Given the following data frames join\merge df1 and df2 to result in df3 in the following manner:
df1.a join on regex matching to df2.a AND df1.b join on regex matching to df2.b
df1 = pd.DataFrame({'a': ['foo', 'bar','zoolo','foo','foo'], 'b': ['tty','abc', 'uhg','abc','tty'], 'c':[14,72,93,33,57]})
|
a |
b |
c |
| 0 |
foo |
tty |
14 |
| 1 |
bar |
abc |
72 |
| 2 |
zoolo |
uhg |
93 |
| 3 |
foo |
abc |
33 |
| 4 |
foo |
tty |
57 |
df2 = pd.DataFrame({'a': ['zool(o|r)','foo|oof','[badr] '],'b': ['uh(f|g)','[ty] ','abc|bac'], 'j': [11,32,65]})
|
a |
b |
j |
| 0 |
zool(o|r) |
uh(f|g) |
11 |
| 1 |
foo|oof |
[ty] |
32 |
| 2 |
[badr] |
abc|bac |
65 |
Result
df3 = pd.DataFrame({'a': ['foo', 'bar','zoolo','foo'], 'b': ['tty','abc', 'uhg','tty'], 'c':[14,72,93,57],'j': [32,65,11,32]}, )
|
a |
b |
c |
j |
| 0 |
foo |
tty |
14 |
32 |
| 1 |
bar |
abc |
72 |
65 |
| 2 |
zoolo |
uhg |
93 |
11 |
| 4 |
foo |
tty |
57 |
32 |
CodePudding user response：
def func(dfRow):
for index, row in df2.iterrows():
if re.match(row["a"], dfRow["a"]) and re.match(row["b"], dfRow["b"]):
return dfRow.append(row[["j"]])
df1.apply(func, axis=1).dropna()
