import tensorflow as tf
import numpy as np
import pandas as pd
import matplotlib.pylab as plt
import matplotlib as mpl

#   讀取資料集
TRIN_URL = 'http://download.tensorflow.org/data/iris_training.csv'            # 資料集下載網址
df_iris = pd.read_csv('./鳶尾花資料集/iris.csv', header=0)                   # 讀取本地csv檔案資料
# print(df_iris)                                                              #   最後一列是分類編號
 

#   處理資料集
iris = np.array(df_iris)                                                        # 將資料轉換成numpy陣列
iris2 = iris[iris[:, -1] < 2]                                                   # 只取分類編號小於2的兩類資料
train_x = iris2[:, 0:2]                                                         # 只取特徵的前兩列
train_x = train_x - np.mean(train_x, axis=0)                                     #
 
 需要將樣本的均值變為0
train_1 = np.ones(train_x.shape[0]).reshape(-1, 1)                               # 生成一個與train_x一樣行數的全1矩陣
# print('train_1:',train_1)
train_x = tf.concat((train_x, train_1), axis=1)                                  # 將train_x擴充一列全1
train_x = tf.cast(train_x,tf.float32)
train_y = iris2[:, -1]  #
 
 標籤
train_y = train_y.reshape(-1,1)
# print(train_x, train_x.shape)
# print('axis = 0:',np.mean(train_x,axis=0))                                    #   axis = 0 求一列的平均值
print(train_y, train_y.shape)

#   設定超參
iter = 2000
learn_rate = 0.1
loss_list = []
acc_list = []
#   初始化訓練引數

w = tf.Variable(np.random.randn(3,1),dtype=tf.float32)
# w = tf.Variable(np.array([1.,1.,1.]).reshape(-1,1),dtype=tf.float32)

for i in range(iter):
    with tf.GradientTape() as tape:
        y_p = 1 / (1 + tf.exp(-(tf.matmul(train_x, w))))
        loss = tf.reduce_mean(-(train_y * tf.math.log(y_p) + (1 - train_y) * tf.math.log(1 - y_p)))
    dloss_dw = tape.gradient(loss, w)
    w.assign_sub(learn_rate * dloss_dw)
    loss_list.append(loss)
    acc = tf.reduce_mean(tf.cast(tf.equal(tf.round(y_p),train_y),dtype = tf.float32))
    acc_list.append(acc)
    if i % 100 == 0:
        print('第{}次, loss:{},acc:{}'.format(i,loss,acc))
        # print('y_p:{}\ntrain_y:{}'.format(y_p, train_y))
        print()


#   預測直線的橫縱座標處理
x1 = train_x[:,0]
print(w[0] * 55)
x2 = -(w[0] * x1 + w[2])/ w[1]
# print('x2:',x2)

#   畫圖
plt.rcParams["font.family"] = 'SimHei'                          # 將字型改為中文
plt.rcParams['axes.unicode_minus'] = False                      # 設定了中文字型預設後，座標的"-"號無法顯示，設定這個引數就可以避免
plt.subplot(221)
cm_pt = mpl.colors.ListedColormap(['red', 'green'])
plt.scatter(x=train_x[:, 0], y=train_x[:, 1], c=train_y, cmap=cm_pt)
plt.plot(x1,x2,label = '預測線')
plt.legend()
plt.subplot(223)
plt.title('損失值')
plt.plot(loss_list)
plt.subplot(224)
plt.title('準確率')
plt.plot(acc_list)
plt.show()