar a qualidade do teu resultado, em cada 100 acertei 94"
  640. ]
  641. },
  642. {
  643. "cell_type": "code",
  644. "execution_count": 17,
  645. "metadata": {},
  646. "outputs": [],
  647. "source": [
  648. "#knnmodel.predict(testset.iloc[:,:-1]) (ver o que colocar aqui)"
  649. ]
  650. },
  651. {
  652. "cell_type": "code",
  653. "execution_count": 18,
  654. "metadata": {},
  655. "outputs": [],
  656. "source": [
  657. "ppl=make_pipeline(StandardScaler(),KNeighborsClassifier(n_neighbors=1)) #ver isto dos vizinhos"
  658. ]
  659. },
  660. {
  661. "cell_type": "code",
  662. "execution_count": 19,
  663. "metadata": {},
  664. "outputs": [],
  665. "source": [
  666. "ppl2=Pipeline(steps=[(\"scaler\",StandardScaler()),(\"3NN\",KNeighborsClassifier(n_neighbors=3))])\n",
  667. "#outra forma de se resolver o ppl2. ver a questão dos vizinhos como se faz"
  668. ]
  669. },
  670. {
  671. "cell_type": "code",
  672. "execution_count": 20,
  673. "metadata": {},
  674. "outputs": [
  675. {
  676. "data": {
  677. "text/plain": [
  678. "Pipeline(memory=None,\n",
  679. " steps=[('standardscaler',\n",
  680. " StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
  681. " ('kneighborsclassifier',\n",
  682. " KNeighborsClassifier(algorithm='auto', leaf_size=30,\n",
  683. " metric='minkowski', metric_params=None,\n",
  684. " n_jobs=None, n_neighbors=1, p=2,\n",
  685. " weights='uniform'))],\n",
  686. " verbose=False)"
  687. ]
  688. },
  689. "execution_count": 20,
  690. "metadata": {},
  691. "output_type": "execute_result"
  692. }
  693. ],
  694. "source": [
  695. "ppl.fit(trainX,trainY) \n",
  696. "#o pipeline é uma linha de transformação \n",
  697. "#- em vez da standardScaler - faz o fit transform para varios modelos?"
  698. ]
  699. },
  700. {
  701. "cell_type": "code",
  702. "execution_count": 21,
  703. "metadata": {},
  704. "outputs": [
  705. {
  706. "data": {
  707. "text/plain": [
  708. "0.5283018867924528"
  709. ]
  710. },
  711. "execution_count": 21,
  712. "metadata": {},
  713. "output_type": "execute_result"
  714. }
  715. ],
  716. "source": [
  717. "ppl.score(testX,testY)"
  718. ]
  719. },
  720. {
  721. "cell_type": "code",
  722. "execution_count": 22,
  723. "metadata": {},
  724. "outputs": [
  725. {
  726. "data": {
  727. "text/plain": [
  728. "Pipeline(memory=None,\n",
  729. " steps=[('scaler',\n",
  730. " StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
  731. " ('3NN',\n",
  732. " KNeighborsClassifier(algorithm='auto', leaf_size=30,\n",
  733. " metric='minkowski', metric_params=None,\n",
  734. " n_jobs=None, n_neighbors=3, p=2,\n",
  735. " weights='uniform'))],\n",
  736. " verbose=False)"
  737. ]
  738. },
  739. "execution_count": 22,
  740. "metadata": {},
  741. "output_type": "execute_result"
  742. }
  743. ],
  744. "source": [
  745. "ppl2.fit(trainX,trainY)"
  746. ]
  747. },
  748. {
  749. "cell_type": "code",
  750. "execution_count": 23,
  751. "metadata": {},
  752. "outputs": [
  753. {
  754. "data": {
  755. "text/plain": [
  756. "0.49056603773584906"
  757. ]
  758. },
  759. "execution_count": 23,
  760. "metadata": {},
  761. "output_type": "execute_result"
  762. }
  763. ],
  764. "source": [
  765. "ppl2.score(testX,testY)"
  766. ]
  767. },
  768. {
  769. "cell_type": "markdown",
  770. "metadata": {},
  771. "source": [
  772. "Quais as previsões erradas? (Vamos ver onde erramos)"
  773. ]
  774. },
  775. {
  776. "cell_type": "code",
  777. "execution_count": 24,
  778. "metadata": {},
  779. "outputs": [
  780. {
  781. "data": {
  782. "text/plain": [
  783. "array([0, 3, 0, 3, 2], dtype=int64)"
  784. ]
  785. },
  786. "execution_count": 24,
  787. "metadata": {},
  788. "output_type": "execute_result"
  789. }
  790. ],
  791. "source": [
  792. "preds=ppl.predict(testX)\n",
  793. "preds[:5]"
  794. ]
  795. },
  796. {
  797. "cell_type": "code",
  798. "execution_count": 25,
  799. "metadata": {},
  800. "outputs": [
  801. {
  802. "data": {
  803. "text/plain": [
  804. "Id\n",
  805. "30 False\n",
  806. "40 True\n",
  807. "0 False\n",
  808. "213 False\n",
  809. "174 True\n",
  810. "Name: 13, dtype: bool"
  811. ]
  812. },
  813. "execution_count": 25,
  814. "metadata": {},
  815. "output_type": "execute_result"
  816. }
  817. ],
  818. "source": [
  819. "errado=(preds !=testY)\n",
  820. "errado.head()\n",
  821. "#Para identificarmos os objetos em que erramos a previsão\n",
  822. "#True é porque está errado e false é onde está certo"
  823. ]
  824. },
  825. {
  826. "cell_type": "code",
  827. "execution_count": 26,
  828. "metadata": {},
  829. "outputs": [
  830. {
  831. "name": "stderr",
  832. "output_type": "stream",
  833. "text": [
  834. "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
  835. "A value is trying to be set on a copy of a slice from a DataFrame\n",
  836. "\n",
  837. "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
  838. " \n",
  839. "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\pandas\\core\\generic.py:8682: SettingWithCopyWarning: \n",
  840. "A value is trying to be set on a copy of a slice from a DataFrame\n",
  841. "\n",
  842. "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
  843. " self._update_inplace(new_data)\n",
  844. "C:\\ProgramData\\Anaconda3\\lib\\site-packages\\IPython\\core\\interactiveshell.py:3325: SettingWithCopyWarning: \n",
  845. "A value is trying to be set on a copy of a slice from a DataFrame\n",
  846. "\n",
  847. "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
  848. " exec(code_obj, self.user_global_ns, self.user_ns)\n"
  849. ]
  850. },
  851. {
  852. "data": {
  853. "text/plain": [
  854. "Id\n",
  855. "30 0\n",
  856. "40 FAIL\n",
  857. "0 0\n",
  858. "213 3\n",
  859. "174 FAIL\n",
  860. "298 1\n",
  861. "280 2\n",
  862. "19 0\n",
  863. "177 1\n",
  864. "240 0\n",
  865. "Name: 13, dtype: object"
  866. ]
  867. },
  868. "execution_count": 26,
  869. "metadata": {},
  870. "output_type": "execute_result"
  871. }
  872. ],
  873. "source": [
  874. "colorseries=testY\n",
  875. "colorseries[errado==True]=\"FAIL\"\n",
  876. "colorseries.head(10)"
  877. ]
  878. },
  879. {
  880. "cell_type": "code",
  881. "execution_count": null,
  882. "metadata": {},
  883. "outputs": [],
  884. "source": []
  885. }
  886. ],
  887. "metadata": {
  888. "kernelspec": {
  889. "display_name": "Python 3",
  890. "language": "python",
  891. "name": "python3"
  892. },
  893. "language_info": {
  894. "codemirror_mode": {
  895. "name": "ipython",
  896. "version": 3
  897. },
  898. "file_extension": ".py",
  899. "mimetype": "text/x-python",
  900. "name": "python",
  901. "nbconvert_exporter": "python",
  902. "pygments_lexer": "ipython3",
  903. "version": "3.7.3"
  904. }
  905. },
  906. "nbformat": 4,
  907. "nbformat_minor": 2
  908. }
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!