print?

%matplotlib inline

import tensorflow as tf
import pandas            as pd
import tensorflow        as tf  
import numpy             as np
import matplotlib.pyplot as plt
# 支持中文
plt.rcParams['font.sans-serif'] = ['SimHei']  # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False  # 用来正常显示负号

from numpy                 import array
from sklearn               import metrics
from sklearn.preprocessing import MinMaxScaler
from tensorflow.keras.models          import Sequential
from tensorflow.keras.layers          import Dense,LSTM,Bidirectional

from numpy.random   import seed

np.__path__

# 确保结果尽可能重现

seed(1)
tf.random.set_seed(1)

# 设置相关参数
n_timestamp  = 5    # 时间戳
n_epochs     = 30    # 训练轮数
# ====================================
#      选择模型：
#            1: 单层 LSTM
#            2: 多层 LSTM
#            3: 双向 LSTM
# ====================================
model_type = 1

# 导入tushare
import tushare as ts

#pro=ts.get_hist_data('603722')
help(ts.pro_api)

# 初始化pro接口
pro = ts.pro_api('66f88ebe33ca20ebed66ba99509f7607b6a05f1e64f51454e3d4021c')

# 拉取数据
data = pro.daily(**{
    "ts_code": '603722.SH',
    "trade_date": "",
    "start_date": "2017-10-1",
    "end_date": "",
    "offset": "",
    "limit": ""
}, fields=[
    "trade_date",
    "open",
    "close",
    "high",
    "low",
    "vol",
    "ts_code"
])

print(data)
data.to_csv("stock603722.csv")

data=pd.read_csv("./stock603722.csv",parse_dates=["trade_date"],index_col="trade_date")[["open","high","low","close"]]
# data=pd.read_csv("./stock688333.csv",parse_dates=["trade_date"])[["trade_date","open","high","low","close","vol"]]
data

mydata=pd.read_csv("./stock603722.csv")[["open","high","low","close"]]
plt.plot(mydata)

myindex=range(1268)
mydata.loc[:]=mydata.iloc[::-1].to_numpy()
mydata[:2]

# 获取股价数据
import pandas as pd
import numpy as np
# import akshare as ak



df3 = data.reset_index().iloc[-30:, :6]  # 取过去30天数据
df3 = df3.dropna(how='any').reset_index(drop=True) #去除空值且从零开始编号索引
df3 = df3.sort_values(by='trade_date', ascending=True)
print(df3.info())

# 均线数据
df3['5'] = df3.close.rolling(5).mean()
df3['10'] = df3.close.rolling(10).mean()

df3.tail()

import matplotlib
matplotlib.style.use('ggplot') #用于调整图标样式，可选
import matplotlib.pyplot as plt
from mplfinance.original_flavor import candlestick2_ohlc
from matplotlib.ticker import FormatStrFormatter

fig, ax = plt.subplots(1, 1, figsize=(8,3), dpi=200)

candlestick2_ohlc(ax,
                  opens = df3[ 'open'].values,
                highs = df3['high'].values,
                lows = df3[ 'low'].values,
                closes = df3['close'].values,
                  width=0.5, colorup="r",colordown="g")

# 显示最高点和最低点
ax.text( df3.high.idxmax(), df3.high.max(),   s =df3.high.max(), fontsize=8)
ax.text( df3.high.idxmin(), df3.high.min()-2, s = df3.high.min(), fontsize=8)

ax.set_facecolor("white")
ax.set_title("阿科力")

# 画均线
plt.plot(df3['5'].values, alpha = 0.5, label='MA5')
plt.plot(df3['10'].values, alpha = 0.5, label='MA10')

ax.legend(facecolor='white', edgecolor='white', fontsize=6)

# 修改x轴坐标
plt.xticks(ticks =  np.arange(0,len(df3)), labels = df3.trade_date.dt.strftime('%Y-%m-%d').to_numpy() )
plt.xticks(rotation=90, size=8)

# 修改y轴坐标
ax.yaxis.set_major_formatter(FormatStrFormatter('%.2f'))

plt.show()

#from matplotlib.dates  import date2num
#data['date']=date2num(data.index.to_pydatetime())
#data=data.sort_values(by="date",ascending=True)
data

# 获取股价数据
import pandas as pd
import numpy as np

df3 = data.reset_index().iloc[:,:]  #取过去30天数据
df3 = df3.dropna(how='any').reset_index(drop=True) #去除空值且从零开始编号索引
print(df3)
df3 = df3.sort_values(by='trade_date', ascending=True)
print(df3.info())

# 均线数据
df3['5'] = df3.close.rolling(5).mean()
df3['10'] = df3.close.rolling(10).mean()
df3['30']=df3.close.rolling(30).mean()

df3

import matplotlib
matplotlib.style.use('ggplot') #用于调整图标样式，可选
import matplotlib.pyplot as plt
from mplfinance.original_flavor import candlestick2_ohlc
from matplotlib.ticker import FormatStrFormatter

fig, ax = plt.subplots(1, 1, figsize=(8,3), dpi=200)

candlestick2_ohlc(ax,
                  opens = df3[ 'open'].values,
                highs = df3['high'].values,
                lows = df3[ 'low'].values,
                closes = df3['close'].values,
                  width=0.5, colorup="r",colordown="g")

# 显示最高点和最低点
ax.text( df3.high.idxmax(), df3.high.max(),   s =df3.high.max(), fontsize=8)
ax.text( df3.high.idxmin(), df3.high.min()-2, s = df3.high.min(), fontsize=8)

ax.set_facecolor("white")
ax.set_title("阿科力")

# 画均线
plt.plot(df3['5'].values, alpha = 0.5, label='MA5')
plt.plot(df3['10'].values, alpha = 0.5, label='MA10')
plt.plot(df3['30'].values,alpha=0.5, label='MA30')

ax.legend(facecolor='white', edgecolor='white', fontsize=6)

# 修改x轴坐标
tempXticks=np.arange(0,len(df3))
#XticksData=data.asfreq("2M").dropna()
nameXticks =  df3.trade_date.dt.strftime('%Y-%m-%d').to_numpy()

plt.xticks(ticks =tempXticks , labels =nameXticks )
plt.xticks(rotation=45, size=8)
#修改X轴间隔
x_major_locator=plt.MultipleLocator(100)
ax.xaxis.set_major_locator(x_major_locator)
ax.spines['bottom'].set_color('red')
ax.spines['left'].set_color('red')
plt.xlim(0,1100)
# 修改y轴坐标
ax.yaxis.set_major_formatter(FormatStrFormatter('%.2f'))

plt.show()

a=np.linspace(0,round(len(df3),0),10,dtype=int).tolist()
b=df3.trade_date.dt.strftime('%Y-%m-%d').to_numpy()
c=data.asfreq("2M").dropna()
#c.dropna().index
cc=c.index.strftime('%Y-%m-%d').to_numpy()
e=df3.trade_date.tolist()
#f=[e[i] for i in cc]
cc

# data1=pd.DataFrame(data,columns=["trade_date","close"])
#data1=pd.DataFrame(data,columns=["close"])
data1=data.iloc[::-1]
data1["open"].plot(label="open")
data1["close"].plot(label="close")
plt.legend()
plt.show()

ddd=[]
for item in data1.items():
    ddd.append(item)
ind=ddd[0][1].values
da1=ddd[1][1].values
index1=[]
data2=[]
for item in ind:
    index1.append(item)
index1.reverse()
for item in da1:
    data2.append(item)
data2.reverse()
data1={'trade_date':index1,'close':data2}
stockdata=pd.DataFrame(data1)

stockdata.items    

df3.iloc[0:3000,1:3].values[:,1]

from matplotlib.dates  import date2num
date2num(df3.trade_date.values)
df3['date']=date2num(df3.trade_date.values)
#df3=df3.sort_values(by="date",ascending=True)

