# Multi-class (Nonlinear) SVM Example
#----------------------------------
#
# This function wll illustrate how to
# implement the gaussian kernel with
# multiple classes on the iris dataset.
#
# Gaussian Kernel:
# K(x1, x2) = exp(-gamma * abs(x1 - x2)^2)
#
# X : (Sepal Length, Petal Width)
# Y: (I. setosa, I. virginica, I. versicolor) (3 classes)
#
# Basic idea: introduce an extra dimension to do
# one vs all classification.
#
# The prediction of a point will be the category with
# the largest margin or distance to boundary.

import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf
from sklearn import datasets
from tensorflow.python.framework import ops
ops.reset_default_graph()

# Create graph
sess = tf.Session()

# Load the data
# 加載iris數(shù)據(jù)集并為每類分離目標值。
# 因為我們想繪制結(jié)果圖，所以只使用花萼長度和花瓣寬度兩個特征。
# 為了便于繪圖，也會分離x值和y值
# iris.data = [(Sepal Length, Sepal Width, Petal Length, Petal Width)]
iris = datasets.load_iris()
x_vals = np.array([[x[0], x[3]] for x in iris.data])
y_vals1 = np.array([1 if y==0 else -1 for y in iris.target])
y_vals2 = np.array([1 if y==1 else -1 for y in iris.target])
y_vals3 = np.array([1 if y==2 else -1 for y in iris.target])
y_vals = np.array([y_vals1, y_vals2, y_vals3])
class1_x = [x[0] for i,x in enumerate(x_vals) if iris.target[i]==0]
class1_y = [x[1] for i,x in enumerate(x_vals) if iris.target[i]==0]
class2_x = [x[0] for i,x in enumerate(x_vals) if iris.target[i]==1]
class2_y = [x[1] for i,x in enumerate(x_vals) if iris.target[i]==1]
class3_x = [x[0] for i,x in enumerate(x_vals) if iris.target[i]==2]
class3_y = [x[1] for i,x in enumerate(x_vals) if iris.target[i]==2]

# Declare batch size
batch_size = 50

# Initialize placeholders
# 數(shù)據(jù)集的維度在變化，從單類目標分類到三類目標分類。
# 我們將利用矩陣傳播和reshape技術一次性計算所有的三類SVM。
# 注意，由于一次性計算所有分類，
# y_target占位符的維度是[3，None]，模型變量b初始化大小為[3，batch_size]
x_data = tf.placeholder(shape=[None, 2], dtype=tf.float32)
y_target = tf.placeholder(shape=[3, None], dtype=tf.float32)
prediction_grid = tf.placeholder(shape=[None, 2], dtype=tf.float32)

# Create variables for svm
b = tf.Variable(tf.random_normal(shape=[3,batch_size]))

# Gaussian (RBF) kernel 核函數(shù)只依賴x_data
gamma = tf.constant(-10.0)
dist = tf.reduce_sum(tf.square(x_data), 1)
dist = tf.reshape(dist, [-1,1])
sq_dists = tf.multiply(2., tf.matmul(x_data, tf.transpose(x_data)))
my_kernel = tf.exp(tf.multiply(gamma, tf.abs(sq_dists)))

# Declare function to do reshape/batch multiplication
# 最大的變化是批量矩陣乘法。
# 最終的結(jié)果是三維矩陣，并且需要傳播矩陣乘法。
# 所以數(shù)據(jù)矩陣和目標矩陣需要預處理，比如xT·x操作需額外增加一個維度。
# 這里創(chuàng)建一個函數(shù)來擴展矩陣維度，然后進行矩陣轉(zhuǎn)置，
# 接著調(diào)用TensorFlow的tf.batch_matmul（）函數(shù)
def reshape_matmul(mat):
  v1 = tf.expand_dims(mat, 1)
  v2 = tf.reshape(v1, [3, batch_size, 1])
  return(tf.matmul(v2, v1))

# Compute SVM Model 計算對偶損失函數(shù)
first_term = tf.reduce_sum(b)
b_vec_cross = tf.matmul(tf.transpose(b), b)
y_target_cross = reshape_matmul(y_target)

second_term = tf.reduce_sum(tf.multiply(my_kernel, tf.multiply(b_vec_cross, y_target_cross)),[1,2])
loss = tf.reduce_sum(tf.negative(tf.subtract(first_term, second_term)))

# Gaussian (RBF) prediction kernel
# 現(xiàn)在創(chuàng)建預測核函數(shù)。
# 要當心reduce_sum（）函數(shù)，這里我們并不想聚合三個SVM預測，
# 所以需要通過第二個參數(shù)告訴TensorFlow求和哪幾個
rA = tf.reshape(tf.reduce_sum(tf.square(x_data), 1),[-1,1])
rB = tf.reshape(tf.reduce_sum(tf.square(prediction_grid), 1),[-1,1])
pred_sq_dist = tf.add(tf.subtract(rA, tf.multiply(2., tf.matmul(x_data, tf.transpose(prediction_grid)))), tf.transpose(rB))
pred_kernel = tf.exp(tf.multiply(gamma, tf.abs(pred_sq_dist)))

# 實現(xiàn)預測核函數(shù)后，我們創(chuàng)建預測函數(shù)。
# 與二類不同的是，不再對模型輸出進行sign（）運算。
# 因為這里實現(xiàn)的是一對多方法，所以預測值是分類器有最大返回值的類別。
# 使用TensorFlow的內(nèi)建函數(shù)argmax（）來實現(xiàn)該功能
prediction_output = tf.matmul(tf.multiply(y_target,b), pred_kernel)
prediction = tf.arg_max(prediction_output-tf.expand_dims(tf.reduce_mean(prediction_output,1), 1), 0)
accuracy = tf.reduce_mean(tf.cast(tf.equal(prediction, tf.argmax(y_target,0)), tf.float32))

# Declare optimizer
my_opt = tf.train.GradientDescentOptimizer(0.01)
train_step = my_opt.minimize(loss)

# Initialize variables
init = tf.global_variables_initializer()
sess.run(init)

# Training loop
loss_vec = []
batch_accuracy = []
for i in range(100):
  rand_index = np.random.choice(len(x_vals), size=batch_size)
  rand_x = x_vals[rand_index]
  rand_y = y_vals[:,rand_index]
  sess.run(train_step, feed_dict={x_data: rand_x, y_target: rand_y})

  temp_loss = sess.run(loss, feed_dict={x_data: rand_x, y_target: rand_y})
  loss_vec.append(temp_loss)

  acc_temp = sess.run(accuracy, feed_dict={x_data: rand_x,
                       y_target: rand_y,
                       prediction_grid:rand_x})
  batch_accuracy.append(acc_temp)

  if (i+1)%25==0:
    print('Step #' + str(i+1))
    print('Loss = ' + str(temp_loss))

# 創(chuàng)建數(shù)據(jù)點的預測網(wǎng)格，運行預測函數(shù)
x_min, x_max = x_vals[:, 0].min() - 1, x_vals[:, 0].max() + 1
y_min, y_max = x_vals[:, 1].min() - 1, x_vals[:, 1].max() + 1
xx, yy = np.meshgrid(np.arange(x_min, x_max, 0.02),
           np.arange(y_min, y_max, 0.02))
grid_points = np.c_[xx.ravel(), yy.ravel()]
grid_predictions = sess.run(prediction, feed_dict={x_data: rand_x,
                          y_target: rand_y,
                          prediction_grid: grid_points})
grid_predictions = grid_predictions.reshape(xx.shape)

# Plot points and grid
plt.contourf(xx, yy, grid_predictions, cmap=plt.cm.Paired, alpha=0.8)
plt.plot(class1_x, class1_y, 'ro', label='I. setosa')
plt.plot(class2_x, class2_y, 'kx', label='I. versicolor')
plt.plot(class3_x, class3_y, 'gv', label='I. virginica')
plt.title('Gaussian SVM Results on Iris Data')
plt.xlabel('Pedal Length')
plt.ylabel('Sepal Width')
plt.legend(loc='lower right')
plt.ylim([-0.5, 3.0])
plt.xlim([3.5, 8.5])
plt.show()

# Plot batch accuracy
plt.plot(batch_accuracy, 'k-', label='Accuracy')
plt.title('Batch Accuracy')
plt.xlabel('Generation')
plt.ylabel('Accuracy')
plt.legend(loc='lower right')
plt.show()

# Plot loss over time
plt.plot(loss_vec, 'k-')
plt.title('Loss per Generation')
plt.xlabel('Generation')
plt.ylabel('Loss')
plt.show()