API tools faq
paste
Advertisement
dfracz3

HW 8 Neural Netowrks second 3 ii dan's version

dfracz3
Mar 19th, 2018
78
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 7.17 KB | None | 0 0
raw download clone embed print report
  1. #
  2. ## -*- coding: utf-8 -*-
  3. #"""
  4. #Created on Sun Mar 11 15:41:00 2018
  5. #
  6. #@author: UIMCC-Staff
  7. #"""
  8. #
  9. ##Libraries needed to run the tool
  10. #import numpy as np
  11. #import pandas as pd
  12. #from sklearn.neural_network import MLPRegressor
  13. #from sklearn.neural_network import MLPClassifier
  14. #from sklearn import preprocessing #to normalize the values
  15. #from sklearn.model_selection import train_test_split
  16. #from sklearn.model_selection import cross_val_score
  17. #import matplotlib.pyplot as plt
  18. #
  19. ##Ask for file name and read the file
  20. ##file_name = raw_input("Name of file:")
  21. #file_name = 'forestfiresw8'
  22. #data = pd.read_csv(file_name + '.csv', header=0)
  23. #
  24. #analysis_type = input("Analysis Type 'R' or 'C': ")
  25. #
  26. ##Print number of rows and colums read
  27. #print("{0} rows and {1} columns".format(len(data.index), len(data.columns)))
  28. #print("")
  29. #
  30. #data["month"] = data["month"].replace("jan",1)
  31. #data["month"] = data["month"].replace("feb",2)
  32. #data["month"] = data["month"].replace("mar",3)
  33. #data["month"] = data["month"].replace("apr",4)
  34. #data["month"] = data["month"].replace("may",5)
  35. #data["month"] = data["month"].replace("jun",6)
  36. #data["month"] = data["month"].replace("jul",7)
  37. #data["month"] = data["month"].replace("aug",8)
  38. #data["month"] = data["month"].replace("sep",9)
  39. #data["month"] = data["month"].replace("oct",10)
  40. #data["month"] = data["month"].replace("nov",11)
  41. #data["month"] = data["month"].replace("dec",12)
  42. #
  43. #
  44. #
  45. ##Defining X1, X2, and all the data X
  46. #X1 = data.FFMC
  47. #X2 = data.DMC
  48. #X3 = data.DC
  49. #X4 = data.ISI
  50. #X5 = data.temp
  51. #X6 = data.RH
  52. #X7 = data.wind
  53. #X8 = data.rain
  54. #X_raw = np.column_stack((X1, X2, X3, X4, X5, X6, X7, X8))
  55. #
  56. ##Normalizing or not the data
  57. #min_max_scaler = preprocessing.MinMaxScaler()
  58. #X = min_max_scaler.fit_transform(X_raw)
  59. ##X = X_raw
  60. ##print(X)
  61. #
  62. ##Defining Y variables depending on whether we have a regression or classification problem
  63. #if analysis_type == 'R':
  64. # Y = data.month
  65. #
  66. #
  67. ##Using Built in train test split function in sklearn
  68. #X_train, X_test, Y_train, Y_test = train_test_split(X, Y, train_size = 0.8)
  69. #
  70. #if analysis_type == 'R':
  71. # neuron_count = []
  72. # for i in range (1,201):
  73. #
  74. # #Fit the neural network for Regression purposes (i.e., you expect a continuous variable out)
  75. # #Note that 'sgd' and 'adam' require a batch_size and the function is not as clear
  76. # acti = ['logistic', 'tanh', 'relu', 'identity']
  77. # algo = ['lbfgs', 'sgd', 'adam']
  78. # learn = ['constant', 'invscaling', 'adaptive']
  79. # neural = MLPRegressor(activation=acti[0], solver=algo[0], batch_size = 1, learning_rate = learn[0], hidden_layer_sizes=(i,))
  80. #
  81. # #Cross validation
  82. # neural_scores = cross_val_score(neural, X_train, Y_train, cv=5)
  83. ## print("Cross Validation Accuracy: {0} (+/- {1})".format(neural_scores.mean().round(2), (neural_scores.std() * 2).round(2)))
  84. ## print("")
  85. # rounded = neural_scores.mean().round(2)
  86. # neuron_count.append(rounded)
  87. #
  88. # #Cross validation
  89. # neural_scores = cross_val_score(neural, X_train, Y_train, cv=15)
  90. # print("Cross Validation Accuracy: {0} (+/- {1})".format(neural_scores.mean().round(2), (neural_scores.std() * 2).round(2)))
  91. ## print("")
  92. #
  93. # #Fitting final neural network
  94. # neural.fit(X_train, Y_train)
  95. # neural_score = neural.score(X_test, Y_test)
  96. ## print("Shape of neural network: {0}".format([coef.shape for coef in neural.coefs_]))
  97. ## print("Coefs: ")
