我不明白为什么第一个例子不起作用而第二个例子起作用。它们是相同的代码，了解为什么不起作用是很有帮助的，谢谢帮助。

def naive_bayes_eval(test_sms_file, f):    Precision=1    est_spam =0    pas_spam =0    nbr_spam =0    for ligne in open(test_sms_file, 'r').readlines():        if("spam" in ligne):            nbr_spam = nbr_spam + 1            if f(ligne.replace('spam',''))== 1:                est_spam=est_spam+1        if("ham" in ligne):            if( f(ligne.replace('ham',''))== 1):                pas_spam = pas_spam+1    recall = est_spam/nbr_spam;             if (est_spam + pas_spam)>0:        Precision = est_spam/(est_spam + pas_spam)    return(recall,Precision)

def naive_bayes_eval(test_sms_file, f):    Precision=1    est_spam =0    pas_spam =0    nbr_spam =0    f=open(test_sms_file, 'r')    lines=f.readlines()    for ligne in lines:        if("spam" in ligne):            nbr_spam = nbr_spam + 1            if f(ligne.replace('spam',''))== 1:                est_spam=est_spam+1        if("ham" in ligne):            if( f(ligne.replace('ham',''))== 1):                pas_spam = pas_spam+1    recall = est_spam/nbr_spam;             if (est_spam + pas_spam)>0:        Precision = est_spam/(est_spam + pas_spam)    return(recall,Precision)

回答：

你是说第一个有效而第二个无效吗？

def naive_bayes_eval(test_sms_file, f):    ...    f=open(test_sms_file, 'r')

这一行覆盖了f参数。