ipospeed.py

# -*-coding=utf-8-*-
__author__ = 'Rocky'
'''
http://30daydo.com
Contact: weigesysu@qq.com
'''
# 查看ipo速度 和指数的关系
import tushare as ts
from datetime import datetime
import pandas as pd
import numpy as np
from pandas import Series

import matplotlib.pyplot as plt

pd.set_option('display.max_rows', None)


class IPOSpeed():

    def __init__(self):

        self.ipo = ts.new_stocks()

        # 日期转化
        self.ipo['ipo_date'] = self.ipo['ipo_date'].astype('datetime64')
        self.start = self.ipo['ipo_date'].iloc[-1]
        self.end = self.ipo['ipo_date'].values[0]
        # 转化类型
        # ipo['ipo_date']=ipo['ipo_date'].astype('datetime64')
        # self.start_d=datetime.datetime.strptime(self.start,'%Y-%m-%d')
        # self.end_d=datetime.datetime.strptime(self.end,'%Y-%m-%d')
        # print(type(self.start_d))
        # period=self.start_d+datetime.timedelta(days=30)
        # print(period.strftime('%Y-%m-%d'))
        # print(ipo[ipo['ipo_date']<np.datetime64(period)])

    def comparation(self):
        # print(self.start)
        # print(self.end)
        delta = 30
        count_list = []
        profit_list = []
        self.period = self.end + np.timedelta64(delta, 'D')
        # print(self.period)
        # print(self.ipo[self.ipo['ipo_date']<self.period])
        # print(type(self.end.tolist()))
        l = self.end.tolist()
        ns = 1e-9
        b = datetime.utcfromtimestamp(l * ns)
        # print(datetime.fromtimestamp(self.end.tolist()))
        c = b.strftime('%Y-%m-%d')
        # print(type(c))
        start_data = self.start
        while start_data < self.end:
            # print(start_data)
            first_date = start_data
            start_data = start_data + np.timedelta64(delta, 'D')
            result = self.ipo[(self.ipo['ipo_date'] >= first_date) & (self.ipo['ipo_date'] < start_data)]
            # print(result)
            count = len(result)

            temp_end_data = start_data + np.timedelta64(-1, 'D')
            t = pd.to_datetime(str(temp_end_data))
            d = t.strftime('%Y-%m-%d')
            t1 = pd.to_datetime(str(first_date))
            d1 = t1.strftime('%Y-%m-%d')
            sz_index_data = ts.get_k_data('399001', index=True, start=d1, end=d)
            # print(index_data)
            # 大盘（深圳，考虑到国家队在上证的操作） 在30天内的收益
            before = sz_index_data['close'].values[0]
            after = sz_index_data['close'].values[-1]
            # profit_index=(index_data['close'][-1]-index_data['close'][0])/index_data['close'][0]*100
            p = round((after - before) / before * 100, 2)
            count_list.append(count)
            profit_list.append(p)

        return count_list, profit_list

    def draw(self):
        count, profit = self.comparation()
        s1 = Series(count, index=range(len(count)))
        s2 = Series(profit, index=range(len(profit)))
        '''
        ax=s1.plot()
        ax2=s2.plot()
        fig=ax.get_figure()
        fig.savefig('count.png')
        fig2=ax2.get_figure()
        fig2.savefig('profit.png')
        '''
        r = s1.corr(s2)


def main():
    obj = IPOSpeed()
    obj.draw()


main()