{"id":204,"date":"2021-09-15T22:34:15","date_gmt":"2021-09-15T14:34:15","guid":{"rendered":"\/?p=204"},"modified":"2021-09-19T21:13:42","modified_gmt":"2021-09-19T13:13:42","slug":"pandas%e7%9a%84apply%e6%96%b9%e6%b3%95","status":"publish","type":"post","link":"\/?p=204","title":{"rendered":"Pandas\u7684apply\u65b9\u6cd5"},"content":{"rendered":"

\u5728\u5904\u7406\u80a1\u7968\u6570\u636e\u65f6\uff0c\u7ecf\u5e38\u9700\u8981\u628a\u4e00\u4e9b\u51fd\u6570\u5e94\u7528\u4e8e\u591a\u884c\u6216\u8005\u591a\u5217\u6570\u636e\u4e0a\uff0c\u8fd9\u65f6\u5c31\u9700\u8981\u4f7f\u7528Series\u548cDataFrame\u7684apply\u65b9\u6cd5\u3002\u672c\u6587\u4ee5sz.002407\u591a\u6c1f\u591a\u57282021\u5e74\u7684\u524d10\u6839\u65e5\u7ebf\u6570\u636e\uff08\u5982\u4e0b\u8868\u6240\u793a\uff09\u4e3a\u4f8b\uff0c\u5c55\u793aapply\u65b9\u6cd5\u7684\u4f7f\u7528\uff0c\u6211\u4eec\u9996\u5148\u5c06\u6570\u636e\u52a0\u8f7d\u5230\u4e00\u4e2a\u7c7b\u578b\u4e3aDataFrame\u7684\u5bf9\u8c61df\u4e2d\u3002<\/p>\n

\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
date<\/th>\nopen<\/th>\nhigh<\/th>\nlow<\/th>\nclose<\/th>\npreclose<\/th>\nvolume<\/th>\namount<\/th>\nadjustflag<\/th>\nturn<\/th>\ntradestatus<\/th>\npctChg<\/th>\npeTTM<\/th>\npbMRQ<\/th>\npsTTM<\/th>\npcfNcfTTM<\/th>\nisST<\/th>\n<\/tr>\n<\/thead>\n
2021-01-04<\/td>\n20.0300000000<\/td>\n21.4000000000<\/td>\n19.9300000000<\/td>\n20.7300000000<\/td>\n20.0300000000<\/td>\n69639314<\/td>\n1442558998.1400<\/td>\n2<\/td>\n12.026700<\/td>\n1<\/td>\n3.494800<\/td>\n-28.520577<\/td>\n5.001054<\/td>\n3.874122<\/td>\n-336.943766<\/td>\n0<\/td>\n<\/tr>\n
2021-01-05<\/td>\n20.5100000000<\/td>\n21.3100000000<\/td>\n19.9500000000<\/td>\n20.6800000000<\/td>\n20.7300000000<\/td>\n68155914<\/td>\n1404877488.0600<\/td>\n2<\/td>\n11.770500<\/td>\n1<\/td>\n-0.241200<\/td>\n-28.451786<\/td>\n4.988992<\/td>\n3.864778<\/td>\n-336.131070<\/td>\n0<\/td>\n<\/tr>\n
2021-01-06<\/td>\n20.8300000000<\/td>\n21.2800000000<\/td>\n19.0600000000<\/td>\n19.7700000000<\/td>\n20.6800000000<\/td>\n69108077<\/td>\n1375667187.1400<\/td>\n2<\/td>\n11.935000<\/td>\n1<\/td>\n-4.400400<\/td>\n-27.199798<\/td>\n4.769457<\/td>\n3.694713<\/td>\n-321.340003<\/td>\n0<\/td>\n<\/tr>\n
2021-01-07<\/td>\n19.5500000000<\/td>\n20.4400000000<\/td>\n19.2100000000<\/td>\n19.8900000000<\/td>\n19.7700000000<\/td>\n57333679<\/td>\n1145163382.0700<\/td>\n2<\/td>\n9.901500<\/td>\n1<\/td>\n0.607000<\/td>\n-27.364895<\/td>\n4.798406<\/td>\n3.717139<\/td>\n-323.290473<\/td>\n0<\/td>\n<\/tr>\n
2021-01-08<\/td>\n19.8900000000<\/td>\n20.1300000000<\/td>\n18.6000000000<\/td>\n19.4900000000<\/td>\n19.8900000000<\/td>\n55294505<\/td>\n1063035986.1300<\/td>\n2<\/td>\n9.549300<\/td>\n1<\/td>\n-2.011100<\/td>\n-26.814570<\/td>\n4.701908<\/td>\n3.642385<\/td>\n-316.788905<\/td>\n0<\/td>\n<\/tr>\n
2021-01-11<\/td>\n18.9000000000<\/td>\n18.9000000000<\/td>\n17.5400000000<\/td>\n17.5500000000<\/td>\n19.4900000000<\/td>\n77040397<\/td>\n1371939544.2400<\/td>\n2<\/td>\n13.304900<\/td>\n1<\/td>\n-9.953800<\/td>\n-24.145496<\/td>\n4.233888<\/td>\n3.279829<\/td>\n-285.256300<\/td>\n0<\/td>\n<\/tr>\n
2021-01-12<\/td>\n17.2000000000<\/td>\n18.3500000000<\/td>\n16.8300000000<\/td>\n18.1300000000<\/td>\n17.5500000000<\/td>\n55091872<\/td>\n970628000.0300<\/td>\n2<\/td>\n9.514400<\/td>\n1<\/td>\n3.304800<\/td>\n-24.943466<\/td>\n4.373811<\/td>\n3.388222<\/td>\n-294.683573<\/td>\n0<\/td>\n<\/tr>\n
2021-01-13<\/td>\n18.2000000000<\/td>\n19.5300000000<\/td>\n18.2000000000<\/td>\n18.7600000000<\/td>\n18.1300000000<\/td>\n68535355<\/td>\n1303591461.9500<\/td>\n2<\/td>\n11.836000<\/td>\n1<\/td>\n3.474900<\/td>\n-25.810228<\/td>\n4.525797<\/td>\n3.505959<\/td>\n-304.923543<\/td>\n0<\/td>\n<\/tr>\n
2021-01-14<\/td>\n18.6900000000<\/td>\n19.6200000000<\/td>\n18.1200000000<\/td>\n19.4000000000<\/td>\n18.7600000000<\/td>\n58518764<\/td>\n1109560714.8600<\/td>\n2<\/td>\n10.106200<\/td>\n1<\/td>\n3.411500<\/td>\n-26.690747<\/td>\n4.680195<\/td>\n3.625566<\/td>\n-315.326052<\/td>\n0<\/td>\n<\/tr>\n
2021-01-15<\/td>\n19.0700000000<\/td>\n19.6800000000<\/td>\n18.1300000000<\/td>\n19.3000000000<\/td>\n19.4000000000<\/td>\n55752774<\/td>\n1052651676.0200<\/td>\n2<\/td>\n9.628500<\/td>\n1<\/td>\n-0.515500<\/td>\n-26.553166<\/td>\n4.656071<\/td>\n3.606877<\/td>\n-313.700660<\/td>\n0<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n
\n

Series\u7684apply\u65b9\u6cd5<\/h1>\n

\u5728Pandas\u4e2d\uff0c\u4eceDataFrame\u53d6\u51fa1\u884c\u6216\u80051\u5217\uff0c\u90fd\u4f1a\u8fd4\u56deSeries\u5bf9\u8c61\uff0c\u4f8b\u5982\uff1a<\/p>\n

# \u53d61\u5217\u6570\u636e\nprint(type(df['date']))\n<\/code><\/pre>\n
<class 'pandas.core.series.Series'>\n<\/code><\/pre>\n
# \u53d61\u884c\u6570\u636e\nprint(type(df.iloc[0]))\n<\/code><\/pre>\n
<class 'pandas.core.series.Series'>\n<\/code><\/pre>\n
\u6211\u4eec\u4ee5date\u5217\u4e3a\u4f8b\uff0c\u6765\u4ecb\u7ecdSeries\u7684apply\u65b9\u6cd5\u7684\u4f7f\u7528\u3002<\/p>\n
\u5728\u539f\u59cb\u7684df\u4e2d\uff0cdate\u5217\u7684\u5143\u7d20\u7c7b\u578b\u5747\u4e3astr\uff0c\u6211\u4eec\u9996\u5148\u5c1d\u8bd5\u53bb\u6389\u5e74\u6708\u65e5\u4e4b\u95f4\u7684”-“\u7b26\u53f7\uff0c\u5373\u5c06”2021-01-04″\u8f6c\u5316\u4e3a”20210104″\u3002<\/p>\n
\u9996\u5148\u6253\u5370\u4e0b\u539f\u59cb\u7684date\u5217\u770b\u4e0b\u7ed3\u679c\uff1a<\/p>\n
print(df['date'])\n<\/code><\/pre>\n
0    2021-01-04\n1    2021-01-05\n2    2021-01-06\n3    2021-01-07\n4    2021-01-08\n5    2021-01-11\n6    2021-01-12\n7    2021-01-13\n8    2021-01-14\n9    2021-01-15\nName: date, dtype: object\n<\/code><\/pre>\n
\u4e0b\u9762\u6211\u4eec\u5b9a\u4e49\u4e00\u4e2a\u51fd\u6570\uff0c\u8f93\u5165\u53c2\u6570\u4e3a\u539f\u59cb\u5b57\u7b26\u4e32\uff0c\u8fd4\u56de\u503c\u4e3a\u66ff\u6362\u540e\u7684\u5b57\u7b26\u4e32\uff0c\u51fd\u6570\u5185\u90e8\u5b9e\u9645\u8c03\u7528\u7684\u662fstr\u7c7b\u7684replace\u51fd\u6570\uff1a<\/p>\n
def str_replace(s):\n    return s.replace('-', '')\n<\/code><\/pre>\n
\u63a5\u7740\u6211\u4eec\u5c06\u81ea\u5b9a\u4e49\u7684str_replace\u4f5c\u4e3a\u53c2\u6570\u4f20\u9012\u7ed9Series\u7684apply\u65b9\u6cd5\uff0c\u5c31\u53ef\u4ee5\u5e94\u7528\u5230Series\u7684\u6bcf\u4e2a\u5143\u7d20\u4e0a\uff1a<\/p>\n
sr = df['date'].apply(str_replace)\nprint(sr)\n<\/code><\/pre>\n
0    20210104\n1    20210105\n2    20210106\n3    20210107\n4    20210108\n5    20210111\n6    20210112\n7    20210113\n8    20210114\n9    20210115\nName: date, dtype: object\n<\/code><\/pre>\n
\u4ece\u6253\u5370\u7ed3\u679c\u53ef\u4ee5\u770b\u51fa\uff0c\u6240\u6709\u7684\u65e5\u671f\u90fd\u6309\u7167\u9884\u671f\u8fdb\u884c\u4e86\u4fee\u6539\u3002<\/p>\n
\u5bf9\u4e8e\u5e26\u591a\u4e2a\u53c2\u6570\u7684\u51fd\u6570\uff0c\u540c\u6837\u53ef\u4ee5\u4f7f\u7528apply\u65b9\u6cd5\u5e94\u7528\u5230Series\u7684\u6240\u6709\u5143\u7d20\u4e0a\u3002\u6211\u4eec\u5bf9str_replace\u51fd\u6570\u8fdb\u884c\u6539\u9020\uff1a<\/p>\n
def str_replace_v2(s, c=''):\n    return s.replace('-', c)\n<\/code><\/pre>\n
\u8be5\u51fd\u6570\u9ed8\u8ba4\u4f7f\u7528\u7a7a\u4e32\u6765\u66ff\u6362”-“\uff0c\u4e5f\u53ef\u4ee5\u4f7f\u7528\u6307\u5b9a\u7684\u5b57\u7b26\u4e32c\u6765\u66ff\u6362\u3002\u5728apply\u4e2d\u4f7f\u7528\u65f6\uff0c\u53ea\u9700\u8981\u6307\u5b9ac\u7684\u503c\u5373\u53ef\uff1a<\/p>\n
sr = df['date'].apply(str_replace_v2, c='\/')\nprint(sr)\n<\/code><\/pre>\n
0    2021\/01\/04\n1    2021\/01\/05\n2    2021\/01\/06\n3    2021\/01\/07\n4    2021\/01\/08\n5    2021\/01\/11\n6    2021\/01\/12\n7    2021\/01\/13\n8    2021\/01\/14\n9    2021\/01\/15\nName: date, dtype: object\n<\/code><\/pre>\n
\u53e6\uff1a\u4e5f\u53ef\u4ee5\u76f4\u63a5\u4f7f\u7528Series\u7684str\u65b9\u6cd5\uff0c\u800c\u4e0d\u9700\u8981\u81ea\u5b9a\u4e49\u51fd\u6570\uff0c\u6765\u5b9e\u73b0\u76f8\u540c\u7684\u529f\u80fd\uff1a<\/p>\n
sr = df['date'].str.replace('-', '')\nprint(sr)\n<\/code><\/pre>\n
\u5bf9\u4e8e\u5217Series\u8fdb\u884c\u7684\u64cd\u4f5c\uff0c\u6211\u4eec\u53ef\u4ee5\u628a\u8ba1\u7b97\u7ed3\u679c\u4fdd\u5b58\u4e3adf\u7684\u65b0\u5217\uff1a<\/p>\n
df['new_date'] = df['date'].apply(str_replace_v2, c='\/')\nprint(df)\n<\/code><\/pre>\n
         date   open   high    low  ...     psTTM   pcfNcfTTM  isST    new_date\n0  2021-01-04  20.03  21.40  19.93  ...  3.874122 -336.943766     0  2021\/01\/04\n1  2021-01-05  20.51  21.31  19.95  ...  3.864778 -336.131070     0  2021\/01\/05\n2  2021-01-06  20.83  21.28  19.06  ...  3.694713 -321.340003     0  2021\/01\/06\n3  2021-01-07  19.55  20.44  19.21  ...  3.717139 -323.290473     0  2021\/01\/07\n4  2021-01-08  19.89  20.13  18.60  ...  3.642385 -316.788905     0  2021\/01\/08\n5  2021-01-11  18.90  18.90  17.54  ...  3.279829 -285.256300     0  2021\/01\/11\n6  2021-01-12  17.20  18.35  16.83  ...  3.388222 -294.683573     0  2021\/01\/12\n7  2021-01-13  18.20  19.53  18.20  ...  3.505959 -304.923543     0  2021\/01\/13\n8  2021-01-14  18.69  19.62  18.12  ...  3.625566 -315.326052     0  2021\/01\/14\n9  2021-01-15  19.07  19.68  18.13  ...  3.606877 -313.700660     0  2021\/01\/15\n\n[10 rows x 18 columns]\n<\/code><\/pre>\n
\u4ece\u6253\u5370\u7ed3\u679c\u53ef\u4ee5\u770b\u5230\uff0cdf\u65b0\u589e\u4e86\u4e00\u5217new_date\uff0c\u4fdd\u5b58\u4e86apply\u7684\u8ba1\u7b97\u7ed3\u679c\u3002<\/p>\n
\n
DataFrame\u7684apply\u65b9\u6cd5<\/h1>\n
DataFrame\u7684apply\u4e0eSeries\u7684apply\u4f7f\u7528\u65b9\u6cd5\u7c7b\u4f3c\uff0c\u4f46\u662f\u9700\u8981\u6307\u5b9a\u51fd\u6570\u662f\u6309\u5217\u8fd8\u662f\u6309\u884c\u8fdb\u884c\uff0c\u901a\u8fc7axis\u53c2\u6570\u8fdb\u884c\u8bbe\u7f6e\uff0caxis\u9ed8\u8ba4\u503c\u4e3a0\uff0c\u8868\u793a\u6309\u5217\u5e94\u7528\u6307\u5b9a\u51fd\u6570\uff0caxis\u4e3a1\u65f6\uff0c\u8868\u793a\u6309\u884c\u5e94\u7528\u6307\u5b9a\u51fd\u6570\u3002<\/p>\n
    \n
  1. \u6309\u5217\u5e94\u7528<\/li>\n<\/ol>\n
    \u5047\u5982\u6211\u4eec\u60f3\u8ba1\u7b97open\u3001high\u3001low\u3001close\u5217\u7684\u5747\u503c\uff0c\u9996\u5148\u53d6\u51fa\u8fd94\u5217\uff0c\u770b\u4e00\u4e0b\u6570\u636e\u7c7b\u578b\uff1a<\/p>\n
    print(type(df[['open', 'high', 'low', 'close']]))\n<\/code><\/pre>\n
    <class 'pandas.core.frame.DataFrame'>\n<\/code><\/pre>\n
    \u53ef\u89c1df[['open', 'high', 'low', 'close']]<\/code>\u7684\u6570\u636e\u7c7b\u578b\u4e3aDataFrame\u3002<\/p>\n
    \u7136\u540e\uff0c\u6211\u4eec\u5b9a\u4e49\u8ba1\u7b97\u5747\u503c\u7684\u51fd\u6570\uff1a<\/p>\n
    def mean_value(row):\n    return numpy.mean(row)\n<\/code><\/pre>\n
    \u8fd9\u91cc\u8c03\u7528numpy\u7684mean\u51fd\u6570\u6765\u8ba1\u7b97\u5747\u503c\u3002\u6700\u540e\u6211\u4eec\u5c31\u53ef\u4ee5\u5229\u7528apply\u65b9\u6cd5\u6765\u8ba1\u7b97\u5404\u5217\u7684\u5747\u503c\uff1a<\/p>\n
    print(df[['open', 'high', 'low', 'close']].apply(mean_value))\n<\/code><\/pre>\n
    open     19.287\nhigh     20.064\nlow      18.557\nclose    19.370\ndtype: float64\n<\/code><\/pre>\n
    apply\u65b9\u6cd5\u8fd8\u662f\u8fd4\u56de\u7684Series\u7c7b\u578b\u7684\u5bf9\u8c61\uff0c\u91cc\u9762\u4fdd\u5b58\u4e86\u5404\u5217\u7684\u5747\u503c\u3002<\/p>\n
    \u4e0a\u9762\u7684\u4ee3\u7801\u53ea\u662f\u7528\u4e8e\u6f14\u793aapply\u7684\u4f7f\u7528\u65b9\u6cd5\uff0c\u4e5f\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u4ee3\u7801\u5b9e\u73b0\u76f8\u540c\u7684\u529f\u80fd\uff1a<\/p>\n
    print(df[['open', 'high', 'low', 'close']].mean())\n<\/code><\/pre>\n
    \u5982\u679c\u60f3\u5728\u5e94\u7528\u51fd\u6570\u4e2d\u8bbf\u95ee\u5230\u5177\u4f53\u5143\u7d20\uff0c\u9700\u8981\u7528\u7d22\u5f15\u6765\u8bbf\u95ee\u3002\u6211\u4eec\u7684\u793a\u4f8b\u4e2d\uff0c\u5404\u884c\u7684\u7d22\u5f15\u662f\u4ece0\u52309\uff0c\u5047\u8bbe\u6211\u4eec\u60f3\u8ba1\u7b97\u7b2c0\u30012\u30015\u884c\u7684\u5747\u503c\uff0c\u53ef\u4ee5\u8fd9\u6837\u6539\u9020\u4ee3\u7801\uff1a<\/p>\n
    def mean_value_v2(row):\n    return (row[0] + row[2] + row[5]) \/ 3\n\nprint(df[['open', 'high', 'low', 'close']].apply(mean_value_v2))\n<\/code><\/pre>\n
    open     19.920000\nhigh     20.526667\nlow      18.843333\nclose    19.350000\ndtype: float64\n<\/code><\/pre>\n
    \u5728\u4ee3\u7801\u4e2d\u53ef\u4ee5\u770b\u5230\uff0c\u4f7f\u7528row[0]\u3001row[2]\u8fd9\u6837\u7684\u65b9\u5f0f\u6765\u8bbf\u95ee\u5177\u4f53\u7684\u5143\u7d20\u3002<\/p>\n
      \n
    1. \u6309\u884c\u5e94\u7528<\/li>\n<\/ol>\n
      \u6309\u884c\u5e94\u7528\u53ef\u4ee5\u5b9e\u73b0DataFrame\u5355\u5217\u6216\u8005\u591a\u5217\u4e4b\u95f4\u7684\u8fd0\u7b97\u3002<\/p>\n
      \u6211\u4eec\u4ee5\u8ba1\u7b97\u632f\u5e45=(high-low)\/preclose\u4e3a\u4f8b\uff1a<\/p>\n
      def amp(col):\n    return (col['high'] - col['low']) \/ col['preclose']\n\nprint(df.apply(amp, axis=1))\n<\/code><\/pre>\n
      0    0.073390\n1    0.065605\n2    0.107350\n3    0.062215\n4    0.076923\n5    0.069779\n6    0.086610\n7    0.073359\n8    0.079957\n9    0.079897\ndtype: float64\n<\/code><\/pre>\n
      \u4e5f\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u4e24\u79cdlambda\u8868\u8fbe\u5f0f\u6765\u5b9e\u73b0\u5747\u53ef\uff1a<\/p>\n
      print(df.apply(lambda x: (x.high - x.low) \/ x.preclose, axis=1))\nprint(df.apply(lambda x: (x['high'] - x['low']) \/ x['preclose'], axis=1))\n<\/code><\/pre>\n
      \u5982\u679c\u60f3\u628a\u8ba1\u7b97\u7ed3\u679c\u4fdd\u5b58\u5728df\u7684\u5217\u4e2d\uff0c\u53ef\u4ee5\u5982\u4e0b\u5b9e\u73b0\uff1a<\/p>\n
      df['amp'] = df.apply(lambda x: (x.high - x.low) \/ x.preclose, axis=1)\nprint(df)\n<\/code><\/pre>\n
               date   open   high    low  ...     psTTM   pcfNcfTTM  isST       amp\n0  2021-01-04  20.03  21.40  19.93  ...  3.874122 -336.943766     0  0.073390\n1  2021-01-05  20.51  21.31  19.95  ...  3.864778 -336.131070     0  0.065605\n2  2021-01-06  20.83  21.28  19.06  ...  3.694713 -321.340003     0  0.107350\n3  2021-01-07  19.55  20.44  19.21  ...  3.717139 -323.290473     0  0.062215\n4  2021-01-08  19.89  20.13  18.60  ...  3.642385 -316.788905     0  0.076923\n5  2021-01-11  18.90  18.90  17.54  ...  3.279829 -285.256300     0  0.069779\n6  2021-01-12  17.20  18.35  16.83  ...  3.388222 -294.683573     0  0.086610\n7  2021-01-13  18.20  19.53  18.20  ...  3.505959 -304.923543     0  0.073359\n8  2021-01-14  18.69  19.62  18.12  ...  3.625566 -315.326052     0  0.079957\n9  2021-01-15  19.07  19.68  18.13  ...  3.606877 -313.700660     0  0.079897\n\n[10 rows x 18 columns]\n<\/code><\/pre>\n