Question

我有一个DataFrame和Series条件，如下所示：

import pandas as pd
import numpy as np
dates = pd.date_range('20170801','20170805')
df = pd.DataFrame({'Sales:[5,10,15,20,25]'}, index = dates)
error_days = pd.Series([True, True,False, False, True], index=df.index)

所以，df是

             Sales
2017-08-01      5
2017-08-02     10
2017-08-03     15
2017-08-04     20
2017-08-05     25

error_days是

2017-08-01     True
2017-08-02     True
2017-08-03    False
2017-08-04    False
2017-08-05     True

如果原始销售额＆gt; 8，我会将那些error_days的销售额更改为8，因此df结果为

             Sales
2017-08-01      5
2017-08-02     8
2017-08-03     15
2017-08-04     20
2017-08-05     8

怎么做？

我知道

df['sales'][error_days] = 8

会更改所有值，但我只想将这些销售额＆gt; 8更改为8。

BTW，请不要在df。

中添加新列

感谢。

Answer 1

首先需要&的链式条件和另一个掩码df['Sales'] > 8。

<强> 1

使用loc：

df.loc[error_days & (df['Sales'] > 8), 'Sales'] = 8 
print (df)
            Sales
2017-08-01      5
2017-08-02      8
2017-08-03     15
2017-08-04     20
2017-08-05      8

<强> 2

或mask：

df['Sales'] = df['Sales'].mask(error_days & (df['Sales'] > 8), 8) 
print (df)
            Sales
2017-08-01      5
2017-08-02      8
2017-08-03     15
2017-08-04     20
2017-08-05      8

第3

或numpy.where：

df['Sales'] = np.where(error_days & (df['Sales'] > 8), 8, df['Sales']) print (df) Sales 2017-08-01 5 2017-08-02 8 2017-08-03 15 2017-08-04 20 2017-08-05 8

Answer 2

IIUC我们可以在这里使用Series.clip_upper()：

export const thumbnails = {
    'cat1': require('./images/cat1.png'),
    'cat2': require('./images/cat2.png'),
    'cat3': require('./images/cat3.png'),
    'cat4': require('./images/cat4.png'),
    'cat5': require('./images/cat5.png'),
    'cat6': require('./images/cat6.png'),
    'cat7': require('./images/cat7.png'),
    'cat8': require('./images/cat8.png'),
    'cat9': require('./images/cat9.png'),
    'cat10': require('./images/cat10.png'),
    'cat11': require('./images/cat11.png'),
    'cat12': require('./images/cat12.png'),
};

export default class App extends React.Component {
  // For List View
  constructor() {
    super();
    const ds = new ListView.DataSource({rowHasChanged: (r1, r2) => r1 !== r2});
    this.state = {
      dataSource: ds.cloneWithRows(['cat1', 'cat2', 'cat3', 'cat4', 'cat5', 'cat6', 'cat7', 'cat8', 'cat9', 'cat10', 'cat11', 'cat12', ]),
    };
  }

  render() {
    return (
        <View style={{flex: 1}}>
            <ScrollableTabView 
      renderTabBar={() => <CustomTabBar />}> 
                <ListView tabLabel="tab1" contentContainerStyle={{flexDirection: 'row', flexWrap: 'wrap', alignItems: 'flex-start',}}
style={{flex:1}} dataSource={this.state.dataSource} renderRow={(rowData) => <Image resizeMode="contain" style={{width: width / 3, height: width / 3}} source={thumbnails[rowData]} />} />
                <View tabLabel="tab2" style={{alignItems: 'center', justifyContent: 'center', height: 300}}>
                    <Text style={{color: '#000'}}>Page 2 here</Text>
                </View>
                <View tabLabel="tab3" style={{alignItems: 'center', justifyContent: 'center', height: 300}}>
                    <Text style={{color: '#000'}}>Page 3 here</Text>
                </View>
                <View tabLabel="tab4" style={{alignItems: 'center', justifyContent: 'center', height: 300}}>
                    <Text style={{color: '#000'}}>Page 4 here</Text>
                </View>
            </ScrollableTabView> 
        </View>
    );
  }
}

如何根据一个列值和其他条件更改DataFrame列值

2 个答案: