Question

我正在尝试根据数据框（'cnms_df'中其他列的多个条件填充列['Mod_unMod']。我写了一些pseudo-code来解释这个困难：

IF (cnms_df['CA'] = "NO"):
    IF (cnms_df['I'] ="X" OR cnms_df['I']="V" OR cnms_df['I']="VE" OR cnms_df['I']="0.2 PCT ANNUAL CHANCE FLOOD HAZARD" OR cnms_df['I']="AREA NOT INCLUDED")
        cnms_df['Mod_unMod'] = "UnMapped"
    ELSE IF (LEFT(cnms_df['STUDY_TYPE].str[:3])="NON")
        cnms_df['Mod_unMod'] ="NON"

    ELSE IF (cnms_df['BJ']="X" OR cnms_df['BJ']="V" OR cnms_df['BJ']="V" OR cnms_df['BJ']="VE" OR cnms_df['BJ']="0.2 PCT ANNUAL CHANCE FLOOD HAZARD" OR cnms_df['BJ']="AREA NOT INCLUDED")   
        cnms_df['Mod_unMod']
    ELSE IF (LEFT(BK2,3)="NON" 
        cnms_df['Mod_unMod']="UnMod"
    ELSE:
        cnms_df['Mod_unMod']="Modernized")
ELSE:
    cnms_df['Mod_unMod'] = "UnMapped"

我已经应用了简单的np.where语句，但是我不确定如何在上述级别上做到这一点。有没有办法以合乎逻辑的方式做这样的事情？

cnms_df['Mod_unMod'] = np.where((cnms_df['CA'] == 'No') & ((cnms_df['I'] ="X") | (cnms_df['I']="V") | (cnms_df['I']="VE") | (cnms_df['I']="0.2 PCT ANNUAL CHANCE FLOOD HAZARD") | (cnms_df['I']="AREA NOT INCLUDED")), "UnMapped", "Modernized")

Answer 1

定义函数以生成新列中的内容当前行为：

def fun(row):
    lst = ['X', 'V', 'VE', '0.2 PCT ANNUAL CHANCE FLOOD HAZARD', 'AREA NOT INCLUDED']
    if row.CA != 'NO:
        return 'UnMapped'
    elif row.I in lst:
        return 'UnMapped'
    elif row.STUDY_TYPE[0:3] == 'NON':
        return 'NON'
    elif row.BJ in lst:
        return 'Mod_unMod'
    elif row.BK2[0:3] == 'NON':
        return 'UnMod'
    else:
        return 'Modernized'

然后将此功能应用于每一行，并将结果保存在新列中：

df['Mod_unMod'] = df.apply(fun, axis=1)

基于多级条件在数据框中填充新列

1 个答案: