API tools faq
paste
Advertisement
11eimilia11

main

11eimilia11
Feb 25th, 2019
148
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Python 94.87 KB | None | 0 0
raw download clone embed print report
  1. import pandas as pd
  2. import cv2
  3. import numpy as np
  4. from Lista02 import FuncoesML as fun
  5. from sklearn.model_selection import KFold
  6. from sklearn.neighbors import KNeighborsClassifier
  7. from sklearn.tree import DecisionTreeClassifier
  8. from sklearn.naive_bayes import GaussianNB
  9. from sklearn.svm import SVC
  10. import time
  11. from sklearn.linear_model import LogisticRegression
  12. from sklearn.neural_network import MLPClassifier
  13. from sklearn.ensemble import RandomForestClassifier
  14. from sklearn.metrics import recall_score
  15. from sklearn.metrics import precision_score
  16. from sklearn.metrics import f1_score
  17.  
  18. #lendo a imagem
  19. print('comecou load images')
  20. squares = []
  21. squares = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/squares/*.jpg', squares)
  22.  
  23. circles = []
  24. circles = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/circles/*.jpg',circles)
  25.  
  26. triangles = []
  27. triangles = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/triangles/*.jpg', triangles)
  28.  
  29. ellipses = []
  30. ellipses = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/ellipses/*.jpg', ellipses)
  31.  
  32. trapezia = []
  33. trapezia = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/trapezia/*.jpg', trapezia)
  34.  
  35. rectangles = []
  36. rectangles = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/rectangles/*.jpg', rectangles)
  37.  
  38. rhombuses = []
  39. rhombuses = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/rhombuses/*.jpg', rhombuses)
  40.  
  41. lines = []
  42. lines = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/lines/*.jpg', lines)
  43.  
  44. hexagons = []
  45. hexagons = fun.loadFiles('C:/Users/Auricelia/Desktop/DataSetsML/New_shapes_dataset/hexagons/*.jpg', hexagons)
  46.  
  47. print('terminou load images')
  48.  
  49. # Selecionando aleatoriamente 72 imagens de cada classe
  50.  
  51. squares_selec, squares_naoselec = fun.seleciona_imagens(squares,72)
  52. circles_selec, circles_naoselec = fun.seleciona_imagens(circles,72)
  53. triangles_selec, triangles_naoselec = fun.seleciona_imagens(triangles,72)
  54. ellipses_selec, ellipses_naoselec = fun.seleciona_imagens(ellipses,72)
  55. trapezia_selec, trapezia_naoselec = fun.seleciona_imagens(trapezia,72)
  56. rectangles_selec, rectangles_naoselec = fun.seleciona_imagens(rectangles,72)
  57. rhombuses_selec, rhombuses_naoselec = fun.seleciona_imagens(rhombuses,72)
  58. lines_selec, lines_naoselec = fun.seleciona_imagens(lines,72)
  59. hexagons_selec, hexagons_naoselec = fun.seleciona_imagens(hexagons,72)
  60.  
  61. #Salvando em pastas diferentes as imagens para seleção de características e as de teste
  62.  
  63. fun.save_images(squares_selec,'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomSquares/')
  64. fun.save_images(squares_naoselec,'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Squares/')
  65.  
  66. fun.save_images(circles_selec,'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomCircles/')
  67. fun.save_images(circles_naoselec,'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Circles/')
  68.  
  69. fun.save_images(triangles_selec,'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomTriangles/')
  70. fun.save_images(triangles_naoselec,'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Triangles/')
  71.  
  72. fun.save_images(ellipses_selec, 'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomEllipses/')
  73. fun.save_images(ellipses_naoselec,'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Ellipses/')
  74.  
  75. fun.save_images(trapezia_selec,'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomTrapezia/')
  76. fun.save_images(trapezia_naoselec,'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Trapezia/')
  77.  
  78. fun.save_images(rectangles_selec,'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomRectangles/')
  79. fun.save_images(rectangles_naoselec, 'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Rectangles/')
  80.  
  81. fun.save_images(rhombuses_selec,'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomRhombuses/')
  82. fun.save_images(rhombuses_naoselec, 'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Rhombuses/')
  83.  
  84. fun.save_images(lines_selec, 'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomLines/')
  85. fun.save_images(lines_naoselec, 'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Lines/')
  86.  
  87. fun.save_images(hexagons_selec, 'C:/Users/Auricelia/Desktop/DataSetsML/Random/RandomHexagons/')
  88. fun.save_images(hexagons_naoselec, 'C:/Users/Auricelia/Desktop/DataSetsML/Not_Random/Hexagons/')
  89.  
  90.  
  91. # PRE PROCESSING
  92.  
  93. #criando cópias de cada uma das pastas para redimensionar as imagens
  94. #quadrados
  95.  
  96. squares16_s = squares_selec.copy()
  97. squares16_n = squares_naoselec.copy()
  98.  
  99. squares32_s = squares_selec.copy()
  100. squares32_n = squares_naoselec.copy()
  101.  
  102. squares64_s = squares_selec.copy()
  103. squares64_n = squares_naoselec.copy()
  104.  
  105. squares128_s = squares_selec.copy()
  106. squares128_n = squares_naoselec.copy()
  107.  
  108. squares16_s = fun.resizeImages(squares16_s,16,16)
  109. squares16_n = fun.resizeImages(squares16_n,16,16)
  110.  
  111. squares32_s = fun.resizeImages(squares32_s,32,32)
  112. squares32_n = fun.resizeImages(squares32_n,32,32)
  113.  
  114. squares64_s = fun.resizeImages(squares64_s,64,64)
  115. squares64_n = fun.resizeImages(squares64_n,64,64)
  116.  
  117. squares128_s = fun.resizeImages(squares128_s,128,128)
  118. squares128_n = fun.resizeImages(squares128_n,128,128)
  119.  
  120. #círculos
  121. circles16_s = circles_selec.copy()
  122. circles16_n = circles_naoselec.copy()
  123.  
  124. circles32_s = circles_selec.copy()
  125. circles32_n = circles_naoselec.copy()
  126.  
  127. circles64_s = circles_selec.copy()
  128. circles64_n = circles_naoselec.copy()
  129.  
  130. circles128_s = circles_selec.copy()
  131. circles128_n = circles_naoselec.copy()
  132.  
  133. circles16_s = fun.resizeImages(circles16_s,16,16)
  134. circles16_n = fun.resizeImages(circles16_n,16,16)
  135.  
  136. circles32_s = fun.resizeImages(circles32_s,32,32)
  137. circles32_n = fun.resizeImages(circles32_n,32,32)
  138.  
  139. circles64_s = fun.resizeImages(circles64_s,64,64)
  140. circles64_n = fun.resizeImages(circles64_n,64,64)
  141. circles128_s = fun.resizeImages(circles128_s,128,128)
  142. circles128_n = fun.resizeImages(circles128_n,128,128)
  143.  
  144. #elipses
  145. ellipsis16_s = ellipses_selec.copy()
  146. ellipsis16_n = ellipses_naoselec.copy()
  147.  
  148. ellipsis32_s = ellipses_selec.copy()
  149. ellipsis32_n = ellipses_naoselec.copy()
  150.  
  151. ellipsis64_s = ellipses_selec.copy()
  152. ellipsis64_n = ellipses_naoselec.copy()
  153.  
  154. ellipsis128_s = ellipses_selec.copy()
  155. ellipsis128_n = ellipses_naoselec.copy()
  156.  
  157. ellipsis16_s = fun.resizeImages(ellipsis16_s,16,16)
  158. ellipsis16_n = fun.resizeImages(ellipsis16_n,16,16)
  159.  
  160. ellipsis32_s = fun.resizeImages(ellipsis32_s,32,32)
  161. ellipsis32_n = fun.resizeImages(ellipsis32_n,32,32)
  162.  
  163. ellipsis64_s = fun.resizeImages(ellipsis64_s,64,64)
  164. ellipsis64_n = fun.resizeImages(ellipsis64_n,64,64)
  165.  
  166. ellipsis128_s = fun.resizeImages(ellipsis128_s,128,128)
  167. ellipsis128_n = fun.resizeImages(ellipsis128_n,128,128)
  168.  
  169. #hexágonos
  170. hexagons16_s = hexagons_selec.copy()
  171. hexagons16_n = hexagons_naoselec.copy()
  172. hexagons32_s = hexagons_selec.copy()
  173. hexagons32_n = hexagons_naoselec.copy()
  174. hexagons64_s = hexagons_selec.copy()
  175. hexagons64_n = hexagons_naoselec.copy()
  176. hexagons128_s = hexagons_selec.copy()
  177. hexagons128_n = hexagons_naoselec.copy()
  178.  
  179. hexagons16_s = fun.resizeImages(hexagons16_s,16,16)
  180. hexagons16_n = fun.resizeImages(hexagons16_n,16,16)
  181. hexagons32_s = fun.resizeImages(hexagons32_s,32,32)
  182. hexagons32_n = fun.resizeImages(hexagons32_n,32,32)
  183. hexagons64_s = fun.resizeImages(hexagons64_s,64,64)
  184. hexagons64_n = fun.resizeImages(hexagons64_n,64,64)
  185. hexagons128_s = fun.resizeImages(hexagons128_s,128,128)
  186. hexagons128_n = fun.resizeImages(hexagons128_n,128,128)
  187.  
  188. #linhas
  189. lines16_s = lines_selec.copy()
  190. lines16_n = lines_naoselec.copy()
  191. lines32_s = lines_selec.copy()
  192. lines32_n = lines_naoselec.copy()
  193. lines64_s = lines_selec.copy()
  194. lines64_n = lines_naoselec.copy()
  195. lines128_s = lines_selec.copy()
  196. lines128_n = lines_naoselec.copy()
  197.  
  198. lines16_s = fun.resizeImages(lines16_s,16,16)
  199. lines16_n = fun.resizeImages(lines16_n,16,16)
  200. lines32_s = fun.resizeImages(lines32_s,32,32)
  201. lines32_n = fun.resizeImages(lines32_n,32,32)
  202. lines64_s = fun.resizeImages(lines64_s,64,64)
  203. lines64_n = fun.resizeImages(lines64_n,64,64)
  204. lines128_s = fun.resizeImages(lines128_s,128,128)
  205. lines128_n = fun.resizeImages(lines128_n,128,128)
  206.  
  207. #retângulos
  208. rectangles16_s = rectangles_selec.copy()
  209. rectangles16_n = rectangles_naoselec.copy()
  210. rectangles32_s = rectangles_selec.copy()
  211. rectangles32_n = rectangles_naoselec.copy()
  212. rectangles64_s = rectangles_selec.copy()
  213. rectangles64_n = rectangles_naoselec.copy()
  214. rectangles128_s = rectangles_selec.copy()
  215. rectangles128_n = rectangles_naoselec.copy()
  216.  
  217. rectangles16_s = fun.resizeImages(rectangles16_s,16,16)
  218. rectangles16_n = fun.resizeImages(rectangles16_n,16,16)
  219. rectangles32_s = fun.resizeImages(rectangles32_s,32,32)
  220. rectangles32_n = fun.resizeImages(rectangles32_n,32,32)
  221. rectangles64_s = fun.resizeImages(rectangles64_s,64,64)
  222. rectangles64_n = fun.resizeImages(rectangles64_n,64,64)
  223. rectangles128_s = fun.resizeImages(rectangles128_s,128,128)
  224. rectangles128_n = fun.resizeImages(rectangles128_n,128,128)
  225.  
  226. #losangos
  227. rhombuses16_s = rhombuses_selec.copy()
  228. rhombuses16_n = rhombuses_naoselec.copy()
  229. rhombuses32_s = rhombuses_selec.copy()
  230. rhombuses32_n = rhombuses_naoselec.copy()
  231. rhombuses64_s = rhombuses_selec.copy()
  232. rhombuses64_n = rhombuses_naoselec.copy()
  233. rhombuses128_s = rhombuses_selec.copy()
  234. rhombuses128_n = rhombuses_naoselec.copy()
  235.  
  236. rhombuses16_s = fun.resizeImages(rhombuses16_s,16,16)
  237. rhombuses16_n = fun.resizeImages(rhombuses16_n,16,16)
  238. rhombuses32_s = fun.resizeImages(rhombuses32_s,32,32)
  239. rhombuses32_n = fun.resizeImages(rhombuses32_n,32,32)
  240. rhombuses64_s = fun.resizeImages(rhombuses64_s,64,64)
  241. rhombuses64_n = fun.resizeImages(rhombuses64_n,64,64)
  242. rhombuses128_s = fun.resizeImages(rhombuses128_s,128,128)
  243. rhombuses128_n = fun.resizeImages(rhombuses128_n,128,128)
  244.  
  245. #trapézios
  246. trapezia16_s = trapezia_selec.copy()
  247. trapezia16_n = trapezia_naoselec.copy()
  248. trapezia32_s = trapezia_selec.copy()
  249. trapezia32_n = trapezia_naoselec.copy()
  250. trapezia64_s = trapezia_selec.copy()
  251. trapezia64_n = trapezia_naoselec.copy()
  252. trapezia128_s = trapezia_selec.copy()
  253. trapezia128_n = trapezia_naoselec.copy()
  254.  
  255. trapezia16_s = fun.resizeImages(trapezia16_s,16,16)
  256. trapezia16_n = fun.resizeImages(trapezia16_n,16,16)
  257. trapezia32_s = fun.resizeImages(trapezia32_s,32,32)
  258. trapezia32_n = fun.resizeImages(trapezia32_n,32,32)
  259. trapezia64_s = fun.resizeImages(trapezia64_s,64,64)
  260. trapezia64_n = fun.resizeImages(trapezia64_n,64,64)
  261. trapezia128_s = fun.resizeImages(trapezia128_s,128,128)
  262. trapezia128_n = fun.resizeImages(trapezia128_n,128,128)
  263.  
  264. #triângulos
  265. triangles16_s = triangles_selec.copy()
  266. triangles16_n = triangles_naoselec.copy()
  267. triangles32_s = triangles_selec.copy()
  268. triangles32_n = triangles_naoselec.copy()
  269. triangles64_s = triangles_selec.copy()
  270. triangles64_n = triangles_naoselec.copy()
  271. triangles128_s = triangles_selec.copy()
  272. triangles128_n = triangles_naoselec.copy()
  273.  
  274. triangles16_s = fun.resizeImages(triangles16_s,16,16)
  275. triangles16_n = fun.resizeImages(triangles16_n,16,16)
  276. triangles32_s = fun.resizeImages(triangles32_s,32,32)
  277. triangles32_n = fun.resizeImages(triangles32_n,32,32)
  278. triangles64_s = fun.resizeImages(triangles64_s,64,64)
  279. triangles64_n = fun.resizeImages(triangles64_n,64,64)
  280. triangles128_s = fun.resizeImages(triangles128_s,128,128)
  281. triangles128_n = fun.resizeImages(triangles128_n,128,128)
  282.  
  283. #convertendo para níveis de cinza
  284.  
  285. squares16_s = fun.grayConversion(squares16_s)
  286. squares16_n = fun.grayConversion(squares16_n)
  287. squares32_s = fun.grayConversion(squares32_s)
  288. squares32_n = fun.grayConversion(squares32_n)
  289. squares64_s = fun.grayConversion(squares64_s)
  290. squares64_n = fun.grayConversion(squares64_n)
  291. squares128_s = fun.grayConversion(squares128_s)
  292. squares128_n = fun.grayConversion(squares128_n)
  293.  
  294.  
  295. circles16_s = fun.grayConversion(circles16_s)
  296. circles16_n = fun.grayConversion(circles16_n)
  297. circles32_s = fun.grayConversion(circles32_s)
  298. circles32_n = fun.grayConversion(circles32_n)
  299. circles64_s = fun.grayConversion(circles64_s)
  300. circles64_n = fun.grayConversion(circles64_n)
  301. circles128_s = fun.grayConversion(circles128_s)
  302. circles128_n = fun.grayConversion(circles128_n)
  303.  
  304. triangles16_s = fun.grayConversion(triangles16_s)
  305. triangles16_n = fun.grayConversion(triangles16_n)
  306. triangles32_s = fun.grayConversion(triangles32_s)
  307. triangles32_n = fun.grayConversion(triangles32_n)
  308. triangles64_s = fun.grayConversion(triangles64_s)
  309. triangles64_n = fun.grayConversion(triangles64_n)
  310. triangles128_s = fun.grayConversion(triangles128_s)
  311. triangles128_n = fun.grayConversion(triangles128_n)
  312.  
  313.  
  314. trapezia16_s = fun.grayConversion(trapezia16_s)
  315. trapezia16_n = fun.grayConversion(trapezia16_n)
  316. trapezia32_s = fun.grayConversion(trapezia32_s)
  317. trapezia32_n = fun.grayConversion(trapezia32_n)
  318. trapezia64_s = fun.grayConversion(trapezia64_s)
  319. trapezia64_n = fun.grayConversion(trapezia64_n)
  320. trapezia128_s = fun.grayConversion(trapezia128_s)
  321. trapezia128_n = fun.grayConversion(trapezia128_n)
  322.  
  323. rhombuses16_s = fun.grayConversion(rhombuses16_s)
  324. rhombuses16_n = fun.grayConversion(rhombuses16_n)
  325. rhombuses32_s = fun.grayConversion(rhombuses32_s)
  326. rhombuses32_n = fun.grayConversion(rhombuses32_n)
  327. rhombuses64_s = fun.grayConversion(rhombuses64_s)
  328. rhombuses64_n = fun.grayConversion(rhombuses64_n)
  329. rhombuses128_s = fun.grayConversion(rhombuses128_s)
  330. rhombuses128_n = fun.grayConversion(rhombuses128_n)
  331.  
  332. rectangles16_s = fun.grayConversion(rectangles16_s)
  333. rectangles16_n = fun.grayConversion(rectangles16_n)
  334. rectangles32_s = fun.grayConversion(rectangles32_s)
  335. rectangles32_n = fun.grayConversion(rectangles32_n)
  336. rectangles64_s = fun.grayConversion(rectangles64_s)
  337. rectangles64_n = fun.grayConversion(rectangles64_n)
  338. rectangles128_s = fun.grayConversion(rectangles128_s)
  339. rectangles128_n = fun.grayConversion(rectangles128_n)
  340.  
  341. lines16_s = fun.grayConversion(lines16_s)
  342. lines16_n = fun.grayConversion(lines16_n)
  343. lines32_s = fun.grayConversion(lines32_s)
  344. lines32_n = fun.grayConversion(lines32_n)
  345. lines64_s = fun.grayConversion(lines64_s)
  346. lines64_n = fun.grayConversion(lines64_n)
  347. lines128_s = fun.grayConversion(lines128_s)
  348. lines128_n = fun.grayConversion(lines128_n)
  349.  
  350. hexagons16_s = fun.grayConversion(hexagons16_s)
  351. hexagons16_n = fun.grayConversion(hexagons16_n)
  352. hexagons32_s = fun.grayConversion(hexagons32_s)
  353. hexagons32_n = fun.grayConversion(hexagons32_n)
  354. hexagons64_s = fun.grayConversion(hexagons64_s)
  355. hexagons64_n = fun.grayConversion(hexagons64_n)
  356. hexagons128_s = fun.grayConversion(hexagons128_s)
  357. hexagons128_n = fun.grayConversion(hexagons128_n)
  358.  
  359. ellipsis16_s = fun.grayConversion(ellipsis16_s)
  360. ellipsis16_n = fun.grayConversion(ellipsis16_n)
  361. ellipsis32_s = fun.grayConversion(ellipsis32_s)
  362. ellipsis32_n = fun.grayConversion(ellipsis32_n)
  363. ellipsis64_s = fun.grayConversion(ellipsis64_s)
  364. ellipsis64_n = fun.grayConversion(ellipsis64_n)
  365. ellipsis128_s = fun.grayConversion(ellipsis128_s)
  366. ellipsis128_n = fun.grayConversion(ellipsis128_n)
  367.  
  368. #aplicando o filtro gaussiano
  369.  
  370. squares16_s = fun.blurConversion(squares16_s,5,0)
  371. squares16_n = fun.blurConversion(squares16_n,5,0)
  372. squares32_s = fun.blurConversion(squares32_s,5,0)
  373. squares32_n = fun.blurConversion(squares32_n,5,0)
  374. squares64_s = fun.blurConversion(squares64_s,5,0)
  375. squares64_n = fun.blurConversion(squares64_n,5,0)
  376. squares128_s = fun.blurConversion(squares128_s,5,0)
  377. squares128_n = fun.blurConversion(squares128_n,5,0)
  378.  
  379. circles16_s = fun.blurConversion(circles16_s, 5, 0)
  380. circles16_n = fun.blurConversion(circles16_n, 5, 0)
  381. circles32_s = fun.blurConversion(circles32_s,5 ,0)
  382. circles32_n = fun.blurConversion(circles32_n,5 ,0)
  383. circles64_s = fun.blurConversion(circles64_s,5,0)
  384. circles64_n = fun.blurConversion(circles64_n,5,0)
  385. circles128_s = fun.blurConversion(circles128_s,5,0)
  386. circles128_n = fun.blurConversion(circles128_n,5,0)
  387.  
  388. triangles16_s = fun.blurConversion(triangles16_s,5,0)
  389. triangles16_n = fun.blurConversion(triangles16_n,5,0)
  390. triangles32_s = fun.blurConversion(triangles32_s,5,0)
  391. triangles32_n = fun.blurConversion(triangles32_n,5,0)
  392. triangles64_s = fun.blurConversion(triangles64_s,5,0)
  393. triangles64_n = fun.blurConversion(triangles64_n,5,0)
  394. triangles128_s = fun.blurConversion(triangles128_s,5,0)
  395. triangles128_n = fun.blurConversion(triangles128_n,5,0)
  396.  
  397. trapezia16_s = fun.blurConversion(trapezia16_s,5,0)
  398. trapezia16_n = fun.blurConversion(trapezia16_n,5,0)
  399. trapezia32_s = fun.blurConversion(trapezia32_s,5,0)
  400. trapezia64_s = fun.blurConversion(trapezia64_s,5,0)
  401. trapezia64_n = fun.blurConversion(trapezia64_n,5,0)
  402. trapezia128_s = fun.blurConversion(trapezia128_s,5,0)
  403. trapezia128_n = fun.blurConversion(trapezia128_n,5,0)
  404.  
  405. rhombuses16_s = fun.blurConversion(rhombuses16_s,5,0)
  406. rhombuses16_n = fun.blurConversion(rhombuses16_n,5,0)
  407. rhombuses32_s = fun.blurConversion(rhombuses32_s,5,0)
  408. rhombuses32_n = fun.blurConversion(rhombuses32_n,5,0)
  409. rhombuses64_s = fun.blurConversion(rhombuses64_s,5,0)
  410. rhombuses64_n = fun.blurConversion(rhombuses64_n,5,0)
  411. rhombuses128_s = fun.blurConversion(rhombuses128_s,5,0)
  412. rhombuses128_n = fun.blurConversion(rhombuses128_n,5,0)
  413.  
  414. rectangles16_s = fun.blurConversion(rectangles16_s,5,0)
  415. rectangles16_n = fun.blurConversion(rectangles16_n,5,0)
  416. rectangles32_s = fun.blurConversion(rectangles32_s,5,0)
  417. rectangles32_n = fun.blurConversion(rectangles32_n,5,0)
  418. rectangles64_s = fun.blurConversion(rectangles64_s,5,0)
  419. rectangles64_n = fun.blurConversion(rectangles64_n,5,0)
  420. rectangles128_s = fun.blurConversion(rectangles128_s,5,0)
  421. rectangles128_n = fun.blurConversion(rectangles128_n,5,0)
  422.  
  423. lines16_s = fun.blurConversion(lines16_s,5,0)
  424. lines16_n = fun.blurConversion(lines16_n,5,0)
  425. lines32_s = fun.blurConversion(lines32_s,5,0)
  426. lines32_n = fun.blurConversion(lines32_n,5,0)
  427. lines64_s = fun.blurConversion(lines64_s,5,0)
  428. lines64_n = fun.blurConversion(lines64_n,5,0)
  429. lines128_s = fun.blurConversion(lines128_s,5,0)
  430. lines128_n = fun.blurConversion(lines128_n,5,0)
  431.  
  432. hexagons16_s = fun.blurConversion(hexagons16_s,5,0)
  433. hexagons16_n = fun.blurConversion(hexagons16_n,5,0)
  434. hexagons32_s = fun.blurConversion(hexagons32_s,5,0)
  435. hexagons32_n = fun.blurConversion(hexagons32_n,5,0)
  436. hexagons64_s = fun.blurConversion(hexagons64_s,5,0)
  437. hexagons64_n = fun.blurConversion(hexagons64_n,5,0)
  438. hexagons128_s = fun.blurConversion(hexagons128_s,5,0)
  439. hexagons128_n = fun.blurConversion(hexagons128_n,5,0)
  440.  
  441. ellipsis16_s = fun.blurConversion(ellipsis16_s,5,0)
  442. ellipsis16_n = fun.blurConversion(ellipsis16_n,5,0)
  443. ellipsis32_s = fun.blurConversion(ellipsis32_s,5,0)
  444. ellipsis32_n = fun.blurConversion(ellipsis32_n,5,0)
  445. ellipsis64_s = fun.blurConversion(ellipsis64_s,5,0)
  446. ellipsis64_n = fun.blurConversion(ellipsis64_n,5,0)
  447. ellipsis128_s = fun.blurConversion(ellipsis128_s,5,0)
  448. ellipsis128_n = fun.blurConversion(ellipsis128_n,5,0)
  449.  
  450.  
  451. #convertendo para binária
  452. squares16_s = fun.binaryConversion(squares16_s,255,31)
  453. squares16_n = fun.binaryConversion(squares16_n,255,31)
  454. squares32_s = fun.binaryConversion(squares32_s,255,31)
  455. squares32_n = fun.binaryConversion(squares32_n,255,31)
  456. squares64_s = fun.binaryConversion(squares64_s,255,31)
  457. squares64_n = fun.binaryConversion(squares64_n,255,31)
  458. squares128_s = fun.binaryConversion(squares128_s,255,31)
  459. squares128_n = fun.binaryConversion(squares128_n,255,31)
  460.  
  461. circles16_s = fun.binaryConversion(circles16_s, 255, 31)
  462. circles16_n = fun.binaryConversion(circles16_n, 255, 31)
  463. circles32_s = fun.binaryConversion(circles32_s,255,31)
  464. circles32_n = fun.binaryConversion(circles32_n,255,31)
  465. circles64_s = fun.binaryConversion(circles64_s,255,31)
  466. circles64_n = fun.binaryConversion(circles64_n,255,31)
  467. circles128_s = fun.binaryConversion(circles128_s,255,31)
  468. circles128_n = fun.binaryConversion(circles128_n,255,31)
  469.  
  470. triangles16_s = fun.binaryConversion(triangles16_s,255,31)
  471. triangles16_n = fun.binaryConversion(triangles16_n,255,31)
  472. triangles32_s = fun.binaryConversion(triangles32_s,255,31)
  473. triangles32_n = fun.binaryConversion(triangles32_n,255,31)
  474. triangles64_s = fun.binaryConversion(triangles64_s,255,31)
  475. triangles64_n = fun.binaryConversion(triangles64_n,255,31)
  476. triangles128_s = fun.binaryConversion(triangles128_s,255,31)
  477. triangles128_n = fun.binaryConversion(triangles128_n,255,31)
  478.  
  479. trapezia16_s = fun.binaryConversion(trapezia16_s,255,31)
  480. trapezia16_n = fun.binaryConversion(trapezia16_n,255,31)
  481. trapezia32_s = fun.binaryConversion(trapezia32_s,255,31)
  482. trapezia32_n = fun.binaryConversion(trapezia32_n,255,31)
  483. trapezia64_s = fun.binaryConversion(trapezia64_s,255,31)
  484. trapezia64_n = fun.binaryConversion(trapezia64_n,255,31)
  485. trapezia128_s = fun.binaryConversion(trapezia128_s,255,31)
  486. trapezia128_n = fun.binaryConversion(trapezia128_n,255,31)
  487.  
  488. rhombuses16_s = fun.binaryConversion(rhombuses16_s,255,31)
  489. rhombuses16_n = fun.binaryConversion(rhombuses16_n,255,31)
  490. rhombuses32_s = fun.binaryConversion(rhombuses32_s,255,31)
  491. rhombuses32_n = fun.binaryConversion(rhombuses32_n,255,31)
  492. rhombuses64_s = fun.binaryConversion(rhombuses64_s,255,31)
  493. rhombuses64_n = fun.binaryConversion(rhombuses64_n,255,31)
  494. rhombuses128_s = fun.binaryConversion(rhombuses128_s,255,31)
  495. rhombuses128_n = fun.binaryConversion(rhombuses128_n,255,31)
  496.  
  497. rectangles16_s = fun.binaryConversion(rectangles16_s,255,31)
  498. rectangles16_n = fun.binaryConversion(rectangles16_n,255,31)
  499. rectangles32_s = fun.binaryConversion(rectangles32_s,255,31)
  500. rectangles32_n = fun.binaryConversion(rectangles32_n,255,31)
  501. rectangles64_s = fun.binaryConversion(rectangles64_s,255,31)
  502. rectangles64_n = fun.binaryConversion(rectangles64_n,255,31)
  503. rectangles128_s = fun.binaryConversion(rectangles128_s,255,31)
  504. rectangles128_n = fun.binaryConversion(rectangles128_n,255,31)
  505.  
  506. lines16_s = fun.binaryConversion(lines16_s,255,31)
  507. lines16_n = fun.binaryConversion(lines16_n,255,31)
  508. lines32_s = fun.binaryConversion(lines32_s,255,31)
  509. lines32_n = fun.binaryConversion(lines32_n,255,31)
  510. lines64_s = fun.binaryConversion(lines64_s,255,31)
  511. lines64_n = fun.binaryConversion(lines64_n,255,31)
  512. lines128_s = fun.binaryConversion(lines128_s,255,31)
  513. lines128_n = fun.binaryConversion(lines128_n,255,31)
  514.  
  515. hexagons16_s = fun.binaryConversion(hexagons16_s,255,31)
  516. hexagons16_n = fun.binaryConversion(hexagons16_n,255,31)
  517. hexagons32_s = fun.binaryConversion(hexagons32_s,255,31)
  518. hexagons32_n = fun.binaryConversion(hexagons32_n,255,31)
  519. hexagons64_s = fun.binaryConversion(hexagons64_s,255,31)
  520. hexagons64_n = fun.binaryConversion(hexagons64_n,255,31)
  521. hexagons128_s = fun.binaryConversion(hexagons128_s,255,31)
  522. hexagons128_n = fun.binaryConversion(hexagons128_n,255,31)
  523.  
  524. ellipsis16_s = fun.binaryConversion(ellipsis16_s,255,31)
  525. ellipsis16_n = fun.binaryConversion(ellipsis16_n,255,31)
  526. ellipsis32_s = fun.binaryConversion(ellipsis32_s,255,31)
  527. ellipsis32_n = fun.binaryConversion(ellipsis32_n,255,31)
  528. ellipsis64_s = fun.binaryConversion(ellipsis64_s,255,31)
  529. ellipsis64_n = fun.binaryConversion(ellipsis64_n,255,31)
  530. ellipsis128_s = fun.binaryConversion(ellipsis128_s,255,31)
  531. ellipsis128_n = fun.binaryConversion(ellipsis128_n,255,31)
  532.  
  533. #invertendo as cores
  534.  
  535. squares16_s = fun.invertConversion(squares16_s)
  536. squares16_n = fun.invertConversion(squares16_n)
  537. squares32_s = fun.invertConversion(squares32_s)
  538. squares32_n = fun.invertConversion(squares32_n)
  539. squares64_s = fun.invertConversion(squares64_s)
  540. squares64_n = fun.invertConversion(squares64_n)
  541. squares128_s = fun.invertConversion(squares128_s)
  542. squares128_n = fun.invertConversion(squares128_n)
  543.  
  544. circles16_s = fun.invertConversion(circles16_s)
  545. circles16_n = fun.invertConversion(circles16_n)
  546. circles32_s = fun.invertConversion(circles32_s)
  547. circles32_n = fun.invertConversion(circles32_n)
  548. circles64_s = fun.invertConversion(circles64_s)
  549. circles64_n = fun.invertConversion(circles64_n)
  550. circles128_s = fun.invertConversion(circles128_s)
  551. circles128_n = fun.invertConversion(circles128_n)
  552.  
  553. triangles16_s = fun.invertConversion(triangles16_s)
  554. triangles16_n = fun.invertConversion(triangles16_n)
  555. triangles32_s = fun.invertConversion(triangles32_s)
  556. triangles32_n = fun.invertConversion(triangles32_n)
  557. triangles64_s = fun.invertConversion(triangles64_s)
  558. triangles64_n = fun.invertConversion(triangles64_n)
  559. triangles128_s = fun.invertConversion(triangles128_s)
  560. triangles128_n = fun.invertConversion(triangles128_n)
  561.  
  562. trapezia16_s = fun.invertConversion(trapezia16_s)
  563. trapezia16_n = fun.invertConversion(trapezia16_n)
  564. trapezia32_s = fun.invertConversion(trapezia32_s)
  565. trapezia32_n = fun.invertConversion(trapezia32_n)
  566. trapezia64_s = fun.invertConversion(trapezia64_s)
  567. trapezia64_n = fun.invertConversion(trapezia64_n)
  568. trapezia128_s = fun.invertConversion(trapezia128_s)
  569. trapezia128_n = fun.invertConversion(trapezia128_n)
  570.  
  571. rhombuses16_s = fun.invertConversion(rhombuses16_s)
  572. rhombuses16_n = fun.invertConversion(rhombuses16_n)
  573. rhombuses32_s = fun.invertConversion(rhombuses32_s)
  574. rhombuses32_n = fun.invertConversion(rhombuses32_n)
  575. rhombuses64_s = fun.invertConversion(rhombuses64_s)
  576. rhombuses64_n = fun.invertConversion(rhombuses64_n)
  577. rhombuses128_s = fun.invertConversion(rhombuses128_s)
  578. rhombuses128_n = fun.invertConversion(rhombuses128_n)
  579.  
  580. rectangles16_s = fun.invertConversion(rectangles16_s)
  581. rectangles16_n = fun.invertConversion(rectangles16_n)
  582. rectangles32_s = fun.invertConversion(rectangles32_s)
  583. rectangles32_n = fun.invertConversion(rectangles32_n)
  584. rectangles64_s = fun.invertConversion(rectangles64_s)
  585. rectangles64_n = fun.invertConversion(rectangles64_n)
  586. rectangles128_s = fun.invertConversion(rectangles128_s)
  587. rectangles128_n = fun.invertConversion(rectangles128_n)
  588.  
  589. lines16_s = fun.invertConversion(lines16_s)
  590. lines16_n = fun.invertConversion(lines16_n)
  591. lines32_s = fun.invertConversion(lines32_s)
  592. lines32_n = fun.invertConversion(lines32_n)
  593. lines64_s = fun.invertConversion(lines64_s)
  594. lines64_n = fun.invertConversion(lines64_n)
  595. lines128_s = fun.invertConversion(lines128_s)
  596. lines128_n = fun.invertConversion(lines128_n)
  597.  
  598. hexagons16_s = fun.invertConversion(hexagons16_s)
  599. hexagons16_n = fun.invertConversion(hexagons16_n)
  600. hexagons32_s = fun.invertConversion(hexagons32_s)
  601. hexagons32_n = fun.invertConversion(hexagons32_n)
  602. hexagons64_s = fun.invertConversion(hexagons64_s)
  603. hexagons64_n = fun.invertConversion(hexagons64_n)
  604. hexagons128_s = fun.invertConversion(hexagons128_s)
  605. hexagons128_n = fun.invertConversion(hexagons128_n)
  606.  
  607. ellipsis16_s = fun.invertConversion(ellipsis16_s)
  608. ellipsis16_n = fun.invertConversion(ellipsis16_n)
  609. ellipsis32_s = fun.invertConversion(ellipsis32_s)
  610. ellipsis32_n = fun.invertConversion(ellipsis32_n)
  611. ellipsis64_s = fun.invertConversion(ellipsis64_s)
  612. ellipsis64_n = fun.invertConversion(ellipsis64_n)
  613. ellipsis128_s = fun.invertConversion(ellipsis128_s)
  614. ellipsis128_n = fun.invertConversion(ellipsis128_n)
  615. print('terminou pre processing')
  616.  
  617. # extraindo caracteristicas das imagens
  618.  
  619. squares128_vector_s = fun.extratorCaracteristicas(squares128_s)
  620. squares128_vector_n = fun.extratorCaracteristicas(squares128_n)
  621. circles128_vector_s = fun.extratorCaracteristicas(circles128_s)
  622. circles128_vector_n = fun.extratorCaracteristicas(circles128_n)
  623. triangles128_vector_s = fun.extratorCaracteristicas(triangles128_s)
  624. triangles128_vector_n = fun.extratorCaracteristicas(triangles128_n)
  625. trapezia128_vector_s = fun.extratorCaracteristicas(trapezia128_s)
  626. trapezia128_vector_n = fun.extratorCaracteristicas(trapezia128_n)
  627. rhombuses128_vector_s = fun.extratorCaracteristicas(rhombuses128_s)
  628. rhombuses128_vector_n = fun.extratorCaracteristicas(rhombuses128_n)
  629. rectangles128_vector_s = fun.extratorCaracteristicas(rectangles128_s)
  630. rectangles128_vector_n = fun.extratorCaracteristicas(rectangles128_n)
  631. lines128_vector_s = fun.extratorCaracteristicas(lines128_s)
  632. lines128_vector_n = fun.extratorCaracteristicas(lines128_n)
  633. hexagons128_vector_s = fun.extratorCaracteristicas(hexagons128_s)
  634. hexagons128_vector_n = fun.extratorCaracteristicas(hexagons128_n)
  635. ellipsis128_vector_s = fun.extratorCaracteristicas(ellipsis128_s)
  636. ellipsis128_vector_n = fun.extratorCaracteristicas(ellipsis128_n)
  637.  
  638. squares64_vector_s = fun.extratorCaracteristicas(squares64_s)
  639. squares64_vector_n = fun.extratorCaracteristicas(squares64_n)
  640. circles64_vector_s = fun.extratorCaracteristicas(circles64_s)
  641. circles64_vector_n = fun.extratorCaracteristicas(circles64_n)
  642. triangles64_vector_s = fun.extratorCaracteristicas(triangles64_s)
  643. triangles64_vector_n = fun.extratorCaracteristicas(triangles64_n)
  644. trapezia64_vector_s = fun.extratorCaracteristicas(trapezia64_s)
  645. trapezia64_vector_n = fun.extratorCaracteristicas(trapezia64_n)
  646. rhombuses64_vector_s = fun.extratorCaracteristicas(rhombuses64_s)
  647. rhombuses64_vector_n = fun.extratorCaracteristicas(rhombuses64_n)
  648. rectangles64_vector_s = fun.extratorCaracteristicas(rectangles64_s)
  649. rectangles64_vector_n = fun.extratorCaracteristicas(rectangles64_n)
  650. lines64_vector_s = fun.extratorCaracteristicas(lines64_s)
  651. lines64_vector_n = fun.extratorCaracteristicas(lines64_n)
  652. hexagons64_vector_s = fun.extratorCaracteristicas(hexagons64_s)
  653. hexagons64_vector_n = fun.extratorCaracteristicas(hexagons64_n)
  654. ellipsis64_vector_s = fun.extratorCaracteristicas(ellipsis64_s)
  655. ellipsis64_vector_n = fun.extratorCaracteristicas(ellipsis64_n)
  656.  
  657. squares32_vector_s = fun.extratorCaracteristicas(squares32_s)
  658. squares32_vector_n = fun.extratorCaracteristicas(squares32_n)
  659. circles32_vector_s = fun.extratorCaracteristicas(circles32_s)
  660. circles32_vector_n = fun.extratorCaracteristicas(circles32_n)
  661. triangles32_vector_s = fun.extratorCaracteristicas(triangles32_s)
  662. triangles32_vector_n = fun.extratorCaracteristicas(triangles32_n)
  663. trapezia32_vector_s = fun.extratorCaracteristicas(trapezia32_s)
  664. trapezia32_vector_n = fun.extratorCaracteristicas(trapezia32_n)
  665. rhombuses32_vector_s = fun.extratorCaracteristicas(rhombuses32_s)
  666. rhombuses32_vector_n = fun.extratorCaracteristicas(rhombuses32_n)
  667. rectangles32_vector_s = fun.extratorCaracteristicas(rectangles32_s)
  668. rectangles32_vector_n = fun.extratorCaracteristicas(rectangles32_n)
  669. lines32_vector_s = fun.extratorCaracteristicas(lines32_s)
  670. lines32_vector_n = fun.extratorCaracteristicas(lines32_n)
  671. hexagons32_vector_s = fun.extratorCaracteristicas(hexagons32_s)
  672. hexagons32_vector_n = fun.extratorCaracteristicas(hexagons32_n)
  673. ellipsis32_vector_s = fun.extratorCaracteristicas(ellipsis32_s)
  674. ellipsis32_vector_n = fun.extratorCaracteristicas(ellipsis32_n)
  675.  
  676.  
  677. squares16_vector_s = fun.extratorCaracteristicas(squares16_s)
  678. squares16_vector_n = fun.extratorCaracteristicas(squares16_n)
  679. circles16_vector_s = fun.extratorCaracteristicas(circles16_s)
  680. circles16_vector_n = fun.extratorCaracteristicas(circles16_n)
  681. triangles16_vector_s = fun.extratorCaracteristicas(triangles16_s)
  682. triangles16_vector_n = fun.extratorCaracteristicas(triangles16_n)
  683. trapezia16_vector_s = fun.extratorCaracteristicas(trapezia16_s)
  684. trapezia16_vector_n = fun.extratorCaracteristicas(trapezia16_n)
  685. rhombuses16_vector_s = fun.extratorCaracteristicas(rhombuses16_s)
  686. rhombuses16_vector_n = fun.extratorCaracteristicas(rhombuses16_n)
  687. rectangles16_vector_s = fun.extratorCaracteristicas(rectangles16_s)
  688. rectangles16_vector_n = fun.extratorCaracteristicas(rectangles16_n)
  689. lines16_vector_s = fun.extratorCaracteristicas(lines16_s)
  690. lines16_vector_n = fun.extratorCaracteristicas(lines16_n)
  691. hexagons16_vector_s = fun.extratorCaracteristicas(hexagons16_s)
  692. hexagons16_vector_n = fun.extratorCaracteristicas(hexagons16_n)
  693. ellipsis16_vector_s = fun.extratorCaracteristicas(ellipsis16_s)
  694. ellipsis16_vector_n = fun.extratorCaracteristicas(ellipsis16_n)
  695.  
  696.  
  697. print('terminou extracao carac')
  698.  
  699. # transformando os vetores em dataframes
  700.  
  701.  
  702. squares128_vector_s = pd.DataFrame(squares128_vector_s)
  703. squares128_vector_n = pd.DataFrame(squares128_vector_n)
  704. circles128_vector_s = pd.DataFrame(circles128_vector_s)
  705. circles128_vector_n = pd.DataFrame(circles128_vector_n)
  706. triangles128_vector_s = pd.DataFrame(triangles128_vector_s)
  707. triangles128_vector_n = pd.DataFrame(triangles128_vector_n)
  708. trapezia128_vector_s = pd.DataFrame(trapezia128_vector_s)
  709. trapezia128_vector_n = pd.DataFrame(trapezia128_vector_n)
  710. rhombuses128_vector_s = pd.DataFrame(rhombuses128_vector_s)
  711. rhombuses128_vector_n = pd.DataFrame(rhombuses128_vector_n)
  712. rectangles128_vector_s = pd.DataFrame(rectangles128_vector_s)
  713. rectangles128_vector_n = pd.DataFrame(rectangles128_vector_n)
  714. lines128_vector_s = pd.DataFrame(lines128_vector_s)
  715. lines128_vector_n = pd.DataFrame(lines128_vector_n)
  716. hexagons128_vector_s = pd.DataFrame(hexagons128_vector_s)
  717. hexagons128_vector_n = pd.DataFrame(hexagons128_vector_n)
  718. ellipsis128_vector_s = pd.DataFrame(ellipsis128_vector_s)
  719. ellipsis128_vector_n = pd.DataFrame(ellipsis128_vector_n)
  720.  
  721. squares32_vector_s = pd.DataFrame(squares32_vector_s)
  722. squares32_vector_n = pd.DataFrame(squares32_vector_n)
  723. circles32_vector_s = pd.DataFrame(circles32_vector_s)
  724. circles32_vector_n = pd.DataFrame(circles32_vector_n)
  725. triangles32_vector_s = pd.DataFrame(triangles32_vector_s)
  726. triangles32_vector_n = pd.DataFrame(triangles32_vector_n)
  727. trapezia32_vector_s = pd.DataFrame(trapezia32_vector_s)
  728. trapezia32_vector_n = pd.DataFrame(trapezia32_vector_n)
  729. rhombuses32_vector_s = pd.DataFrame(rhombuses32_vector_s)
  730. rhombuses32_vector_n = pd.DataFrame(rhombuses32_vector_n)
  731. rectangles32_vector_s = pd.DataFrame(rectangles32_vector_s)
  732. rectangles32_vector_n = pd.DataFrame(rectangles32_vector_n)
  733. hexagons32_vector_s = pd.DataFrame(hexagons32_vector_s)
  734. hexagons32_vector_n = pd.DataFrame(hexagons32_vector_n)
  735. ellipsis32_vector_s = pd.DataFrame(ellipsis32_vector_s)
  736. ellipsis32_vector_n = pd.DataFrame(ellipsis32_vector_n)
  737. lines32_vector_s = pd.DataFrame(lines32_vector_s)
  738. lines32_vector_n = pd.DataFrame(lines32_vector_n)
  739.  
  740. squares64_vector_s = pd.DataFrame(squares64_vector_s)
  741. squares64_vector_n = pd.DataFrame(squares64_vector_n)
  742. circles64_vector_s = pd.DataFrame(circles64_vector_s)
  743. circles64_vector_n = pd.DataFrame(circles64_vector_n)
  744. triangles64_vector_s = pd.DataFrame(triangles64_vector_s)
  745. triangles64_vector_n = pd.DataFrame(triangles64_vector_n)
  746. trapezia64_vector_s = pd.DataFrame(trapezia64_vector_s)
  747. trapezia64_vector_n = pd.DataFrame(trapezia64_vector_n)
  748. rhombuses64_vector_s = pd.DataFrame(rhombuses64_vector_s)
  749. rhombuses64_vector_n = pd.DataFrame(rhombuses64_vector_n)
  750. rectangles64_vector_s = pd.DataFrame(rectangles64_vector_s)
  751. rectangles64_vector_n = pd.DataFrame(rectangles64_vector_n)
  752. lines64_vector_s = pd.DataFrame(lines64_vector_s)
  753. lines64_vector_n = pd.DataFrame(lines64_vector_n)
  754. hexagons64_vector_s = pd.DataFrame(hexagons64_vector_s)
  755. hexagons64_vector_n = pd.DataFrame(hexagons64_vector_n)
  756. ellipsis64_vector_s = pd.DataFrame(ellipsis64_vector_s)
  757. ellipsis64_vector_n = pd.DataFrame(ellipsis64_vector_n)
  758.  
  759.  
  760. circles16_vector_s = pd.DataFrame(circles16_vector_s)
  761. circles16_vector_n = pd.DataFrame(circles16_vector_n)
  762. squares16_vector_s = pd.DataFrame(squares16_vector_s)
  763. squares16_vector_n = pd.DataFrame(squares16_vector_n)
  764. triangles16_vector_s = pd.DataFrame(triangles16_vector_s)
  765. triangles16_vector_n = pd.DataFrame(triangles16_vector_n)
  766. trapezia16_vector_s = pd.DataFrame(trapezia16_vector_s)
  767. trapezia16_vector_n = pd.DataFrame(trapezia16_vector_n)
  768. rhombuses16_vector_s = pd.DataFrame(rhombuses16_vector_s)
  769. rhombuses16_vector_n = pd.DataFrame(rhombuses16_vector_n)
  770. rectangles16_vector_s = pd.DataFrame(rectangles16_vector_s)
  771. rectangles16_vector_n = pd.DataFrame(rectangles16_vector_n)
  772. lines16_vector_s = pd.DataFrame(lines16_vector_s)
  773. lines16_vector_n = pd.DataFrame(lines16_vector_n)
  774. hexagons16_vector_s = pd.DataFrame(hexagons16_vector_s)
  775. hexagons16_vector_n = pd.DataFrame(hexagons16_vector_n)
  776. ellipsis16_vector_s = pd.DataFrame(ellipsis16_vector_s)
  777. ellipsis16_vector_n = pd.DataFrame(ellipsis16_vector_n)
  778.  
  779.  
  780. print('terminou transformar em dataframe')
  781.  
  782. #incluindo a classe nos dataframes
  783.  
  784. squares128_vector_s['Classe'] = 'square'
  785. squares128_vector_n['Classe'] = 'square'
  786. circles128_vector_s['Classe'] = 'circle'
  787. circles128_vector_n['Classe'] = 'circle'
  788. triangles128_vector_s['Classe'] = 'triangle'
  789. triangles128_vector_n['Classe'] = 'triangle'
  790. trapezia128_vector_s['Classe'] = 'trapezia'
  791. trapezia128_vector_n['Classe'] = 'trapezia'
  792. rhombuses128_vector_s['Classe'] = 'rhombuse'
  793. rhombuses128_vector_n['Classe'] = 'rhombuse'
  794. rectangles128_vector_s['Classe'] = 'rectangle'
  795. rectangles128_vector_n['Classe'] = 'rectangle'
  796. lines128_vector_s['Classe'] = 'line'
  797. lines128_vector_n['Classe'] = 'line'
  798. hexagons128_vector_s['Classe'] = 'hexagon'
  799. hexagons128_vector_n['Classe'] = 'hexagon'
  800. ellipsis128_vector_s['Classe'] = 'ellipse'
  801. ellipsis128_vector_n['Classe'] = 'ellipse'
  802.  
  803. squares32_vector_s['Classe'] = 'square'
  804. squares32_vector_n['Classe'] = 'square'
  805. circles32_vector_s['Classe'] = 'circle'
  806. circles32_vector_n['Classe'] = 'circle'
  807. triangles32_vector_s['Classe'] = 'triangle'
  808. triangles32_vector_n['Classe'] = 'triangle'
  809. trapezia32_vector_s['Classe'] = 'trapezia'
  810. trapezia32_vector_n['Classe'] = 'trapezia'
  811. rhombuses32_vector_s['Classe'] = 'rhombuse'
  812. rhombuses32_vector_n['Classe'] = 'rhombuse'
  813. rectangles32_vector_s['Classe'] = 'rectangle'
  814. rectangles32_vector_n['Classe'] = 'rectangle'
  815. lines32_vector_s['Classe'] = 'line'
  816. lines32_vector_n['Classe'] = 'line'
  817. hexagons32_vector_s['Classe'] = 'hexagon'
  818. hexagons32_vector_n['Classe'] = 'hexagon'
  819. ellipsis32_vector_s['Classe'] = 'ellipse'
  820. ellipsis32_vector_n['Classe'] = 'ellipse'
  821.  
  822. squares64_vector_s['Classe'] = 'square'
  823. squares64_vector_n['Classe'] = 'square'
  824. circles64_vector_s['Classe'] = 'circle'
  825. circles64_vector_n['Classe'] = 'circle'
  826. triangles64_vector_s['Classe'] = 'triangle'
  827. triangles64_vector_n['Classe'] = 'triangle'
  828. trapezia64_vector_s['Classe'] = 'trapezia'
  829. trapezia64_vector_n['Classe'] = 'trapezia'
  830. rhombuses64_vector_s['Classe'] = 'rhombuse'
  831. rhombuses64_vector_n['Classe'] = 'rhombuse'
  832. rectangles64_vector_s['Classe'] = 'rectangle'
  833. rectangles64_vector_n['Classe'] = 'rectangle'
  834. lines64_vector_s['Classe'] = 'line'
  835. lines64_vector_n['Classe'] = 'line'
  836. hexagons64_vector_s['Classe'] = 'hexagon'
  837. hexagons64_vector_n['Classe'] = 'hexagon'
  838. ellipsis64_vector_s['Classe'] = 'ellipse'
  839. ellipsis64_vector_n['Classe'] = 'ellipse'
  840.  
  841.  
  842. squares16_vector_s['Classe'] = 'square'
  843. squares16_vector_n['Classe'] = 'square'
  844. circles16_vector_s['Classe'] = 'circle'
  845. circles16_vector_n['Classe'] = 'circle'
  846. triangles16_vector_s['Classe'] = 'triangle'
  847. triangles16_vector_n['Classe'] = 'triangle'
  848. trapezia16_vector_s['Classe'] = 'trapezia'
  849. trapezia16_vector_n['Classe'] = 'trapezia'
  850. rhombuses16_vector_s['Classe'] = 'rhombuse'
  851. rhombuses16_vector_n['Classe'] = 'rhombuse'
  852. rectangles16_vector_s['Classe'] = 'rectangle'
  853. rectangles16_vector_n['Classe'] = 'rectangle'
  854. lines16_vector_s['Classe'] = 'line'
  855. lines16_vector_n['Classe'] = 'line'
  856. hexagons16_vector_s['Classe'] = 'hexagon'
  857. hexagons16_vector_n['Classe'] = 'hexagon'
  858. ellipsis16_vector_s['Classe'] = 'ellipse'
  859. ellipsis16_vector_n['Classe'] = 'ellipse'
  860.  
  861.  
  862.  
  863. dfs64_s = [squares64_vector_s,circles64_vector_s,triangles64_vector_s,trapezia64_vector_s,rhombuses64_vector_s,
  864.          rectangles64_vector_s,lines64_vector_s,hexagons64_vector_s,ellipsis64_vector_s]
  865.  
  866. dfs64_n = [squares64_vector_n,circles64_vector_n,triangles64_vector_n,trapezia64_vector_n,rhombuses64_vector_n,
  867.          rectangles64_vector_n,lines64_vector_n,hexagons64_vector_n,ellipsis64_vector_n]
  868.  
  869. dfs128_s = [squares128_vector_s,circles128_vector_s,triangles128_vector_s,trapezia128_vector_s,rhombuses128_vector_s,
  870.           rectangles128_vector_s,lines128_vector_s,hexagons128_vector_s,ellipsis128_vector_s]
  871.  
  872. dfs128_n = [squares128_vector_n,circles128_vector_n,triangles128_vector_n,trapezia128_vector_n,rhombuses128_vector_n,
  873.           rectangles128_vector_n,lines128_vector_n,hexagons128_vector_n,ellipsis128_vector_n]
  874.  
  875. dfs32_s = [squares32_vector_s,circles32_vector_s,triangles32_vector_s,trapezia32_vector_s,rhombuses32_vector_s,
  876.          rectangles32_vector_s,lines32_vector_s,hexagons32_vector_s,ellipsis32_vector_s]
  877.  
  878. dfs32_n = [squares32_vector_n,circles32_vector_n,triangles32_vector_n,trapezia32_vector_n,rhombuses32_vector_n,
  879.          rectangles32_vector_n,lines32_vector_n,hexagons32_vector_n,ellipsis32_vector_n]
  880.  
  881. dfs16_s = [squares16_vector_s,circles16_vector_s,triangles16_vector_s,trapezia16_vector_s,rhombuses16_vector_s,
  882.        rectangles16_vector_s,lines16_vector_s,hexagons16_vector_s,ellipsis16_vector_s]
  883. dfs16_n = [squares16_vector_n,circles16_vector_n,triangles16_vector_n,trapezia16_vector_n,rhombuses16_vector_n,
  884.        rectangles16_vector_n,lines16_vector_n,hexagons16_vector_n,ellipsis16_vector_n]
  885.  
  886.  
  887.  
  888. # USANDO AS IMAGENS 128x128
  889.  
  890. dataFrame128_s = pd.concat(dfs128_s, ignore_index=True)
  891. dataFrame128_2_s = dataFrame128_s.copy()
  892. del dataFrame128_s['Classe']
  893. # dataFrame128_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_128_s_NOCLASS.csv')
  894. dataFrame128_s = fun.normalizar(dataFrame128_s)
  895. dataFrame128_s.fillna(0)
  896. dataFrame128_s['Classe'] = dataFrame128_2_s['Classe']
  897. dataFrame128_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_128_s.csv')
  898.  
  899. dataFrame128_n = pd.concat(dfs128_n, ignore_index=True)
  900. dataFrame128_n.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_128_n.csv')
  901. dataFrame128_2_n = dataFrame128_n.copy()
  902. del dataFrame128_n['Classe']
  903. dataFrame128_n = fun.normalizar(dataFrame128_n)
  904. dataFrame128_n.fillna(0)
  905. dataFrame128_n['Classe'] = dataFrame128_2_n['Classe']
  906.  
  907. dataFrame64_s = pd.concat(dfs64_s, ignore_index=True)
  908. dataFrame64_s = dataFrame64_s.fillna(0)
  909. dataFrame64_2_s = dataFrame64_s.copy()
  910. del dataFrame64_s['Classe']
  911. # dataFrame64_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_64_sNOCLASS.csv')
  912. dataFrame64_s = fun.normalizar(dataFrame64_s)
  913. dataFrame64_s['Classe'] = dataFrame64_2_s['Classe']
  914. dataFrame64_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_64_s.csv')
  915.  
  916. dataFrame64_n = pd.concat(dfs64_n, ignore_index=True)
  917. dataFrame64_n.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_64_n.csv')
  918. dataFrame64_n = dataFrame64_n.fillna(0)
  919. dataFrame64_2_n = dataFrame64_n.copy()
  920. del dataFrame64_n['Classe']
  921. dataFrame64_n = fun.normalizar(dataFrame64_n)
  922. dataFrame64_n['Classe'] = dataFrame64_2_n['Classe']
  923.  
  924. dataFrame32_s = pd.concat(dfs32_s, ignore_index=True)
  925. dataFrame32_s = dataFrame32_s.fillna(0)
  926. dataFrame32_2_s = dataFrame32_s.copy()
  927. del dataFrame32_s['Classe']
  928. # dataFrame32_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_32_sNOCLASS.csv')
  929. dataFrame32_s = fun.normalizar(dataFrame32_s)
  930. dataFrame32_s['Classe'] = dataFrame32_2_s['Classe']
  931. dataFrame32_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_32_s.csv')
  932.  
  933. dataFrame32_n = pd.concat(dfs32_n, ignore_index=True)
  934. dataFrame32_n.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_32_n.csv')
  935. dataFrame32_n = dataFrame32_n.fillna(0)
  936. dataFrame32_2_n = dataFrame32_n.copy()
  937. del dataFrame32_n['Classe']
  938. dataFrame32_n = fun.normalizar(dataFrame32_n)
  939. dataFrame32_n['Classe'] = dataFrame32_2_n['Classe']
  940.  
  941. dataFrame16_s = pd.concat(dfs16_s, ignore_index=True)
  942. dataFrame16_s = dataFrame16_s.fillna(0)
  943. dataFrame16_2_s = dataFrame16_s.copy()
  944. del dataFrame16_s['Classe']
  945. # dataFrame16_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_16_sNOCLASS.csv')
  946. dataFrame16_s = fun.normalizar(dataFrame16_s)
  947. dataFrame16_s['Classe'] = dataFrame16_2_s['Classe']
  948. dataFrame16_s.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_16_s.csv')
  949.  
  950. dataFrame16_n = pd.concat(dfs16_n, ignore_index=True)
  951. dataFrame16_n.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Images_16_n.csv')
  952. dataFrame16_n = dataFrame16_n.fillna(0)
  953. dataFrame16_2_n = dataFrame16_n.copy()
  954. del dataFrame16_n['Classe']
  955. dataFrame16_n = fun.normalizar(dataFrame16_n)
  956. dataFrame16_n['Classe'] = dataFrame16_2_n['Classe']
  957.  
  958. # Criando o objeto do tipo k-folds com 10 folds
  959. # kfold = KFold(10, True, 1)
  960.  
  961.  
  962. # Criando o k-fold com 5 folds para execução do algoritmo genético
  963. kfold = KFold(5, True, 1)
  964.  
  965. #Inicializando o Classificador do algoritmo genético
  966.  
  967. # Random Forest Classifier
  968. RandomForest = RandomForestClassifier()
  969. RandomForest_acerto = []
  970. RandomForest_accmedia = []
  971.  
  972. #criando a população com 20 cromossomos de tamanho 38
  973. cromossomos = fun.create_population(20, 38)
  974.  
  975.  
