追梦小狂魔的博客

max_min_scaler = lambda x : (x-np.min(x))/(np.max(x)-np.min(x))train_df['average_montly_hours']=train_df[['average_montly_hours']].apply(max_min_scaler)train_df

colorMap = {elem:index+1 for index,elem in enumerate(set(train_df["salary"]))}train_df['salary'] = train_df['salary'].map(colorMap)train_df

trian里面有id和isbn 分别在user和book的csv文件中，布标整合成一个trainbook里面的信息都有isbn，根据这个整合train=train_df.copy()# for use in list(user.columns[1:].values):# train[use]=Nonefor bo in list(book.columns[1:].values): train[bo]=Nonetrain先创建出列名和内容，替换for i in tq

右键选择记事本方式打开编码方式如图另存为这三步都选中，直接替换原格式，变成utf-8

train[['label','Western']]=train[['Western','label']]train这样把两列数据换了过来，但是标签还没换过来

使用sklearnfrom sklearn.utils import shuffle df = shuffle(df) 或者用自带的df.sample(frac=1)frac决定打乱的比例，例如0.3就是30%保持索引df.sample(frac=1).reset_index(drop=True)