import numpy as np
import matplotlib.pyplot as plt
import pandas as pd


# 指数成分股，总市值
q_50 = query(
 valuation.code, valuation.market_cap
).filter(
    # valuation.code.in_(get_index_stocks('000852.XSHG'))  #中证1000
    # valuation.code.in_(get_index_stocks('000905.XSHG')) #中证500
    valuation.code.in_(get_index_stocks('000300.XSHG')) #沪深300
).order_by(
    valuation.market_cap.desc())
df_50 = get_fundamentals(q_50, '2019-01-01') 
df_50.head()

# 指数成分股，总市值绘图
import matplotlib.pyplot as plt

plt.figure(figsize=(10,5))
plt.bar(range(len(list(df_50.market_cap))), list(df_50.market_cap),edgecolor='blue')
plt.title('排序后市值分布')
plt.show()

# 绘制直方图 

plt.figure(figsize=(10,5))
plt.hist(df_50.market_cap, bins=50)
plt.xlabel('stocks')
plt.ylabel('Number of market_cap')
plt.title('直方图分布')
plt.show()

# 指数成分股，相关统计数据

# 均值
market_cap_mean_50 = mean(list(df_50.market_cap))

# 中位数
market_cap_median_50 = median(list(df_50.market_cap))

# 标准差
market_cap_std_50 = std(list(df_50.market_cap))

# 极差
market_cap_ptp_50 = ptp(list(df_50.market_cap))


print('均值:',market_cap_mean_50,'中位数:',market_cap_median_50,'标准差:',market_cap_std_50,'极差:',market_cap_ptp_50)

"""
绘制箱体图
Created on 2017.09.04 by ForestNeo
"""
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd


def list_generator(number, min, max):
    dataList = list()
    for i in range(1, number):
        dataList.append(np.random.randint(min, max))
    return dataList
 
list1 = list_generator(100, 20, 80)
 
data = pd.DataFrame({
    "dataSet1":list1,
})
 
df_50.boxplot(figsize = (4,5))
plt.title('成分股市值箱体图')
plt.show()

# 求出每一只个股的sw_l1申万一级行业名称
# 同时使用try和pass跳过错误

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

industry = get_industry(list(df_50.code))

stock_dict={}

for stock in industry.keys():
    try:
        stock_dict[stock] = industry[stock]['sw_l1']['industry_name'] # 取到每个股票的【申万一级】行业名称
    except:
        pass
    
# 转换为DF数据格式
DFindustry = pd.DataFrame(list(stock_dict.values()),index=stock_dict.keys(),columns=['industry_name'])
DFindustry.head()

# 绘制行业分布饼图

import numpy as np
import matplotlib.pyplot as plt

tt = DFindustry.groupby(['industry_name']).size()
tt1 = list(tt.index.values)
tt2 = list(tt.values)

labels = tt1
fracs = tt2

plt.figure(figsize=(10,8))

plt.axes(aspect=1)  # set this , Figure is round, otherwise it is an ellipse
plt.pie(x=fracs, labels=labels, explode=None,autopct='%3.1f %%',shadow=True)
plt.title('成分股行业分布饼图')
plt.show()

# 分布行业数量
industryN = len(tt1) 

# 每个行业平均个股数量
average_N = np.array(tt2).mean()

# 每个行业个股数量标准差
STD_N = np.array(tt2).std()

# 每个行业个股分布变异系数
CV_N = STD_N / average_N

print('行业数量:',industryN,'平均个股数量:',average_N,'个股数量标准差:',STD_N,'个股分布变异系数:',CV_N)

# 成交额分析
money_test = get_price( list(df_50.code), start_date='2018-01-01', end_date='2019-01-01', frequency='daily', fields=None, skip_paused=False, fq='none', count=None)
money_test

money_sort = money_test.money.mean().sort_values(ascending=False)

plt.figure(figsize=(10,5))
plt.bar(range(len(list(money_sort))), list(money_sort),edgecolor='blue')
plt.ylabel('成交额')
plt.title('成分股过去1年的日成交额均值')
plt.show()

print('成分股过去1年的日成交额均值的均值:',money_sort.mean())
print('成分股过去1年的日成交额均值的中位数:',money_sort.median())

close_sort = money_test.close.mean().sort_values(ascending=False)

plt.figure(figsize=(10,5))
plt.bar(range(len(list(close_sort))), list(close_sort),edgecolor='blue')
plt.ylabel('不复权收盘价')
plt.title('成分股过去1年的不复权收盘价均值')
plt.show()


print('成分股过去1年的不复权收盘价均值的均值:',close_sort.mean())
print('成分股过去1年的不复权收盘价均值的中位数:',close_sort.median())

# 通过rolling_mean函数求去价格量纲的波动率

mov_std = money_test.close.rolling(20).std() / money_test.close.rolling(20).mean()

std_sort = mov_std.mean().sort_values(ascending=False) 

std_sort

plt.figure(figsize=(10,5))
plt.bar(range(len(list(std_sort))), list(std_sort),edgecolor='blue')
plt.ylabel('去价格量纲的波动率均值')
plt.title('去价格量纲的波动率均值')
plt.show()


print('成分股过去1年的去价格量纲的波动率均值:',std_sort.mean())
print('成分股过去1年的去价格量纲的波动率的中位数:',std_sort.median())

# ER效率系数

date_list = []
df_ER = pd.DataFrame()
for i in range(20,len(money_test.close)):
    temp = money_test.close.iloc[i-20:i,:]
    Direction = abs(temp.iloc[0,:] - temp.iloc[-1,:])
    Volatility = ((temp - temp.shift(1)).abs()).sum()
    ER = Direction/Volatility
    date = money_test.close.index.tolist()[i].date()
    df_ER = df_ER.append(pd.DataFrame(ER).T)
    
    date_list.append(date)
    print(date)
df_ER.index = date_list
print(df_ER.head())

# 过去1年效率系数均值

ERmean = df_ER.mean().sort_values(ascending=False)

plt.figure(figsize=(10,5))
plt.bar(range(len(list(ERmean))), list(ERmean),edgecolor='blue')
plt.ylabel('效率系数')
plt.title('过去1年效率系数均值')
plt.show()


print('成分股过去1年效率系数均值的均值:',ERmean.mean())
print('成分股过去1年效率系数均值的中位数:',ERmean.median())

# 各项统计数据整理存储
storage_hs300 = [market_cap_mean_50,market_cap_std_50,CV_N,money_sort.mean(),close_sort.mean(),close_sort.median(),std_sort.mean(),std_sort.median(),ERmean.mean()]
storage_hs300

# 各项统计数据整理存储
storage_zz500 = [market_cap_mean_50,market_cap_std_50,CV_N,money_sort.mean(),close_sort.mean(),close_sort.median(),std_sort.mean(),std_sort.median(),ERmean.mean()]
storage_zz500

# 各项统计数据整理存储
storage_zz1000 = [market_cap_mean_50,market_cap_std_50,CV_N,money_sort.mean(),close_sort.mean(),close_sort.median(),std_sort.mean(),std_sort.median(),ERmean.mean()]
storage_zz1000

# 指数成分股相关性

from jqdata import *

import warnings  
warnings.filterwarnings('ignore')

# 交易日时间段
start_day = '2015-01-01'
end_day = '2019-04-01'
trade_day_list = get_trade_days(start_date=start_day, end_date=end_day)

corr_list = []
index_list = []

# 每5天计算一次截面相关性
for i in range(0,len(list(trade_day_list)),5):
    
    each_day = list(trade_day_list)[i]
    index_list.append(each_day)
    
    q_50 = query(
     valuation.code).filter(
        valuation.code.in_(get_index_stocks('000852.XSHG'))  #中证1000
        # valuation.code.in_(get_index_stocks('000905.XSHG')) #中证500
        # valuation.code.in_(get_index_stocks('000300.XSHG')) #沪深300
    ).order_by(
        valuation.market_cap.desc())
    df_50 = get_fundamentals(q_50, each_day) 

    Each_Price_df=get_price(list(df_50.code),end_date=each_day,count=20,fields=['close'],frequency='daily')['close']
    Each_corr = Each_Price_df.corr().mean().mean()
    corr_list.append(Each_corr)

DF_corr_list = pd.DataFrame(index = index_list)
DF_corr_list['corr_list'] = corr_list

DF_corr_list['price_list'] = get_price('000852.XSHG', start_date=start_day, end_date=end_day, \
                                       frequency='1d', fields=['close'])['close']

mpl.rcParams['font.family']='serif'
mpl.rcParams['axes.unicode_minus']=False # 处理负号
(DF_corr_list/DF_corr_list.iloc[0]).plot(figsize=(12,7),grid='on')

print('指数成分股两两间相关系数 = ',DF_corr_list['corr_list'].mean())

# 指数成分股相关性

from jqdata import *

import warnings  
warnings.filterwarnings('ignore')

# 交易日时间段
start_day = '2015-01-01'
end_day = '2019-04-01'
trade_day_list = get_trade_days(start_date=start_day, end_date=end_day)

corr_list = []
index_list = []