  976. for cromo in cromossomos:
  977.     #retorna as posições do array usadas pelo cromossomo
  978.     positions = fun.positions_chromossome(cromo)
  979.  
  980.     df_classe = fun.decode_chromossome(cromo)
  981.     df_classe_2 = df_classe.copy()
  982.     print('df com classe ')
  983.     print(df_classe_2)
  984.     del df_classe['Classe']
  985.  
  986.     imagens = np.array(df_classe)
  987.     caracteristicas = fun.carac_imagens(positions, imagens)
  988.     print('array com caracteristicas selecionadas ')
  989.     print(caracteristicas)
  990.     caracteristicas['Classe'] = df_classe_2['Classe']
  991.     print('array carac selec e classes')
  992.     print(caracteristicas)
  993.  
  994.     for x in range(0, 5):
  995.  
  996.         tempo1 = time.time()
  997.         cols = list(caracteristicas.columns)
  998.         print('colunas com classe')
  999.         print(cols)
  1000.         cols.remove('Classe')
  1001.         print('colunas sem classe')
  1002.         print(cols)
  1003.         df_images_noclass = caracteristicas[cols]
  1004.         df_images_class = caracteristicas['Classe']
  1005.         c = kfold.split(caracteristicas)
  1006.  
  1007.         for train_index, test_index in c:
  1008.             noclass_train, noclass_test = df_images_noclass.iloc[train_index], df_images_noclass.iloc[test_index]
  1009.             class_train, class_test = df_images_class.iloc[train_index], df_images_class.iloc[test_index]
  1010.  
  1011.             RandomForest_inicio = time.time()
  1012.             RandomForest.fit(noclass_train, class_train)
  1013.             RandomForest_acerto.append(RandomForest.score(noclass_test, class_test))
  1014.  
  1015.         df_classe_2 = df_classe_2.sample(frac=1)
  1016.         print("Terminou a ", x)
  1017.         tempo2 = time.time()
  1018.         print("Tempo da rodada ", x, (tempo2 - tempo1) / 60)
  1019.  
  1020.     RandomForest_accmedia.append(np.mean(RandomForest_acerto))
  1021.  
  1022.     tempofinal = time.time()
  1023.  
  1024. print('acuracia media ', RandomForest_accmedia)
  1025. print('acuracia')
  1026. print(RandomForest_acerto)
  1027. print('cromossomos')
  1028. print(cromossomos)
  1029.  
  1030. # passando a função que retorna os dois melhores individuos, as suas acuracias e o array de cromossomos
  1031. #atualizado
  1032. melhores_ind, best_acuracia, cromossomos = fun.get_best_cromossomos(RandomForest_accmedia,cromossomos)
  1033.  
  1034. #realizando o torneio para selecionar 10 pais
  1035. pais_torneio = []
  1036. for i in range(0,10):
  1037.     aux, cromossomos = fun.tournament_selection(RandomForest_accmedia,cromossomos)
  1038.     pais_torneio.append(aux)
  1039.  
  1040.  
  1041. #escolhendo os pais aleatoriamente
  1042. pais_pares = fun.generate_parents(pais_torneio)
  1043.  
  1044.  
  1045. #gerando filhos com operador crossover
  1046. offspring = []
  1047. for x in range(0, 5):
  1048.     filhos = []
  1049.     filhos.append(fun.crossover(0.9, pais_pares[x]))
  1050.  
  1051.     for f in filhos:
  1052.  
  1053.         f = fun.mutation(0.05, f)
  1054.         offspring.append(f)
  1055.  
  1056.  
  1057. #garantindo o elitismo
  1058. for ind in melhores_ind:
  1059.     offspring.append(ind)
  1060.  
  1061.  
  1062. print('offspring')
  1063. print(offspring)
  1064. print('melhores individios')
  1065. print(melhores_ind)
  1066.  
  1067. # melhores, acuracias_best, cromossomos = fun.get_best_cromossomos(RandomForest_acerto, cromossomos)
  1068. # print('melhores cromossomos', melhores)
  1069. # print('melhores acuracias', acuracias_best)
  1070. # filhos,pais = fun.genetic_algorithm(melhores,cromossomos,RandomForest_acerto)
  1071. # print('filhos')
  1072. # print(filhos)
  1073. # print('pais')
  1074. # print(pais)
  1075.  
  1076. # Instanciando os algoritmos e seus vetores de tempo e acurácia
  1077. '''
  1078. #instanciando DataFrame com dados de saida
  1079. DadosSaida = []
  1080. DadosSaida = pd.DataFrame(DadosSaida)
  1081.  
  1082.  
  1083. # Random Forest Classifier
  1084. RandomForest = RandomForestClassifier()
  1085. RandomForest_acerto = []
  1086. RandomForest_tempo = []
  1087. RandomForest_precision = []
  1088. RandomForest_recall = []
  1089. RandomForest_fscore = []
  1090.  
  1091. # SVM com função de kernel linear
  1092. SVMachine_L = SVC(kernel='linear')
  1093. SVMachine_L_acerto = []
  1094. SVMachine_L_tempo = []
  1095. SVMachine_L_precision = []
  1096. SVMachine_L_recall = []
  1097. SVMachine_L_fscore = []
  1098.  
  1099. #SVM com função de kernel RBF
  1100. SVMachine_RBF = SVC(kernel='rbf', gamma='scale')
  1101. SVMachine_RBF_acerto = []
  1102. SVMachine_RBF_tempo = []
  1103. SVMachine_RBF_precision = []
  1104. SVMachine_RBF_recall = []
  1105. SVMachine_RBF_fscore = []
  1106.  
  1107. # KNN com k = 3, 5, 7
  1108. KNN_3 = KNeighborsClassifier(n_neighbors=3, metric='euclidean')
  1109. KNN_3_acerto = []
  1110. KNN_3_tempo = []
  1111. KNN_3_precision = []
  1112. KNN_3_recall = []
  1113. KNN_3_fscore = []
  1114.  
  1115. KNN_5 = KNeighborsClassifier(n_neighbors=5, metric='euclidean')
  1116. KNN_5_acerto = []
  1117. KNN_5_tempo = []
  1118. KNN_5_precision = []
  1119. KNN_5_recall = []
  1120. KNN_5_fscore = []
  1121.  
  1122. KNN_7 =  KNeighborsClassifier(n_neighbors=7, metric='euclidean')
  1123. KNN_7_acerto = []
  1124. KNN_7_tempo = []
  1125. KNN_7_precision = []
  1126. KNN_7_recall = []
  1127. KNN_7_fscore = []
  1128.  
  1129. # KNN Ponderado com k = 3, 5, 7
  1130. KNNP_3 = KNeighborsClassifier(n_neighbors=3, weights='distance',metric='euclidean')
  1131. KNNP_3_acerto = []
  1132. KNNP_3_tempo = []
  1133. KNNP_3_precision = []
  1134. KNNP_3_recall = []
  1135. KNNP_3_fscore = []
  1136.  
  1137. KNNP_5 = KNeighborsClassifier(n_neighbors=5, weights='distance', metric='euclidean')
  1138. KNNP_5_acerto = []
  1139. KNNP_5_tempo = []
  1140. KNNP_5_precision = []
  1141. KNNP_5_recall = []
  1142. KNNP_5_fscore = []
  1143.  
  1144. KNNP_7 = KNeighborsClassifier(n_neighbors=7, weights='distance', metric='euclidean')
  1145. KNNP_7_acerto = []
  1146. KNNP_7_tempo = []
  1147. KNNP_7_precision = []
  1148. KNNP_7_recall = []
  1149. KNNP_7_fscore = []
  1150.  
  1151. # Naïve Bayes
  1152. NaiveBayes = GaussianNB()
  1153. NaiveBayes_acerto = []
  1154. NaiveBayes_tempo = []
  1155. NaiveBayes_precision = []
  1156. NaiveBayes_recall = []
  1157. NaiveBayes_fscore = []
  1158.  
  1159. # Árvore de decisão
  1160. DecisionTree = DecisionTreeClassifier()
  1161. DecisionTree_acerto = []
  1162. DecisionTree_tempo = []
  1163. DecisionTree_precision = []
  1164. DecisionTree_recall = []
  1165. DecisionTree_fscore = []
  1166.  
  1167. # MultiLayer Perceptron
  1168. MLP = MLPClassifier()
  1169. MLP_acerto = []
  1170. MLP_tempo = []
  1171. MLPP_precision = []
  1172. MLP_recall = []
  1173. MLP_fscore = []
  1174.  
  1175. # Regressão Logística
  1176. RegrLogistica = LogisticRegression(solver='lbfgs')
  1177. RegrLogistica_acerto = []
  1178. RegrLogistica_tempo = []
  1179. RegreLogistica_precision = []
  1180. RegrLogistica_recall = []
  1181. RegrLogistica_fscore = []
  1182.  
  1183. # ____________________ USANDO IMAGENS 128x128
  1184.  
  1185. print('comecou o K fold')
  1186.  
  1187. tempoinicial = time.time()
  1188.  
  1189. for x in range(0, 10):
  1190.  
  1191.    tempo1 = time.time()
  1192.    cols = list(dataFrame128.columns)
  1193.    cols.remove('Classe')
  1194.    df_images_noclass = dataFrame128[cols]
  1195.    df_images_class = dataFrame128['Classe']
  1196.    c = kfold.split(dataFrame128)
  1197.  
  1198.    for train_index, test_index in c:
  1199.  
  1200.        noclass_train, noclass_test = df_images_noclass.iloc[train_index], df_images_noclass.iloc[test_index]
  1201.        class_train, class_test = df_images_class.iloc[train_index], df_images_class.iloc[test_index]
  1202.  
  1203.        KNN3_inicio = time.time()
  1204.        KNN_3.fit(noclass_train, class_train)
  1205.        KNN_3_acerto.append(KNN_3.score(noclass_test, class_test))
  1206.        KNN_3_recall. append(recall_score(class_test, KNN_3.predict(noclass_test),average='weighted'))
  1207.        KNN_3_precision.append(precision_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1208.        KNN_3_fscore.append(f1_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1209.        KNN3_fim = time.time()
  1210.        KNN_3_tempo.append(KNN3_fim - KNN3_inicio)
  1211.  
  1212.        KNN5_inicio = time.time()
  1213.        KNN_5.fit(noclass_train, class_train)
  1214.        KNN_5_acerto.append(KNN_5.score(noclass_test, class_test))
  1215.        KNN_5_recall.append(recall_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1216.        KNN_5_precision.append(precision_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1217.        KNN_5_fscore.append(f1_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1218.        KNN5_fim = time.time()
  1219.        KNN_5_tempo.append(KNN5_fim - KNN5_inicio)
  1220.  
  1221.  
  1222.  
  1223.        KNN7_inicio = time.time()
  1224.        KNN_7.fit(noclass_train, class_train)
  1225.        KNN_7_acerto.append(KNN_3.score(noclass_test, class_test))
  1226.        KNN_7_precision.append(precision_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1227.        KNN_7_recall.append(recall_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1228.        KNN_7_fscore.append(f1_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1229.        KNN7_fim = time.time()
  1230.        KNN_7_tempo.append(KNN3_fim - KNN3_inicio)
  1231.  
  1232.        KNNP3_inicio = time.time()
  1233.        KNNP_3.fit(noclass_train, class_train)
  1234.        KNNP_3_acerto.append(KNNP_3.score(noclass_test, class_test))
  1235.        KNNP_3_precision.append(precision_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1236.        KNNP_3_recall.append(recall_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1237.        KNNP_3_fscore.append(f1_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1238.        KNNP3_fim = time.time()
  1239.        KNNP_3_tempo.append(KNNP3_fim - KNNP3_inicio)
  1240.  
  1241.        KNNP5_inicio = time.time()
  1242.        KNNP_5.fit(noclass_train, class_train)
  1243.        KNNP_5_acerto.append(KNNP_5.score(noclass_test, class_test))
  1244.        KNNP_5_precision.append(precision_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1245.        KNNP_5_recall.append(recall_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1246.        KNNP_5_fscore.append(f1_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1247.        KNNP5_fim = time.time()
  1248.        KNNP_5_tempo.append(KNNP5_fim - KNNP5_inicio)
  1249.  
  1250.        KNNP7_inicio = time.time()
  1251.        KNNP_7.fit(noclass_train, class_train)
  1252.        KNNP_7_acerto.append(KNNP_7.score(noclass_test, class_test))
  1253.        KNNP_7_precision.append(precision_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1254.        KNNP_7_recall.append(recall_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1255.        KNNP_7_fscore.append(f1_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1256.        KNNP7_fim = time.time()
  1257.        KNNP_7_tempo.append(KNNP7_fim - KNNP7_inicio)
  1258.  
  1259.        NaiveBayes_inicio = time.time()
  1260.        NaiveBayes.fit(noclass_train, class_train)
  1261.        NaiveBayes_acerto.append(NaiveBayes.score(noclass_test, class_test))
  1262.        NaiveBayes_precision.append(precision_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1263.        NaiveBayes_recall.append(recall_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1264.        NaiveBayes_fscore.append(f1_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1265.        NaiveBayes_fim = time.time()
  1266.        NaiveBayes_tempo.append(NaiveBayes_fim - NaiveBayes_inicio)
  1267.  
  1268.        DecisionTree_inicio = time.time()
  1269.        DecisionTree.fit(noclass_train, class_train)
  1270.        DecisionTree_acerto.append(DecisionTree.score(noclass_test, class_test))
  1271.        DecisionTree_precision.append(precision_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1272.        DecisionTree_recall.append(recall_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1273.        DecisionTree_fscore.append(f1_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1274.        DecisionTree_fim = time.time()
  1275.        DecisionTree_tempo.append(DecisionTree_fim - DecisionTree_inicio)
  1276.  
  1277.        SVMachine_L_inicio = time.time()
  1278.        SVMachine_L.fit(noclass_train, class_train)
  1279.        SVMachine_L_acerto.append(SVMachine_L.score(noclass_test, class_test))
  1280.        SVMachine_L_precision.append(precision_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1281.        SVMachine_L_recall.append(recall_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1282.        SVMachine_L_fscore.append(f1_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1283.        SVMachine_L_fim = time.time()
  1284.        SVMachine_L_tempo.append(SVMachine_L_fim - SVMachine_L_inicio)
  1285.  
  1286.        SVMachine_RBF_inicio = time.time()
  1287.        SVMachine_RBF.fit(noclass_train, class_train)
  1288.        SVMachine_RBF_acerto.append(SVMachine_RBF.score(noclass_test, class_test))
  1289.        SVMachine_RBF_recall.append(recall_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1290.        SVMachine_RBF_precision.append(precision_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1291.        SVMachine_RBF_fscore.append(f1_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1292.        SVMachine_RBF_fim = time.time()
  1293.        SVMachine_RBF_tempo.append(SVMachine_RBF_fim - SVMachine_RBF_inicio)
  1294.  
  1295.        RegrLogistica_inicio = time.time()
  1296.        RegrLogistica.fit(noclass_train, class_train)
  1297.        RegrLogistica_acerto.append(RegrLogistica.score(noclass_test, class_test))
  1298.        RegreLogistica_precision.append(precision_score(class_test, (RegrLogistica.predict(noclass_test)),average='weighted'))
  1299.        RegrLogistica_recall.append(recall_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1300.        RegrLogistica_fscore.append(f1_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1301.        RegrLogistica_fim = time.time()
  1302.        RegrLogistica_tempo.append(RegrLogistica_fim - RegrLogistica_inicio)
  1303.  
