Question

我想在数据框中添加一个new_column，其中指示符为'LH'或'RH'。我已经在Additional_info列上尝试了以下代码。

LH = ['lhd','lh','lhd','left','le']
RH = [ 'rhd','rh','rhd','right,'re']
lh_rh= match_id[['MATA_info','tech_info','additional_info']]
lh_rh['additional_info']= lh_rh['additional_info'].str.lower()
Right = lh_rh.loc[lh_rh['additional_info'].isin(RH)]
left = lh_rh.loc[lh_rh['additional_info'].isin(LH)]

如果LH和RH中的关键字与“ MATA_info”，“ tech_info”和“ additional_info”列匹配，则我想进行匹配，则应创建名称为“ Relation”的其他列并将其分配给LH或相对湿度。

   MATA_info  tech_info   additional_info  Relation
3,50X085Right    F85        NAN              RH
3,50X085Left     F85        lh               LH

Answer 1

将str.contains与apply结合使用，然后检查是否DataFrame.any每行至少有一个True，最后将其传递给numpy.select：

LH = ['lhd','lh','lhd','left','le']
RH = [ 'rhd','rh','rhd','right','re']
lh_rh= match_id[['MATA_info','tech_info','additional_info']]
m1 = lh_rh.apply(lambda x: x.str.contains('|'.join(LH), na=False, case=False)).any(axis=1)
m2 = lh_rh.apply(lambda x: x.str.contains('|'.join(RH), na=False, case=False)).any(axis=1)

match_id['Relation'] = np.select([m1, m2], ['LH','RH'], default=np.nan)

print (match_id)
       MATA_info tech_info additional_info Relation
0  3,50X085Right       F85             NAN       RH
1   3,50X085Left       F85              lh       LH
2           4,56        %T              jj      nan

编辑：

pat1 = '|'.join(r"\b{}\b".format(x) for x in LH)
pat2 = '|'.join(r"\b{}\b".format(x) for x in RH)
m1 = lh_rh.apply(lambda x: x.str.contains(pat1, na=False, case=False)).any(axis=1)
m2 = lh_rh.apply(lambda x: x.str.contains(pat2, na=False, case=False)).any(axis=1)

字符串匹配并相应地分配标签

1 个答案: