大飞

大飞 关注TA

挑战一切!

大飞

大飞

关注TA

挑战一切!

  •  世界的顶端
  • 混口饭吃
  • 写了333,609字

该文章投稿至Nemo社区   Python  板块 复制链接


使用Python挑选基金

发布于 2018/02/28 17:11 2,025浏览 0回复 5,965

参考文章:http://www.sohu.com/a/149042886_572440

如下筛选出最优基金“东吴阿尔法灵活配置混合”

[127 rows x 15 columns]

3年,2年,1年选取:
['华泰柏瑞创新动力混合', '申万菱信沪深300价值指数', '博时产业新动力混合', '农银策略价值混合', '交银消费新驱动股票', '海富通中证100', '富国天惠成长混合A', '银河蓝筹混合', '兴全沪深300指数(LOF)', '融通转型三动力灵活配置混合']

1年,半年,3月,1月选取:
{'诺德成长优势混合', '东吴阿尔法灵活配置混合', '前海开源工业革命4.0混合', '国联安鑫富混合C'}

3年,2年,1年,半年,3个月,1个月选取:
{'诺德成长优势混合', '东吴阿尔法灵活配置混合'}

3年,2年,1年,半年,3个月,1个月,一周选取:
{'东吴阿尔法灵活配置混合'}

Process finished with exit code 0


一.抓取天天基金的首页的基金信息并存储带scv

    

def get_fund():  # 取得基金列表
    # 先凑一个我们需要的URL出来
    max_jj = '5000'  # 调试5 工作5000
    fromstr = datetime.datetime.now().strftime('%Y-%m-%d')
    url = "http://fund.eastmoney.com/data/rankhandler.aspx?op=ph&dt=kf&ft=all&rs=&gs=0&sc=zzf&st=desc&sd=#custday&ed=#nowdate&qdii=&tabSubtype=,,,,,&pi=1&pn=#count&dx=1"
    url = url.replace('#count', max_jj)
    url = url.replace('#nowdate', fromstr)
    tostr = (datetime.datetime.now() - datetime.timedelta(days=5 * 365 + 1)).strftime('%Y-%m-%d')
    url = url.replace('#custday', tostr)
    # 取得文本
    s = get_html(url)
    # 去掉冗余信息,这里应该有更好的方法,但懒得折腾了,这样简单,十分钟搞定
    s = s[22:-159]
    for x in ['"', "'", ']', '[']:
        s = s.replace(x, '')
    lst = s.split(',')
    lst = split_list(lst, 25)
    frame = pd.DataFrame(lst,
                         columns=['code', 'name', 'py', '3', '4', 'jz', 'day1', 'week1', 'month1', 'month3', 'month6',
                                  'year1', 'year2', 'year3', 'year0', 'yearall', 'fromdate', '17', 'year5', '19', '20',
                                  '21', '22', '23', '24'])
    frame = frame.iloc[:, [0, 1, 2, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18]]
    frame.to_csv('fund.csv')
    return frame

二.筛选数据

 df_full =  df_full.sort_values(by='year1', axis=0, ascending=False)
    df = df_full.head(X)
    for xx in ['year1','year2','year3','year5','month6']:
        tmp = df_full.sort_values(by=xx, axis=0, ascending=False).head(X)
        df=df.merge(tmp,on=['code'])

    df=df.iloc[:,0:15]
    df.to_csv('result.csv')
    print('1年选取:')
    print(df)
    print('')

完整代码

import pandas as pd
import requests
import datetime


def get_html(url):  # 取得HTML文本
    try:
        r = requests.get(url)
        r.raise_for_status()
        r.encoding = 'utf-8'
        return r.text
    except:
        return ""


# 转换list维度,从1维到2维
def split_list(datas, n):
    length = len(datas)
    size = length // n + 1 if length % n else length // n
    _datas = []
    for i in range(size):
        start = i * n
        end = (i + 1) * n
        _datas.append(datas[start: end])
    return _datas


def get_fund():  # 取得基金列表
    # 先凑一个我们需要的URL出来
    max_jj = '5000'  # 调试5 工作5000
    fromstr = datetime.datetime.now().strftime('%Y-%m-%d')
    url = "http://fund.eastmoney.com/data/rankhandler.aspx?op=ph&dt=kf&ft=all&rs=&gs=0&sc=zzf&st=desc&sd=#custday&ed=#nowdate&qdii=&tabSubtype=,,,,,&pi=1&pn=#count&dx=1"
    url = url.replace('#count', max_jj)
    url = url.replace('#nowdate', fromstr)
    tostr = (datetime.datetime.now() - datetime.timedelta(days=5 * 365 + 1)).strftime('%Y-%m-%d')
    url = url.replace('#custday', tostr)
    # 取得文本
    s = get_html(url)
    # 去掉冗余信息,这里应该有更好的方法,但懒得折腾了,这样简单,十分钟搞定
    s = s[22:-159]
    for x in ['"', "'", ']', '[']:
        s = s.replace(x, '')
    lst = s.split(',')
    lst = split_list(lst, 25)
    frame = pd.DataFrame(lst,
                         columns=['code', 'name', 'py', '3', '4', 'jz', 'day1', 'week1', 'month1', 'month3', 'month6',
                                  'year1', 'year2', 'year3', 'year0', 'yearall', 'fromdate', '17', 'year5', '19', '20',
                                  '21', '22', '23', '24'])
    frame = frame.iloc[:, [0, 1, 2, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18]]
    frame.to_csv('fund.csv')
    return frame


def main():
    get_fund()  # 如果每次都需要用最新数据,用这句
    df_full = pd.read_csv('fund.csv')  # 节省网络流量,就用这句
    X = 500  # 取排名前多少的基金

    df_full =  df_full.sort_values(by='year1', axis=0, ascending=False)
    df = df_full.head(X)
    for xx in ['year1','year2','year3','year5','month6']:
        tmp = df_full.sort_values(by=xx, axis=0, ascending=False).head(X)
        df=df.merge(tmp,on=['code'])

    df=df.iloc[:,0:15]
    df.to_csv('result.csv')
    print('1年选取:')
    print(df)
    print('')

    ## 三年以来
    y3_index = df_full.sort_values(by=['year3'], ascending=False).head(X).name
    ## 二年以来
    y2_index = df_full.sort_values(by=['year2'], ascending=False).head(X).name
    ## 一年以来
    y1_index = df_full.sort_values(by=['year1'], ascending=False).head(X).name
    ## 六月以来
    m6_index = df_full.sort_values(by=['month6'], ascending=False).head(X).name
    ## 三月以来
    m3_index = df_full.sort_values(by=['month3'], ascending=False).head(X).name
    ## 一月以来
    m1_index = df_full.sort_values(by=['month1'], ascending=False).head(X).name
    ## 一周以来
    w1_index = df_full.sort_values(by=['week1'], ascending=False).head(X).name
    y3_index_set = set(y3_index)
    y2_index_set = set(y2_index)
    y1_index_set = set(y1_index)
    m6_index_set = set(m6_index)
    m3_index_set = set(m3_index)
    m1_index_set = set(m1_index)
    w1_index_set = set(w1_index)
    print('3年,2年,1年选取:')
    print(y3_index_set & y2_index_set & y1_index_set)
    print('')
    print('1年,半年,3月,1月选取:')
    print(y1_index_set,m6_index_set & m3_index_set & m1_index_set)
    print('')
    print('3年,2年,1年,半年,3个月,1个月选取:')
    print(y3_index_set & y2_index_set & y1_index_set & m6_index_set & m3_index_set & m1_index_set)
    print('')
    print('3年,2年,1年,半年,3个月,1个月,一周选取:')
    print(y3_index_set & y2_index_set & y1_index_set & m6_index_set & m3_index_set & m1_index_set & w1_index_set)


main()

201802101053029521

本文标签
 {{tag}}
点了个评