  1304.        MLP_inicio = time.time()
  1305.        MLP.fit(noclass_train, class_train)
  1306.        MLP_acerto.append(MLP.score(noclass_test, class_test))
  1307.        MLPP_precision.append(precision_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1308.        MLP_recall.append(recall_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1309.        MLP_fscore.append(f1_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1310.        MLP_fim = time.time()
  1311.        MLP_tempo.append(MLP_fim - MLP_inicio)
  1312.  
  1313.        RandomForest_inicio = time.time()
  1314.        RandomForest.fit(noclass_train, class_train)
  1315.        RandomForest_acerto.append(RandomForest.score(noclass_test, class_test))
  1316.        RandomForest_recall.append(recall_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1317.        RandomForest_precision.append(precision_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1318.        RandomForest_fscore.append(f1_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1319.        RandomForest_fim = time.time()
  1320.        RandomForest_tempo.append(RandomForest_fim - RandomForest_inicio)
  1321.  
  1322.  
  1323.    dataFrame128 = dataFrame128.sample(frac=1)
  1324.    print("Terminou a ", x)
  1325.    tempo2 = time.time()
  1326.    print("Tempo da rodada ", x, (tempo2 - tempo1) / 60)
  1327.  
  1328. tempofinal = time.time()
  1329.  
  1330. fun.tendencia_central('KNN k = 3', KNN_3_acerto, KNN_3_tempo)
  1331. fun.tendencia_central('KNN k = 5', KNN_5_acerto, KNN_5_tempo)
  1332. fun.tendencia_central('KNN k = 7', KNN_7_acerto, KNN_7_tempo)
  1333. fun.tendencia_central('KNN Ponderado k = 3', KNNP_3_acerto, KNNP_3_tempo)
  1334. fun.tendencia_central('KNN Ponderado k = 5', KNNP_5_acerto, KNNP_5_tempo)
  1335. fun.tendencia_central('KNN Ponderado k = 7', KNNP_7_acerto, KNNP_7_tempo)
  1336. fun.tendencia_central('Naïve Bayes', NaiveBayes_acerto, NaiveBayes_tempo)
  1337. fun.tendencia_central('Árvore de decisão', DecisionTree_acerto, DecisionTree_tempo)
  1338. fun.tendencia_central('MLP', MLP_acerto, MLP_tempo)
  1339. fun.tendencia_central('Regressão Logística', RegrLogistica_acerto, RegrLogistica_tempo)
  1340. fun.tendencia_central('SVM linear', SVMachine_L_acerto, SVMachine_L_tempo)
  1341. fun.tendencia_central('SVM RBF', SVMachine_RBF_acerto, SVMachine_RBF_tempo)
  1342. fun.tendencia_central('Random Forest', RandomForest_acerto,RandomForest_tempo)
  1343.  
  1344. Acuracia128 = [KNN_3_acerto,KNN_5_acerto,KNN_7_acerto,KNNP_3_acerto,KNNP_5_acerto,KNNP_7_acerto,
  1345.               NaiveBayes_acerto,DecisionTree_acerto,MLP_acerto,RegrLogistica_acerto,SVMachine_L_acerto,
  1346.               SVMachine_RBF_acerto,RandomForest_acerto]
  1347. Precision128 = [KNN_3_precision,KNN_5_precision,KNN_7_precision,KNNP_3_precision,KNNP_5_precision,KNNP_7_precision,
  1348.         NaiveBayes_precision,DecisionTree_precision,MLPP_precision,RegreLogistica_precision,SVMachine_L_precision,
  1349.         SVMachine_RBF_precision,RandomForest_precision]
  1350. Recall128 = [KNN_3_recall,KNN_5_recall,KNN_7_recall,KNNP_3_recall,KNNP_5_recall,KNNP_7_recall,NaiveBayes_recall,
  1351.          DecisionTree_recall,MLP_recall,RegrLogistica_recall,SVMachine_L_recall,SVMachine_RBF_recall,RandomForest_recall]
  1352.  
  1353. Fscore128 = [KNN_3_fscore,KNN_5_fscore,KNN_7_fscore,KNNP_3_fscore,KNN_5_fscore,KNN_7_fscore,NaiveBayes_fscore,DecisionTree_fscore,
  1354.          MLP_fscore,RegrLogistica_fscore,SVMachine_L_fscore, SVMachine_RBF_fscore,RandomForest_fscore]
  1355.  
  1356. Acuracia128 = pd.DataFrame(Acuracia128)
  1357. Precision128 = pd.DataFrame(Precision128)
  1358. Recall128 = pd.DataFrame(Recall128)
  1359. Fscore128 = pd.DataFrame(Fscore128)
  1360.  
  1361. Acuracia128.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Acuracia128.csv')
  1362. Precision128.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Precision128.csv')
  1363. Recall128.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Recall128.csv')
  1364. Fscore128.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Fscore128.csv')
  1365.  
  1366. mediasacuracias = {
  1367.  
  1368.  
  1369.    "KNN Ponderado k = 3": np.mean(KNNP_3_acerto),
  1370.    "KNN Ponderado k = 5": np.mean(KNNP_5_acerto),
  1371.    "KNN Ponderado k = 7": np.mean(KNNP_7_acerto),
  1372.    "Naive Bayes": np.mean(NaiveBayes_acerto),
  1373.    "KNN k = 3": np.mean(KNN_3_acerto),
  1374.    "KNN k = 5": np.mean(KNN_5_acerto),
  1375.    "KNN k = 7": np.mean(KNN_7_acerto),
  1376.    "Decision Tree": np.mean(DecisionTree_acerto),
  1377.    "SVM Linear": np.mean(SVMachine_L_acerto),
  1378.    "SVM RBF": np.mean(SVMachine_RBF_acerto),
  1379.    "Regressao Logistica": np.mean(RegrLogistica_acerto),
  1380.    "MLP": np.mean(MLP_acerto),
  1381.    "Random Forest": np.mean(RandomForest_acerto)
  1382. }
  1383.  
  1384.  
  1385.  
  1386. mediasacuracias = sorted(mediasacuracias.items(),
  1387.                         key=lambda x: x[1])
  1388. print(mediasacuracias)
  1389. print("Tempo total: ", (tempofinal - tempoinicial) / 60)
  1390.  
  1391.  
  1392. ## USANDO A BASE COM IMAGENS DE 64x64
  1393.  
  1394.  
  1395.  
  1396. print('comecou o K fold')
  1397.  
  1398. tempoinicial = time.time()
  1399.  
  1400. for x in range(0, 10):
  1401.  
  1402.    tempo1 = time.time()
  1403.    cols = list(dataFrame64.columns)
  1404.    cols.remove('Classe')
  1405.    df_images_noclass = dataFrame64[cols]
  1406.    df_images_class = dataFrame64['Classe']
  1407.    c = kfold.split(dataFrame64)
  1408.  
  1409.    for train_index, test_index in c:
  1410.  
  1411.        noclass_train, noclass_test = df_images_noclass.iloc[train_index], df_images_noclass.iloc[test_index]
  1412.        class_train, class_test = df_images_class.iloc[train_index], df_images_class.iloc[test_index]
  1413.  
  1414.        KNN3_inicio = time.time()
  1415.        KNN_3.fit(noclass_train, class_train)
  1416.        KNN_3_acerto.append(KNN_3.score(noclass_test, class_test))
  1417.        KNN_3_recall. append(recall_score(class_test, KNN_3.predict(noclass_test),average='weighted'))
  1418.        KNN_3_precision.append(precision_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1419.        KNN_3_fscore.append(f1_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1420.        KNN3_fim = time.time()
  1421.        KNN_3_tempo.append(KNN3_fim - KNN3_inicio)
  1422.  
  1423.        KNN5_inicio = time.time()
  1424.        KNN_5.fit(noclass_train, class_train)
  1425.        KNN_5_acerto.append(KNN_5.score(noclass_test, class_test))
  1426.        KNN_5_recall.append(recall_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1427.        KNN_5_precision.append(precision_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1428.        KNN_5_fscore.append(f1_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1429.        KNN5_fim = time.time()
  1430.        KNN_5_tempo.append(KNN5_fim - KNN5_inicio)
  1431.  
  1432.  
  1433.  
  1434.        KNN7_inicio = time.time()
  1435.        KNN_7.fit(noclass_train, class_train)
  1436.        KNN_7_acerto.append(KNN_3.score(noclass_test, class_test))
  1437.        KNN_7_precision.append(precision_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1438.        KNN_7_recall.append(recall_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1439.        KNN_7_fscore.append(f1_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1440.        KNN7_fim = time.time()
  1441.        KNN_7_tempo.append(KNN3_fim - KNN3_inicio)
  1442.  
  1443.        KNNP3_inicio = time.time()
  1444.        KNNP_3.fit(noclass_train, class_train)
  1445.        KNNP_3_acerto.append(KNNP_3.score(noclass_test, class_test))
  1446.        KNNP_3_precision.append(precision_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1447.        KNNP_3_recall.append(recall_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1448.        KNNP_3_fscore.append(f1_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1449.        KNNP3_fim = time.time()
  1450.        KNNP_3_tempo.append(KNNP3_fim - KNNP3_inicio)
  1451.  
  1452.        KNNP5_inicio = time.time()
  1453.        KNNP_5.fit(noclass_train, class_train)
  1454.        KNNP_5_acerto.append(KNNP_5.score(noclass_test, class_test))
  1455.        KNNP_5_precision.append(precision_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1456.        KNNP_5_recall.append(recall_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1457.        KNNP_5_fscore.append(f1_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1458.        KNNP5_fim = time.time()
  1459.        KNNP_5_tempo.append(KNNP5_fim - KNNP5_inicio)
  1460.  
  1461.        KNNP7_inicio = time.time()
  1462.        KNNP_7.fit(noclass_train, class_train)
  1463.        KNNP_7_acerto.append(KNNP_7.score(noclass_test, class_test))
  1464.        KNNP_7_precision.append(precision_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1465.        KNNP_7_recall.append(recall_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1466.        KNNP_7_fscore.append(f1_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1467.        KNNP7_fim = time.time()
  1468.        KNNP_7_tempo.append(KNNP7_fim - KNNP7_inicio)
  1469.  
  1470.        NaiveBayes_inicio = time.time()
  1471.        NaiveBayes.fit(noclass_train, class_train)
  1472.        NaiveBayes_acerto.append(NaiveBayes.score(noclass_test, class_test))
  1473.        NaiveBayes_precision.append(precision_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1474.        NaiveBayes_recall.append(recall_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1475.        NaiveBayes_fscore.append(f1_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1476.        NaiveBayes_fim = time.time()
  1477.        NaiveBayes_tempo.append(NaiveBayes_fim - NaiveBayes_inicio)
  1478.  
  1479.        DecisionTree_inicio = time.time()
  1480.        DecisionTree.fit(noclass_train, class_train)
  1481.        DecisionTree_acerto.append(DecisionTree.score(noclass_test, class_test))
  1482.        DecisionTree_precision.append(precision_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1483.        DecisionTree_recall.append(recall_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1484.        DecisionTree_fscore.append(f1_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1485.        DecisionTree_fim = time.time()
  1486.        DecisionTree_tempo.append(DecisionTree_fim - DecisionTree_inicio)
  1487.  
  1488.        SVMachine_L_inicio = time.time()
  1489.        SVMachine_L.fit(noclass_train, class_train)
  1490.        SVMachine_L_acerto.append(SVMachine_L.score(noclass_test, class_test))
  1491.        SVMachine_L_precision.append(precision_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1492.        SVMachine_L_recall.append(recall_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1493.        SVMachine_L_fscore.append(f1_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1494.        SVMachine_L_fim = time.time()
  1495.        SVMachine_L_tempo.append(SVMachine_L_fim - SVMachine_L_inicio)
  1496.  
  1497.        SVMachine_RBF_inicio = time.time()
  1498.        SVMachine_RBF.fit(noclass_train, class_train)
  1499.        SVMachine_RBF_acerto.append(SVMachine_RBF.score(noclass_test, class_test))
  1500.        SVMachine_RBF_recall.append(recall_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1501.        SVMachine_RBF_precision.append(precision_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1502.        SVMachine_RBF_fscore.append(f1_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1503.        SVMachine_RBF_fim = time.time()
  1504.        SVMachine_RBF_tempo.append(SVMachine_RBF_fim - SVMachine_RBF_inicio)
  1505.  
  1506.        RegrLogistica_inicio = time.time()
  1507.        RegrLogistica.fit(noclass_train, class_train)
  1508.        RegrLogistica_acerto.append(RegrLogistica.score(noclass_test, class_test))
  1509.        RegreLogistica_precision.append(precision_score(class_test, (RegrLogistica.predict(noclass_test)),average='weighted'))
  1510.        RegrLogistica_recall.append(recall_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1511.        RegrLogistica_fscore.append(f1_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1512.        RegrLogistica_fim = time.time()
  1513.        RegrLogistica_tempo.append(RegrLogistica_fim - RegrLogistica_inicio)
  1514.  
  1515.        MLP_inicio = time.time()
  1516.        MLP.fit(noclass_train, class_train)
  1517.        MLP_acerto.append(MLP.score(noclass_test, class_test))
  1518.        MLPP_precision.append(precision_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1519.        MLP_recall.append(recall_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1520.        MLP_fscore.append(f1_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1521.        MLP_fim = time.time()
  1522.        MLP_tempo.append(MLP_fim - MLP_inicio)
  1523.  
  1524.        RandomForest_inicio = time.time()
  1525.        RandomForest.fit(noclass_train, class_train)
  1526.        RandomForest_acerto.append(RandomForest.score(noclass_test, class_test))
  1527.        RandomForest_recall.append(recall_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1528.        RandomForest_precision.append(precision_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1529.        RandomForest_fscore.append(f1_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1530.        RandomForest_fim = time.time()
  1531.        RandomForest_tempo.append(RandomForest_fim - RandomForest_inicio)
  1532.  
  1533.  
  1534.    dataFrame64 = dataFrame64.sample(frac=1)
  1535.    print("Terminou a ", x)
  1536.    tempo2 = time.time()
  1537.    print("Tempo da rodada ", x, (tempo2 - tempo1) / 60)
  1538.  
  1539. tempofinal = time.time()
  1540.  
  1541. fun.tendencia_central('KNN k = 3', KNN_3_acerto, KNN_3_tempo)
  1542. fun.tendencia_central('KNN k = 5', KNN_5_acerto, KNN_5_tempo)
  1543. fun.tendencia_central('KNN k = 7', KNN_7_acerto, KNN_7_tempo)
  1544. fun.tendencia_central('KNN Ponderado k = 3', KNNP_3_acerto, KNNP_3_tempo)
  1545. fun.tendencia_central('KNN Ponderado k = 5', KNNP_5_acerto, KNNP_5_tempo)
  1546. fun.tendencia_central('KNN Ponderado k = 7', KNNP_7_acerto, KNNP_7_tempo)
  1547. fun.tendencia_central('Naïve Bayes', NaiveBayes_acerto, NaiveBayes_tempo)
  1548. fun.tendencia_central('Árvore de decisão', DecisionTree_acerto, DecisionTree_tempo)
  1549. fun.tendencia_central('MLP', MLP_acerto, MLP_tempo)
  1550. fun.tendencia_central('Regressão Logística', RegrLogistica_acerto, RegrLogistica_tempo)
  1551. fun.tendencia_central('SVM linear', SVMachine_L_acerto, SVMachine_L_tempo)
  1552. fun.tendencia_central('SVM RBF', SVMachine_RBF_acerto, SVMachine_RBF_tempo)
  1553. fun.tendencia_central('Random Forest', RandomForest_acerto,RandomForest_tempo)
  1554.  