# 每5天计算一次截面相关性
for i in range(0,len(list(trade_day_list)),5):
    
    each_day = list(trade_day_list)[i]
    index_list.append(each_day)
    
    q_50 = query(
     valuation.code).filter(
        # valuation.code.in_(get_index_stocks('000852.XSHG'))  #中证1000
        valuation.code.in_(get_index_stocks('000905.XSHG')) #中证500
        # valuation.code.in_(get_index_stocks('000300.XSHG')) #沪深300
    ).order_by(
        valuation.market_cap.desc())
    df_50 = get_fundamentals(q_50, each_day) 

    Each_Price_df=get_price(list(df_50.code),end_date=each_day,count=20,fields=['close'],frequency='daily')['close']
    Each_corr = Each_Price_df.corr().mean().mean()
    corr_list.append(Each_corr)

DF_corr_list = pd.DataFrame(index = index_list)
DF_corr_list['corr_list'] = corr_list

DF_corr_list['price_list'] = get_price('000905.XSHG', start_date=start_day, end_date=end_day, \
                                       frequency='1d', fields=['close'])['close']

mpl.rcParams['font.family']='serif'
mpl.rcParams['axes.unicode_minus']=False # 处理负号
(DF_corr_list/DF_corr_list.iloc[0]).plot(figsize=(12,7),grid='on')

print('指数成分股两两间相关系数 = ',DF_corr_list['corr_list'].mean())

# 指数成分股相关性

from jqdata import *

import warnings  
warnings.filterwarnings('ignore')

# 交易日时间段
start_day = '2015-01-01'
end_day = '2019-04-01'
trade_day_list = get_trade_days(start_date=start_day, end_date=end_day)

corr_list = []
index_list = []

# 每5天计算一次截面相关性
for i in range(0,len(list(trade_day_list)),5):
    
    each_day = list(trade_day_list)[i]
    index_list.append(each_day)
    
    q_50 = query(
     valuation.code).filter(
        # valuation.code.in_(get_index_stocks('000852.XSHG'))  #中证1000
        # valuation.code.in_(get_index_stocks('000905.XSHG')) #中证500
        valuation.code.in_(get_index_stocks('000300.XSHG')) #沪深300
    ).order_by(
        valuation.market_cap.desc())
    df_50 = get_fundamentals(q_50, each_day) 

    Each_Price_df=get_price(list(df_50.code),end_date=each_day,count=20,fields=['close'],frequency='daily')['close']
    Each_corr = Each_Price_df.corr().mean().mean()
    corr_list.append(Each_corr)

DF_corr_list = pd.DataFrame(index = index_list)
DF_corr_list['corr_list'] = corr_list

DF_corr_list['price_list'] = get_price('000300.XSHG', start_date=start_day, end_date=end_day, \
                                       frequency='1d', fields=['close'])['close']

mpl.rcParams['font.family']='serif'
mpl.rcParams['axes.unicode_minus']=False # 处理负号
(DF_corr_list/DF_corr_list.iloc[0]).plot(figsize=(12,7),grid='on')

print('指数成分股两两间相关系数 = ',DF_corr_list['corr_list'].mean())

# 获得三大指数价格，并净值绘图

index_list = ['000852.XSHG','000905.XSHG','000300.XSHG']
Price_df=get_price(index_list,end_date='2019-04-18',count=1000,fields=['close'],frequency='daily')['close']
mpl.rcParams['font.family']='serif'
mpl.rcParams['axes.unicode_minus']=False # 处理负号
(Price_df/Price_df.iloc[0]).plot(figsize=(12,7),grid='on')

# 获得价格的滑动窗标准差，并绘图

import pandas as pd

Price_df_std = Price_df.rolling(window=20).std()

# 同样方法，获得价格均值
mov_close = Price_df.rolling(window=20).mean()

# 将价格归一化后（通过去价格量纲）的动量，装入DataFrame，并赋予时间轴
df_STD = pd.DataFrame(Price_df_std/mov_close)

df_STD = df_STD.dropna()

mpl.rcParams['font.family']='serif'
mpl.rcParams['axes.unicode_minus']=False # 处理负号
df_STD.plot(figsize=(12,7),grid='on')
df_STD.mean()

# 相关性矩阵绘图

import seaborn as sns
import matplotlib as mpl

fig = plt.figure(figsize= (6,6))
ax = fig.add_subplot(111)
ax = sns.heatmap(Price_df.corr(),\
                 annot=True,annot_kws={'size':12,'weight':'bold'})

# 相关性矩阵绘图

import seaborn as sns
import matplotlib as mpl

fig = plt.figure(figsize= (6,6))
ax = fig.add_subplot(111)
ax = sns.heatmap(df_STD.corr(),\
                 annot=True,annot_kws={'size':12,'weight':'bold'})