  98. ## print("")
  99. ## print(neural.coefs_[0].round(2))
  100. ## print("")
  101. ## print(neural.coefs_[1].round(2))
  102. ## print("")
  103. ## print("Intercepts: {0}".format(neural.intercepts_))
  104. ## print("")
  105. ## print("Loss: {0}".format(neural.loss_))
  106. ## print("")
  107. ## print("Iteration: {0}".format(neural.n_iter_))
  108. ## print("")
  109. ## print("Layers: {0}".format(neural.n_layers_))
  110. ## print("")
  111. ## print("Outputs: {0}".format(neural.n_outputs_))
  112. ## print("")
  113. ## print("Output Activation: {0}".format(neural.out_activation_)) #identity because we are looking for a value
  114. ## print("")
  115. #
  116. # #Assess the fitted Neural Network
  117. # print("Y test and predicted")
  118. # print(Y_test.values)
  119. # print(neural.predict(X_test).round(1))
  120. # print("")
  121. # print("Accuracy as Pearson's R2: {0}".format(neural_score.round(4)))
  122. # print("")
  123. #
  124. ##fix, ax = plt.subplots()
  125. ##ax.scatter(Y_test.values, neural.predict(X_test), edgecolors=(0, 0, 0))
  126. ##ax.plot([Y.min(0), Y.max(100)], [X.min(0), X.max(100)], 'k--', lw=3)
  127. ##ax.set_xlabel('Measured Grade')
  128. ##ax.set_xlabel('Predicted Grade')
  129. #
  130. #
  131. #fix, ax = plt.subplots()
  132. #ax.scatter(neural.predict(X_test), Y_test.values)
  133. #plt.axis([0, 100, 0, 100])
  134. #plt.grid(True)
  135. #ax.set_xlabel('Number of Neurons')
  136. #ax.set_ylabel('Accuracy')
  137. ##else:
  138. ## #Fit the neural network for Classification purposes (i.e., you don't expect a continuous variable out).
  139. ## #Note that 'sgd' and 'adam' require a batch_size and the function is not as clear
  140. ## acti = ['logistic', 'tanh', 'relu', 'identity']
  141. ## algo = ['lbfgs', 'sgd', 'adam']
  142. ## learn = ['constant', 'invscaling', 'adaptive']
  143. ## neural = MLPClassifier(activation=acti[2], solver=algo[2], batch_size = 1, learning_rate = learn[2], hidden_layer_sizes=(7,500))
  144. ##
  145. ## #Cross validation
  146. ## neural_scores = cross_val_score(neural, X_train, Y_train, cv=5)
  147. ## print("Cross Validation Accuracy: {0} (+/- {1})".format(neural_scores.mean().round(2), (neural_scores.std() * 2).round(2)))
  148. ## print("")
  149. ##
  150. ## #Fitting final neural network
  151. ## neural.fit(X_train, Y_train)
  152. ## neural_score = neural.score(X_test, Y_test)
  153. ## print("Classes: {0}".format(neural.classes_))
  154. ## print("")
  155. ## print("Shape of neural network: {0}".format([coef.shape for coef in neural.coefs_]))
  156. ## print("")
  157. ## print("Coefs: ")
  158. ## print(neural.coefs_[0].round(2))
  159. ## print("")
  160. ## print(neural.coefs_[1].round(2))
  161. ## print("")
  162. ## print("Intercepts: {0}".format(neural.intercepts_))
  163. ## print("")
  164. ## print("Loss: {0}".format(neural.loss_))
  165. ## print("")
  166. ## print("Iteration: {0}".format(neural.n_iter_))
  167. ## print("")
  168. ## print("Layers: {0}".format(neural.n_layers_))
  169. ## print("")
  170. ## print("Outputs: {0}".format(neural.n_outputs_))
  171. ## print("")
  172. ## print("Output Activation: {0}".format(neural.out_activation_)) #softmax to get a probability between 0 and 1
  173. ## print("")
  174. ##
  175. ## #Assess the fitted Neural Network
  176. ## print("Y test and predicted")
  177. ## print(Y_test.values)
  178. ## print(neural.predict(X_test))
  179. ## print("")
  180. ## print("Mean Accuracy: {0}".format(neural_score.round(4)))
  181. ## print("")
  182. ##
  183. ###fix, ax = plt.subplots()
  184. ###ax.scatter(Y_test, X_test, edgecolors=(0, 0, 0))
  185. ###ax.plot([y.min(0), y.max(100)], [x.min(0), x.max(100)], 'k--', lw=4)
  186. ###ax.set_xlabel('Measured')
  187. ###ax.set_xlabel('Predicted')
  188. ###
  189. ####fig, ax =plt.subplots
  190. ####ax.scatter(y_predicted, edgecolors=(0, 0, 0))
  191. ###ax.plot
  192. ###plt.xlabel('X values')
  193. ###plt.ylabel('Area')
  194. ###plt.savefig(file_name, dpi=300)
  195. ###plt.show()
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!