  1555. mediasacuracias = {
  1556.  
  1557.  
  1558.    "KNN Ponderado k = 3": np.mean(KNNP_3_acerto),
  1559.    "KNN Ponderado k = 5": np.mean(KNNP_5_acerto),
  1560.    "KNN Ponderado k = 7": np.mean(KNNP_7_acerto),
  1561.    "Naive Bayes": np.mean(NaiveBayes_acerto),
  1562.    "KNN k = 3": np.mean(KNN_3_acerto),
  1563.    "KNN k = 5": np.mean(KNN_5_acerto),
  1564.    "KNN k = 7": np.mean(KNN_7_acerto),
  1565.    "Decision Tree": np.mean(DecisionTree_acerto),
  1566.    "SVM Linear": np.mean(SVMachine_L_acerto),
  1567.    "SVM RBF": np.mean(SVMachine_RBF_acerto),
  1568.    "Regressao Logistica": np.mean(RegrLogistica_acerto),
  1569.    "MLP": np.mean(MLP_acerto),
  1570.    "Random Forest": np.mean(RandomForest_acerto)
  1571. }
  1572.  
  1573. mediasacuracias = sorted(mediasacuracias.items(),
  1574.                         key=lambda x: x[1])
  1575. print(mediasacuracias)
  1576. print("Tempo total: ", (tempofinal - tempoinicial) / 60)
  1577. Acuracia64 = [KNN_3_acerto,KNN_5_acerto,KNN_7_acerto,KNNP_3_acerto,KNNP_5_acerto,KNNP_7_acerto,
  1578.               NaiveBayes_acerto,DecisionTree_acerto,MLP_acerto,RegrLogistica_acerto,SVMachine_L_acerto,
  1579.               SVMachine_RBF_acerto,RandomForest_acerto]
  1580. Precision64 = [KNN_3_precision,KNN_5_precision,KNN_7_precision,KNNP_3_precision,KNNP_5_precision,KNNP_7_precision,
  1581.         NaiveBayes_precision,DecisionTree_precision,MLPP_precision,RegreLogistica_precision,SVMachine_L_precision,
  1582.         SVMachine_RBF_precision,RandomForest_precision]
  1583. Recall64 = [KNN_3_recall,KNN_5_recall,KNN_7_recall,KNNP_3_recall,KNNP_5_recall,KNNP_7_recall,NaiveBayes_recall,
  1584.          DecisionTree_recall,MLP_recall,RegrLogistica_recall,SVMachine_L_recall,SVMachine_RBF_recall,RandomForest_recall]
  1585.  
  1586. Fscore64 = [KNN_3_fscore,KNN_5_fscore,KNN_7_fscore,KNNP_3_fscore,KNN_5_fscore,KNN_7_fscore,NaiveBayes_fscore,DecisionTree_fscore,
  1587.          MLP_fscore,RegrLogistica_fscore,SVMachine_L_fscore, SVMachine_RBF_fscore,RandomForest_fscore]
  1588.  
  1589. Acuracia64 = pd.DataFrame(Acuracia64)
  1590. Precision64 = pd.DataFrame(Precision64)
  1591. Recall64 = pd.DataFrame(Recall64)
  1592. Fscore64 = pd.DataFrame(Fscore64)
  1593.  
  1594. Acuracia64.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Acuracia64.csv')
  1595. Precision64.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Precision64.csv')
  1596. Recall64.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Recall64.csv')
  1597. Fscore64.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Fscore64.csv')
  1598.  
  1599.  
  1600.  
  1601.  
  1602. # USANDO BASE DE 32x32
  1603.  
  1604.  
  1605.  
  1606. print('comecou o K fold')
  1607.  
  1608. tempoinicial = time.time()
  1609.  
  1610. for x in range(0, 10):
  1611.  
  1612.    tempo1 = time.time()
  1613.    cols = list(dataFrame32.columns)
  1614.    cols.remove('Classe')
  1615.    df_images_noclass = dataFrame32[cols]
  1616.    df_images_class = dataFrame32['Classe']
  1617.    c = kfold.split(dataFrame32)
  1618.  
  1619.    for train_index, test_index in c:
  1620.  
  1621.        noclass_train, noclass_test = df_images_noclass.iloc[train_index], df_images_noclass.iloc[test_index]
  1622.        class_train, class_test = df_images_class.iloc[train_index], df_images_class.iloc[test_index]
  1623.  
  1624.        KNN3_inicio = time.time()
  1625.        KNN_3.fit(noclass_train, class_train)
  1626.        KNN_3_acerto.append(KNN_3.score(noclass_test, class_test))
  1627.        KNN_3_recall. append(recall_score(class_test, KNN_3.predict(noclass_test),average='weighted'))
  1628.        KNN_3_precision.append(precision_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1629.        KNN_3_fscore.append(f1_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1630.        KNN3_fim = time.time()
  1631.        KNN_3_tempo.append(KNN3_fim - KNN3_inicio)
  1632.  
  1633.        KNN5_inicio = time.time()
  1634.        KNN_5.fit(noclass_train, class_train)
  1635.        KNN_5_acerto.append(KNN_5.score(noclass_test, class_test))
  1636.        KNN_5_recall.append(recall_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1637.        KNN_5_precision.append(precision_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1638.        KNN_5_fscore.append(f1_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1639.        KNN5_fim = time.time()
  1640.        KNN_5_tempo.append(KNN5_fim - KNN5_inicio)
  1641.  
  1642.  
  1643.  
  1644.        KNN7_inicio = time.time()
  1645.        KNN_7.fit(noclass_train, class_train)
  1646.        KNN_7_acerto.append(KNN_3.score(noclass_test, class_test))
  1647.        KNN_7_precision.append(precision_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1648.        KNN_7_recall.append(recall_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1649.        KNN_7_fscore.append(f1_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1650.        KNN7_fim = time.time()
  1651.        KNN_7_tempo.append(KNN3_fim - KNN3_inicio)
  1652.  
  1653.        KNNP3_inicio = time.time()
  1654.        KNNP_3.fit(noclass_train, class_train)
  1655.        KNNP_3_acerto.append(KNNP_3.score(noclass_test, class_test))
  1656.        KNNP_3_precision.append(precision_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1657.        KNNP_3_recall.append(recall_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1658.        KNNP_3_fscore.append(f1_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1659.        KNNP3_fim = time.time()
  1660.        KNNP_3_tempo.append(KNNP3_fim - KNNP3_inicio)
  1661.  
  1662.        KNNP5_inicio = time.time()
  1663.        KNNP_5.fit(noclass_train, class_train)
  1664.        KNNP_5_acerto.append(KNNP_5.score(noclass_test, class_test))
  1665.        KNNP_5_precision.append(precision_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1666.        KNNP_5_recall.append(recall_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1667.        KNNP_5_fscore.append(f1_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1668.        KNNP5_fim = time.time()
  1669.        KNNP_5_tempo.append(KNNP5_fim - KNNP5_inicio)
  1670.  
  1671.        KNNP7_inicio = time.time()
  1672.        KNNP_7.fit(noclass_train, class_train)
  1673.        KNNP_7_acerto.append(KNNP_7.score(noclass_test, class_test))
  1674.        KNNP_7_precision.append(precision_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1675.        KNNP_7_recall.append(recall_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1676.        KNNP_7_fscore.append(f1_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1677.        KNNP7_fim = time.time()
  1678.        KNNP_7_tempo.append(KNNP7_fim - KNNP7_inicio)
  1679.  
  1680.        NaiveBayes_inicio = time.time()
  1681.        NaiveBayes.fit(noclass_train, class_train)
  1682.        NaiveBayes_acerto.append(NaiveBayes.score(noclass_test, class_test))
  1683.        NaiveBayes_precision.append(precision_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1684.        NaiveBayes_recall.append(recall_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1685.        NaiveBayes_fscore.append(f1_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1686.        NaiveBayes_fim = time.time()
  1687.        NaiveBayes_tempo.append(NaiveBayes_fim - NaiveBayes_inicio)
  1688.  
  1689.        DecisionTree_inicio = time.time()
  1690.        DecisionTree.fit(noclass_train, class_train)
  1691.        DecisionTree_acerto.append(DecisionTree.score(noclass_test, class_test))
  1692.        DecisionTree_precision.append(precision_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1693.        DecisionTree_recall.append(recall_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1694.        DecisionTree_fscore.append(f1_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1695.        DecisionTree_fim = time.time()
  1696.        DecisionTree_tempo.append(DecisionTree_fim - DecisionTree_inicio)
  1697.  
  1698.        SVMachine_L_inicio = time.time()
  1699.        SVMachine_L.fit(noclass_train, class_train)
  1700.        SVMachine_L_acerto.append(SVMachine_L.score(noclass_test, class_test))
  1701.        SVMachine_L_precision.append(precision_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1702.        SVMachine_L_recall.append(recall_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1703.        SVMachine_L_fscore.append(f1_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1704.        SVMachine_L_fim = time.time()
  1705.        SVMachine_L_tempo.append(SVMachine_L_fim - SVMachine_L_inicio)
  1706.  
  1707.        SVMachine_RBF_inicio = time.time()
  1708.        SVMachine_RBF.fit(noclass_train, class_train)
  1709.        SVMachine_RBF_acerto.append(SVMachine_RBF.score(noclass_test, class_test))
  1710.        SVMachine_RBF_recall.append(recall_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1711.        SVMachine_RBF_precision.append(precision_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1712.        SVMachine_RBF_fscore.append(f1_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1713.        SVMachine_RBF_fim = time.time()
  1714.        SVMachine_RBF_tempo.append(SVMachine_RBF_fim - SVMachine_RBF_inicio)
  1715.  
  1716.        RegrLogistica_inicio = time.time()
  1717.        RegrLogistica.fit(noclass_train, class_train)
  1718.        RegrLogistica_acerto.append(RegrLogistica.score(noclass_test, class_test))
  1719.        RegreLogistica_precision.append(precision_score(class_test, (RegrLogistica.predict(noclass_test)),average='weighted'))
  1720.        RegrLogistica_recall.append(recall_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1721.        RegrLogistica_fscore.append(f1_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1722.        RegrLogistica_fim = time.time()
  1723.        RegrLogistica_tempo.append(RegrLogistica_fim - RegrLogistica_inicio)
  1724.  
  1725.        MLP_inicio = time.time()
  1726.        MLP.fit(noclass_train, class_train)
  1727.        MLP_acerto.append(MLP.score(noclass_test, class_test))
  1728.        MLPP_precision.append(precision_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1729.        MLP_recall.append(recall_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1730.        MLP_fscore.append(f1_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1731.        MLP_fim = time.time()
  1732.        MLP_tempo.append(MLP_fim - MLP_inicio)
  1733.  
  1734.        RandomForest_inicio = time.time()
  1735.        RandomForest.fit(noclass_train, class_train)
  1736.        RandomForest_acerto.append(RandomForest.score(noclass_test, class_test))
  1737.        RandomForest_recall.append(recall_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1738.        RandomForest_precision.append(precision_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1739.        RandomForest_fscore.append(f1_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1740.        RandomForest_fim = time.time()
  1741.        RandomForest_tempo.append(RandomForest_fim - RandomForest_inicio)
  1742.  
  1743.  
  1744.    dataFrame32 = dataFrame32.sample(frac=1)
  1745.    print("Terminou a ", x)
  1746.    tempo2 = time.time()
  1747.    print("Tempo da rodada ", x, (tempo2 - tempo1) / 60)
  1748.  
  1749. tempofinal = time.time()
  1750.  
  1751. fun.tendencia_central('KNN k = 3', KNN_3_acerto, KNN_3_tempo)
  1752. fun.tendencia_central('KNN k = 5', KNN_5_acerto, KNN_5_tempo)
  1753. fun.tendencia_central('KNN k = 7', KNN_7_acerto, KNN_7_tempo)
  1754. fun.tendencia_central('KNN Ponderado k = 3', KNNP_3_acerto, KNNP_3_tempo)
  1755. fun.tendencia_central('KNN Ponderado k = 5', KNNP_5_acerto, KNNP_5_tempo)
  1756. fun.tendencia_central('KNN Ponderado k = 7', KNNP_7_acerto, KNNP_7_tempo)
  1757. fun.tendencia_central('Naïve Bayes', NaiveBayes_acerto, NaiveBayes_tempo)
  1758. fun.tendencia_central('Árvore de decisão', DecisionTree_acerto, DecisionTree_tempo)
  1759. fun.tendencia_central('MLP', MLP_acerto, MLP_tempo)
  1760. fun.tendencia_central('Regressão Logística', RegrLogistica_acerto, RegrLogistica_tempo)
  1761. fun.tendencia_central('SVM linear', SVMachine_L_acerto, SVMachine_L_tempo)
  1762. fun.tendencia_central('SVM RBF', SVMachine_RBF_acerto, SVMachine_RBF_tempo)
  1763. fun.tendencia_central('Random Forest', RandomForest_acerto,RandomForest_tempo)
  1764.  
  1765. mediasacuracias = {
  1766.  
  1767.  
  1768.    "KNN Ponderado k = 3": np.mean(KNNP_3_acerto),
  1769.    "KNN Ponderado k = 5": np.mean(KNNP_5_acerto),
  1770.    "KNN Ponderado k = 7": np.mean(KNNP_7_acerto),
  1771.    "Naive Bayes": np.mean(NaiveBayes_acerto),
  1772.    "KNN k = 3": np.mean(KNN_3_acerto),
  1773.    "KNN k = 5": np.mean(KNN_5_acerto),
  1774.    "KNN k = 7": np.mean(KNN_7_acerto),
  1775.    "Decision Tree": np.mean(DecisionTree_acerto),
  1776.    "SVM Linear": np.mean(SVMachine_L_acerto),
  1777.    "SVM RBF": np.mean(SVMachine_RBF_acerto),
  1778.    "Regressao Logistica": np.mean(RegrLogistica_acerto),
  1779.    "MLP": np.mean(MLP_acerto),
  1780.    "Random Forest": np.mean(RandomForest_acerto)
  1781. }
  1782.  
  1783. mediasacuracias = sorted(mediasacuracias.items(),
  1784.                         key=lambda x: x[1])
  1785. print(mediasacuracias)
  1786. print("Tempo total: ", (tempofinal - tempoinicial) / 60)
  1787. Acuracia32 = [KNN_3_acerto,KNN_5_acerto,KNN_7_acerto,KNNP_3_acerto,KNNP_5_acerto,KNNP_7_acerto,
  1788.               NaiveBayes_acerto,DecisionTree_acerto,MLP_acerto,RegrLogistica_acerto,SVMachine_L_acerto,
  1789.               SVMachine_RBF_acerto,RandomForest_acerto]
  1790. Precision32 = [KNN_3_precision,KNN_5_precision,KNN_7_precision,KNNP_3_precision,KNNP_5_precision,KNNP_7_precision,
  1791.         NaiveBayes_precision,DecisionTree_precision,MLPP_precision,RegreLogistica_precision,SVMachine_L_precision,
  1792.         SVMachine_RBF_precision,RandomForest_precision]
  1793. Recall32 = [KNN_3_recall,KNN_5_recall,KNN_7_recall,KNNP_3_recall,KNNP_5_recall,KNNP_7_recall,NaiveBayes_recall,
  1794.          DecisionTree_recall,MLP_recall,RegrLogistica_recall,SVMachine_L_recall,SVMachine_RBF_recall,RandomForest_recall]
  1795.  
