预处理归一化

import pandas as pd
df=pd.read_csv('energy.csv')




orilist=[]
for i in df['999']:
    if i[:4] not in orilist:
        orilist.append(i[:4])
dffinal = pd.DataFrame(columns=('999', '0', '1', '2', '3', '4', '5', '6'))
for j in orilist:
    df1 = pd.DataFrame(columns=('999', '0', '1', '2', '3', '4', '5', '6'))
##############################################2Lie
    for i in df['999']:
        newlist=[]
        if i[:4]==j:
            a=df.loc[df['999']==i]
            df1 = df1.append(a, ignore_index=True)
    # print(df1)
    s = (df1['2'] - df1['2'].min()) / (df1['2'].max() - df1['2'].min())
    df1 = df1.drop(['2'], axis=1)
    # print(df1)
    # 把规格化的那一列插入到数组中
    df1.insert(3, '2', s)

    # print(df1)
    # print(s)
##############################################3Lie
    s = (df1['3'] - df1['3'].min()) / (df1['3'].max() - df1['3'].min())
    df1 = df1.drop(['3'], axis=1)
    # print(df1)

    df1.insert(4, '3', s)
##############################################4Lie
    s = (df1['4'] - df1['4'].min()) / (df1['4'].max() - df1['4'].min())
    df1 = df1.drop(['4'], axis=1)
    # print(df1)

    df1.insert(5, '4', s)
##############################################4Lie
    s = (df1['5'] - df1['5'].min()) / (df1['5'].max() - df1['5'].min())
    df1 = df1.drop(['5'], axis=1)
    # print(df1)

    df1.insert(6, '5', s)
##############################################4Lie
    s = (df1['6'] - df1['6'].min()) / (df1['6'].max() - df1['6'].min())
    df1 = df1.drop(['6'], axis=1)
    # print(df1)

    df1.insert(7, '6', s)
    dffinal = dffinal.append(df1, ignore_index=True)

dffinal=dffinal.fillna(value=0)
print(dffinal)


dffinal.to_csv('final.csv')