-1

I have two dataframes,

new1.
      Name       city
 0    sri won    chn
 1    pechi won  pune
 2    Ram won    mum
 0    pec won    kerala

new3
    req
0   pec
1   mut

I tried, 

mask=new1.Name.str.contains("|".join(new3.req.values.tolist()))
new1[mask]

I am getting,

 new1[mask]
      Name       city
 1  pechi won    pune
 0  pec won      kerala

As "pechi" contains "pec", it took this valu. but I want the exact match between the values not "contains"

my desired output is,

 new1[mask]
      Name       city
 0  pec won      kerala
Pyd
  • 6,017
  • 18
  • 52
  • 109

2 Answers2

1

You need \b that means "word boundary":

a = r'\b(' + "|".join(new3.req.values.tolist()) + r')\b'
print (a)
\b(pec|mut)\b

mask=new1.Name.str.contains(a)
df = new1[mask]
print (df)
      Name    city
0  pec won  kerala
jezrael
  • 822,522
  • 95
  • 1,334
  • 1,252
  • wow!, It worked perfectly, could you please explain what the first line of code does ??? – Pyd Aug 09 '17 at 06:40
  • You can check [this](https://stackoverflow.com/a/10138154/2901002) for explain `word boundary` (my English is horrible, especially for deep explanations) – jezrael Aug 09 '17 at 06:42
0

You need space in separator

In [1350]: new1
Out[1350]:
        Name    city
0    sri won     chn
1  pechi won    pune
2    Ram won     mum
0    pec won  kerala

In [1351]: new3
Out[1351]:
   req
0  pec
1  mut

In [1352]: ' | '.join(new3.req)
Out[1352]: 'pec | mut'

In [1353]: new1.Name.str.contains(' | '.join(new3.req))
Out[1353]:
0    False
1    False
2    False
0     True
Name: Name, dtype: bool

In [1354]: new1[new1.Name.str.contains(' | '.join(new3.req))]
Out[1354]:
      Name    city
0  pec won  kerala
Zero
  • 74,117
  • 18
  • 147
  • 154