  1796. Fscore32 = [KNN_3_fscore,KNN_5_fscore,KNN_7_fscore,KNNP_3_fscore,KNN_5_fscore,KNN_7_fscore,NaiveBayes_fscore,DecisionTree_fscore,
  1797.          MLP_fscore,RegrLogistica_fscore,SVMachine_L_fscore, SVMachine_RBF_fscore,RandomForest_fscore]
  1798.  
  1799. Acuracia32 = pd.DataFrame(Acuracia32)
  1800. Precision32 = pd.DataFrame(Precision32)
  1801. Recall32 = pd.DataFrame(Recall32)
  1802. Fscore32 = pd.DataFrame(Fscore32)
  1803.  
  1804. Acuracia32.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Acuracia32.csv')
  1805. Precision32.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Precision32.csv')
  1806. Recall32.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Recall32.csv')
  1807. Fscore32.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Fscore32.csv')
  1808.  
  1809.  
  1810. '''
  1811.  
  1812. '''
  1813.  
  1814. # USANDO A BASE COM IMAGENS 16x16
  1815.  
  1816. print('comecou o K fold')
  1817.  
  1818. tempoinicial = time.time()
  1819.  
  1820. for x in range(0, 10):
  1821.  
  1822.    tempo1 = time.time()
  1823.    cols = list(dataFrame16.columns)
  1824.    cols.remove('Classe')
  1825.    df_images_noclass = dataFrame16[cols]
  1826.    df_images_class = dataFrame16['Classe']
  1827.    c = kfold.split(dataFrame16)
  1828.  
  1829.    for train_index, test_index in c:
  1830.  
  1831.        noclass_train, noclass_test = df_images_noclass.iloc[train_index], df_images_noclass.iloc[test_index]
  1832.        class_train, class_test = df_images_class.iloc[train_index], df_images_class.iloc[test_index]
  1833.  
  1834.        KNN3_inicio = time.time()
  1835.        KNN_3.fit(noclass_train, class_train)
  1836.        KNN_3_acerto.append(KNN_3.score(noclass_test, class_test))
  1837.        KNN_3_recall. append(recall_score(class_test, KNN_3.predict(noclass_test),average='weighted'))
  1838.        KNN_3_precision.append(precision_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1839.        KNN_3_fscore.append(f1_score(class_test,KNN_3.predict(noclass_test),average='weighted'))
  1840.        KNN3_fim = time.time()
  1841.        KNN_3_tempo.append(KNN3_fim - KNN3_inicio)
  1842.  
  1843.        KNN5_inicio = time.time()
  1844.        KNN_5.fit(noclass_train, class_train)
  1845.        KNN_5_acerto.append(KNN_5.score(noclass_test, class_test))
  1846.        KNN_5_recall.append(recall_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1847.        KNN_5_precision.append(precision_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1848.        KNN_5_fscore.append(f1_score(class_test,KNN_5.predict(noclass_test),average='weighted'))
  1849.        KNN5_fim = time.time()
  1850.        KNN_5_tempo.append(KNN5_fim - KNN5_inicio)
  1851.  
  1852.  
  1853.        KNN7_inicio = time.time()
  1854.        KNN_7.fit(noclass_train, class_train)
  1855.        KNN_7_acerto.append(KNN_3.score(noclass_test, class_test))
  1856.        KNN_7_precision.append(precision_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1857.        KNN_7_recall.append(recall_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1858.        KNN_7_fscore.append(f1_score(class_test,KNN_7.predict(noclass_test),average='weighted'))
  1859.        KNN7_fim = time.time()
  1860.        KNN_7_tempo.append(KNN3_fim - KNN3_inicio)
  1861.  
  1862.        KNNP3_inicio = time.time()
  1863.        KNNP_3.fit(noclass_train, class_train)
  1864.        KNNP_3_acerto.append(KNNP_3.score(noclass_test, class_test))
  1865.        KNNP_3_precision.append(precision_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1866.        KNNP_3_recall.append(recall_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1867.        KNNP_3_fscore.append(f1_score(class_test,KNNP_3.predict(noclass_test),average='weighted'))
  1868.        KNNP3_fim = time.time()
  1869.        KNNP_3_tempo.append(KNNP3_fim - KNNP3_inicio)
  1870.  
  1871.        KNNP5_inicio = time.time()
  1872.        KNNP_5.fit(noclass_train, class_train)
  1873.        KNNP_5_acerto.append(KNNP_5.score(noclass_test, class_test))
  1874.        KNNP_5_precision.append(precision_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1875.        KNNP_5_recall.append(recall_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1876.        KNNP_5_fscore.append(f1_score(class_test,KNNP_5.predict(noclass_test),average='weighted'))
  1877.        KNNP5_fim = time.time()
  1878.        KNNP_5_tempo.append(KNNP5_fim - KNNP5_inicio)
  1879.  
  1880.        KNNP7_inicio = time.time()
  1881.        KNNP_7.fit(noclass_train, class_train)
  1882.        KNNP_7_acerto.append(KNNP_7.score(noclass_test, class_test))
  1883.        KNNP_7_precision.append(precision_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1884.        KNNP_7_recall.append(recall_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1885.        KNNP_7_fscore.append(f1_score(class_test,KNNP_7.predict(noclass_test),average='weighted'))
  1886.        KNNP7_fim = time.time()
  1887.        KNNP_7_tempo.append(KNNP7_fim - KNNP7_inicio)
  1888.  
  1889.        NaiveBayes_inicio = time.time()
  1890.        NaiveBayes.fit(noclass_train, class_train)
  1891.        NaiveBayes_acerto.append(NaiveBayes.score(noclass_test, class_test))
  1892.        NaiveBayes_precision.append(precision_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1893.        NaiveBayes_recall.append(recall_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1894.        NaiveBayes_fscore.append(f1_score(class_test,NaiveBayes.predict(noclass_test),average='weighted'))
  1895.        NaiveBayes_fim = time.time()
  1896.        NaiveBayes_tempo.append(NaiveBayes_fim - NaiveBayes_inicio)
  1897.  
  1898.        DecisionTree_inicio = time.time()
  1899.        DecisionTree.fit(noclass_train, class_train)
  1900.        DecisionTree_acerto.append(DecisionTree.score(noclass_test, class_test))
  1901.        DecisionTree_precision.append(precision_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1902.        DecisionTree_recall.append(recall_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1903.        DecisionTree_fscore.append(f1_score(class_test,DecisionTree.predict(noclass_test),average='weighted'))
  1904.        DecisionTree_fim = time.time()
  1905.        DecisionTree_tempo.append(DecisionTree_fim - DecisionTree_inicio)
  1906.  
  1907.        SVMachine_L_inicio = time.time()
  1908.        SVMachine_L.fit(noclass_train, class_train)
  1909.        SVMachine_L_acerto.append(SVMachine_L.score(noclass_test, class_test))
  1910.        SVMachine_L_precision.append(precision_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1911.        SVMachine_L_recall.append(recall_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1912.        SVMachine_L_fscore.append(f1_score(class_test,SVMachine_L.predict(noclass_test),average='weighted'))
  1913.        SVMachine_L_fim = time.time()
  1914.        SVMachine_L_tempo.append(SVMachine_L_fim - SVMachine_L_inicio)
  1915.  
  1916.        SVMachine_RBF_inicio = time.time()
  1917.        SVMachine_RBF.fit(noclass_train, class_train)
  1918.        SVMachine_RBF_acerto.append(SVMachine_RBF.score(noclass_test, class_test))
  1919.        SVMachine_RBF_recall.append(recall_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1920.        SVMachine_RBF_precision.append(precision_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1921.        SVMachine_RBF_fscore.append(f1_score(class_test,SVMachine_RBF.predict(noclass_test),average='weighted'))
  1922.        SVMachine_RBF_fim = time.time()
  1923.        SVMachine_RBF_tempo.append(SVMachine_RBF_fim - SVMachine_RBF_inicio)
  1924.  
  1925.        RegrLogistica_inicio = time.time()
  1926.        RegrLogistica.fit(noclass_train, class_train)
  1927.        RegrLogistica_acerto.append(RegrLogistica.score(noclass_test, class_test))
  1928.        RegreLogistica_precision.append(precision_score(class_test, (RegrLogistica.predict(noclass_test)),average='weighted'))
  1929.        RegrLogistica_recall.append(recall_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1930.        RegrLogistica_fscore.append(f1_score(class_test,RegrLogistica.predict(noclass_test),average='weighted'))
  1931.        RegrLogistica_fim = time.time()
  1932.        RegrLogistica_tempo.append(RegrLogistica_fim - RegrLogistica_inicio)
  1933.  
  1934.        MLP_inicio = time.time()
  1935.        MLP.fit(noclass_train, class_train)
  1936.        MLP_acerto.append(MLP.score(noclass_test, class_test))
  1937.        MLPP_precision.append(precision_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1938.        MLP_recall.append(recall_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1939.        MLP_fscore.append(f1_score(class_test,MLP.predict(noclass_test),average='weighted'))
  1940.        MLP_fim = time.time()
  1941.        MLP_tempo.append(MLP_fim - MLP_inicio)
  1942.  
  1943.        RandomForest_inicio = time.time()
  1944.        RandomForest.fit(noclass_train, class_train)
  1945.        RandomForest_acerto.append(RandomForest.score(noclass_test, class_test))
  1946.        RandomForest_recall.append(recall_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1947.        RandomForest_precision.append(precision_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1948.        RandomForest_fscore.append(f1_score(class_test,RandomForest.predict(noclass_test),average='weighted'))
  1949.        RandomForest_fim = time.time()
  1950.        RandomForest_tempo.append(RandomForest_fim - RandomForest_inicio)
  1951.  
  1952.  
  1953.    dataFrame16 = dataFrame16.sample(frac=1)
  1954.    print("Terminou a ", x)
  1955.    tempo2 = time.time()
  1956.    print("Tempo da rodada ", x, (tempo2 - tempo1) / 60)
  1957.  
  1958. tempofinal = time.time()
  1959.  
  1960. fun.tendencia_central('KNN k = 3', KNN_3_acerto, KNN_3_tempo)
  1961. fun.tendencia_central('KNN k = 5', KNN_5_acerto, KNN_5_tempo)
  1962. fun.tendencia_central('KNN k = 7', KNN_7_acerto, KNN_7_tempo)
  1963. fun.tendencia_central('KNN Ponderado k = 3', KNNP_3_acerto, KNNP_3_tempo)
  1964. fun.tendencia_central('KNN Ponderado k = 5', KNNP_5_acerto, KNNP_5_tempo)
  1965. fun.tendencia_central('KNN Ponderado k = 7', KNNP_7_acerto, KNNP_7_tempo)
  1966. fun.tendencia_central('Naïve Bayes', NaiveBayes_acerto, NaiveBayes_tempo)
  1967. fun.tendencia_central('Árvore de decisão', DecisionTree_acerto, DecisionTree_tempo)
  1968. fun.tendencia_central('MLP', MLP_acerto, MLP_tempo)
  1969. fun.tendencia_central('Regressão Logística', RegrLogistica_acerto, RegrLogistica_tempo)
  1970. fun.tendencia_central('SVM linear', SVMachine_L_acerto, SVMachine_L_tempo)
  1971. fun.tendencia_central('SVM RBF', SVMachine_RBF_acerto, SVMachine_RBF_tempo)
  1972. fun.tendencia_central('Random Forest', RandomForest_acerto,RandomForest_tempo)
  1973.  
  1974. mediasacuracias = {
  1975.  
  1976.  
  1977.    "KNN Ponderado k = 3": np.mean(KNNP_3_acerto),
  1978.    "KNN Ponderado k = 5": np.mean(KNNP_5_acerto),
  1979.    "KNN Ponderado k = 7": np.mean(KNNP_7_acerto),
  1980.    "Naive Bayes": np.mean(NaiveBayes_acerto),
  1981.    "KNN k = 3": np.mean(KNN_3_acerto),
  1982.    "KNN k = 5": np.mean(KNN_5_acerto),
  1983.    "KNN k = 7": np.mean(KNN_7_acerto),
  1984.    "Decision Tree": np.mean(DecisionTree_acerto),
  1985.    "SVM Linear": np.mean(SVMachine_L_acerto),
  1986.    "SVM RBF": np.mean(SVMachine_RBF_acerto),
  1987.    "Regressao Logistica": np.mean(RegrLogistica_acerto),
  1988.    "MLP": np.mean(MLP_acerto),
  1989.    "Random Forest": np.mean(RandomForest_acerto)
  1990. }
  1991.  
  1992. mediasacuracias = sorted(mediasacuracias.items(),
  1993.                         key=lambda x: x[1])
  1994. print(mediasacuracias)
  1995. print("Tempo total: ", (tempofinal - tempoinicial) / 60)
  1996. Acuracia16 = [KNN_3_acerto,KNN_5_acerto,KNN_7_acerto,KNNP_3_acerto,KNNP_5_acerto,KNNP_7_acerto,
  1997.               NaiveBayes_acerto,DecisionTree_acerto,MLP_acerto,RegrLogistica_acerto,SVMachine_L_acerto,
  1998.               SVMachine_RBF_acerto,RandomForest_acerto]
  1999. Precision16 = [KNN_3_precision,KNN_5_precision,KNN_7_precision,KNNP_3_precision,KNNP_5_precision,KNNP_7_precision,
  2000.         NaiveBayes_precision,DecisionTree_precision,MLPP_precision,RegreLogistica_precision,SVMachine_L_precision,
  2001.         SVMachine_RBF_precision,RandomForest_precision]
  2002. Recall16 = [KNN_3_recall,KNN_5_recall,KNN_7_recall,KNNP_3_recall,KNNP_5_recall,KNNP_7_recall,NaiveBayes_recall,
  2003.          DecisionTree_recall,MLP_recall,RegrLogistica_recall,SVMachine_L_recall,SVMachine_RBF_recall,RandomForest_recall]
  2004.  
  2005. Fscore16 = [KNN_3_fscore,KNN_5_fscore,KNN_7_fscore,KNNP_3_fscore,KNN_5_fscore,KNN_7_fscore,NaiveBayes_fscore,DecisionTree_fscore,
  2006.          MLP_fscore,RegrLogistica_fscore,SVMachine_L_fscore, SVMachine_RBF_fscore,RandomForest_fscore]
  2007.  
  2008. Acuracia16 = pd.DataFrame(Acuracia16)
  2009. Precision16 = pd.DataFrame(Precision16)
  2010. Recall16 = pd.DataFrame(Recall16)
  2011. Fscore16 = pd.DataFrame(Fscore16)
  2012.  
  2013. Acuracia16.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Acuracia16.csv')
  2014. Precision16.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Precision16.csv')
  2015. Recall16.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Recall16.csv')
  2016. Fscore16.to_csv('C:/Users/Auricelia/Desktop/DataSetsML/Fscore16.csv')
  2017. '''
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!