使用Pandas apply基于GPA、SAT生成新列报TypeError错误如何解决

错误1：运算符优先级误用 #

& 作为位运算符优先级远高于比较运算符（>/</==等），你写的 x['GPA W'] < 2.0 & x['SAT'] < 940 会被Python优先计算 2.0 & x['SAT']，浮点数和SAT分数/NaN做位运算直接触发类型错误。
在单值条件判断（apply逐行处理的是单个标量值）场景下直接用逻辑运算符 and 即可，不需要用 &。

错误2：NaN值判断方法错误 #

NaN 是特殊浮点值，和任何值（包括自身）做 == 比较都会返回False，不能用 x['SAT'] == NaN 判断空值，需要用 pd.isna(x['SAT']) 兼容判断各类空值。

修正后完整代码 #

import pandas as pd

def f(x):
    gpa = x['GPA W']
    sat = x['SAT']
    if gpa < 2.0 and pd.isna(sat): return "2 yr"
    elif gpa < 2.0 and sat < 940: return "2 yr"
    elif gpa < 2.0 and sat >= 940 and sat <= 1050: return "Non-Selective 4 yr"
    elif gpa < 2.0 and sat >= 1060 and sat <= 1150: return "Somewhat Selective"
    elif gpa < 2.0 and sat >=1160: return "Somewhat Selective"
    elif gpa >= 2.0 and gpa <=2.4 and pd.isna(sat): return "Non-Selective 4 yr"
    elif gpa >= 2.0 and gpa <=2.4 and sat < 940: return "Non-Selective 4 yr"
    elif gpa >= 2.0 and gpa <=2.4 and sat >= 940 and sat <= 1050: return "Somewhat Selective"
    elif gpa >= 2.0 and gpa <=2.4 and sat >= 1060 and sat <= 1150: return "Somewhat Selective"
    elif gpa >= 2.0 and gpa <=2.4 and sat > 1160: return "Selective/Very Selective"
    elif gpa >= 2.5 and gpa <=2.9 and pd.isna(sat): return "Somewhat Selective"
    elif gpa >= 2.5 and gpa <=2.9 and sat < 940: return "Somewhat Selective"
    elif gpa >= 2.5 and gpa <=2.9 and sat >= 940 and sat <= 1050: return "Somewhat Selective"
    elif gpa >= 2.5 and gpa <=2.9 and sat >= 1060 and sat <= 1150: return "Selective"
    elif gpa >= 2.5 and gpa <=2.9 and sat > 1160: return "Selective/Very Selective"
    elif gpa >= 3.0 and gpa <=3.4 and pd.isna(sat): return "Selective"
    elif gpa >= 3.0 and gpa <=3.4 and sat < 940: return "Somewhat Selective"
    elif gpa >= 3.0 and gpa <=3.4 and sat >= 940 and sat <= 1050: return "Selective"
    elif gpa >= 3.0 and gpa <=3.4 and sat >= 1060 and sat <= 1150: return "Selective/Very Selective"
    elif gpa >= 3.0 and gpa <=3.4 and sat > 1160: return "Very Selective"
    elif gpa >= 3.5 and gpa <=4.0 and pd.isna(sat): return "Selective"
    elif gpa >= 3.5 and gpa <=4.0 and sat < 940: return "Selective"
    elif gpa >= 3.5 and gpa <=4.0 and sat >= 940 and sat <= 1050: return "Selective/Very Selective"
    elif gpa >= 3.5 and gpa <=4.0 and sat >= 1060 and sat <= 1150: return "Selective/Very Selective"
    elif gpa >= 3.5 and gpa <=4.0 and sat > 1160: return "Very Selective"
    else: return "Unknown"

dat['Selectivity'] = dat.apply(f, axis=1)

可选优化建议 #

如果数据量较大（十万行以上），可以改用pd.cut打GPA分段标签+np.select向量化判断的写法，执行性能比逐行apply高数十倍；也可以把重复的GPA范围判断提取为公共变量，降低代码冗余和写错的概率。

内容的提问来源于stack exchange，提问作者Mkmkmkmk