#df3=df3[['date','open', 'high', 'low', 'close']]
ind=np.arange(0,3000)
dataf1=df3[['date','open', 'high', 'low', 'close']]

dataf1.set_index("date")

#
#拟合开始
#
training_set = dataf1.iloc[0:512, 4:5].to_numpy()#要提取一列数据，否则会出错
test_set     = dataf1.iloc[973 - 300:, 4:5].to_numpy()
print(training_set)

#将数据归一化，范围是0到1
sc  = MinMaxScaler(feature_range=(0, 1))
training_set_scaled = sc.fit_transform(training_set)
testing_set_scaled  = sc.transform(test_set) 

# 取前 n_timestamp 天的数据为 X；n_timestamp+1天数据为 Y。
def data_split(sequence, n_timestamp):
    X = []
    y = []
    for i in range(len(sequence)):
        end_ix = i + n_timestamp
        
        if end_ix > len(sequence)-1:
            break
            
        seq_x, seq_y = sequence[i:end_ix], sequence[end_ix]
        X.append(seq_x)
        y.append(seq_y)
    return array(X), array(y)

X_train, y_train = data_split(training_set_scaled, n_timestamp)

X_train          = X_train.reshape(X_train.shape[0], X_train.shape[1], 1)

X_test, y_test   = data_split(testing_set_scaled, n_timestamp)
X_test           = X_test.reshape(X_test.shape[0], X_test.shape[1], 1)

#五、构建模型
# 建构 LSTM模型
model_type=3
if model_type == 1:
    # 单层 LSTM
    model = Sequential()
    model.add(LSTM(units=50, activation='relu',
                   input_shape=(X_train.shape[1], 1)))
    model.add(Dense(units=1))
if model_type == 2:
    # 多层 LSTM
    model = Sequential()
    model.add(LSTM(units=50, activation='relu', return_sequences=True,
                   input_shape=(X_train.shape[1], 1)))
    model.add(LSTM(units=50, activation='relu'))
    model.add(Dense(1))
if model_type == 3:
    # 双向 LSTM
    model = Sequential()
    model.add(Bidirectional(LSTM(50, activation='relu'),
                            input_shape=(X_train.shape[1], 1)))
    model.add(Dense(1))

model.summary()  # 输出模型结构

model.compile(optimizer=tf.keras.optimizers.Adam(0.001),
              loss='mean_squared_error')  # 损失函数用均方误差
#七、训练模型
history = model.fit(X_train, y_train,
                    batch_size=64,
                    epochs=n_epochs,
                    validation_data=(X_test, y_test),
                    validation_freq=1)  # 测试的epoch间隔数

model.summary()

plt.plot(history.history['loss'], label='Training Loss')
plt.plot(history.history['val_loss'], label='Validation Loss')
#plt.title('Training and Validation Loss')
plt.legend()
plt.show()

predicted_stock_price = model.predict(
    X_test)                        # 测试集输入模型进行预测
predicted_stock_price = sc.inverse_transform(
    predicted_stock_price)  # 对预测数据还原---从（0，1）反归一化到原始范围
real_stock_price = sc.inverse_transform(y_test)  # 对真实数据还原---从（0，1）反归一化到原始范围

# 画出真实数据和预测数据的对比曲线
plt.plot(real_stock_price, color='red', label='Stock Price')
plt.plot(predicted_stock_price, color='blue', label='Predicted Stock Price')
plt.title('阿科力')
plt.xlabel('Time')
plt.ylabel('Stock Price')
plt.legend()
plt.show()

MSE = metrics.mean_squared_error(predicted_stock_price, real_stock_price)
RMSE = metrics.mean_squared_error(predicted_stock_price, real_stock_price)**0.5
MAE = metrics.mean_absolute_error(predicted_stock_price, real_stock_price)
R2 = metrics.r2_score(predicted_stock_price, real_stock_price)

print('均方误差: %.5f' % MSE)
print('均方根误差: %.5f' % RMSE)
print('平均绝对误差: %.5f' % MAE)
print('R2: %.5f' % R2)