I want to create a new column in my dataframe if the column contains any of the values from a column of a second dataframe.
First dataframe
WXYnineZAB
EFGsixHIJ
QRSeightTUV
GHItwoJKL
YZAfiveBCD
EFGsixHIJ
MNOthreePQR
ABConeDEF
MNOthreePQR
MNOthreePQR
YZAfiveBCD
WXYnineZAB
GHItwoJKL
KLMsevenNOP
EFGsixHIJ
ABConeDEF
KLMsevenNOP
QRSeightTUV
STUfourVWX
STUfourVWX
KLMsevenNOP
WXYnineZAB
CDEtenFGH
YZAfiveBCD
CDEtenFGH
QRSeightTUV
ABConeDEF
STUfourVWX
CDEtenFGH
GHItwoJKL
Second Dataframe
one
three
five
seven
nine
Output DataFrame
WXYnineZAB,nine
EFGsixHIJ,***
QRSeightTUV,***
GHItwoJKL,***
YZAfiveBCD,five
EFGsixHIJ,***
MNOthreePQR,three
ABConeDEF,one
MNOthreePQR,three
MNOthreePQR,three
YZAfiveBCD,five
WXYnineZAB,nine
GHItwoJKL,***
KLMsevenNOP,seven
EFGsixHIJ,***
ABConeDEF,one
KLMsevenNOP,seven
QRSeightTUV,***
STUfourVWX,***
STUfourVWX,***
KLMsevenNOP,seven
WXYnineZAB,nine
CDEtenFGH,***
YZAfiveBCD,five
CDEtenFGH,***
QRSeightTUV,***
ABConeDEF,one
STUfourVWX,***
CDEtenFGH,***
GHItwoJKL,***
To explain it easily I made the first dataframe be 3chars + search string + 3chars, but my actual file doesn't have any consistency like this.
Source DFs:
In [172]: d1
Out[172]:
txt
0 WXYnineZAB
1 EFGsixHIJ
2 QRSeightTUV
3 GHItwoJKL
4 YZAfiveBCD
.. ...
25 QRSeightTUV
26 ABConeDEF
27 STUfourVWX
28 CDEtenFGH
29 GHItwoJKL
[30 rows x 1 columns]
In [173]: d2
Out[173]:
word
0 one
1 three
2 five
3 seven
4 nine
generate RegEx pattern from the second DataFrame:
In [174]: pat = r'({})'.format(d2['word'].str.cat(sep='|'))
In [175]: pat
Out[175]: '(one|three|five|seven|nine)'
extract words matching the RegEx pattern and assign them as a new column:
In [176]: d1['new'] = d1['txt'].str.extract(pat, expand=False)
In [177]: d1
Out[177]:
txt new
0 WXYnineZAB nine
1 EFGsixHIJ NaN
2 QRSeightTUV NaN
3 GHItwoJKL NaN
4 YZAfiveBCD five
.. ... ...
25 QRSeightTUV NaN
26 ABConeDEF one
27 STUfourVWX NaN
28 CDEtenFGH NaN
29 GHItwoJKL NaN
[30 rows x 2 columns]
you can also fill NaN's if you want in the same step:
In [178]: d1['new'] = d1['txt'].str.extract(pat, expand=False).fillna('***')
In [179]: d1
Out[179]:
txt new
0 WXYnineZAB nine
1 EFGsixHIJ ***
2 QRSeightTUV ***
3 GHItwoJKL ***
4 YZAfiveBCD five
.. ... ...
25 QRSeightTUV ***
26 ABConeDEF one
27 STUfourVWX ***
28 CDEtenFGH ***
29 GHItwoJKL ***
[30 rows x 2 columns]
If you want to avoid RegEx, here is a purely list-based solution:
# Sample DataFrames (structure is borrowed from MaxU)
d1 = pd.DataFrame({'txt':['WXYnineZAB','EFGsixHIJ','QRSeightTUV','GHItwoJKL']})
d2 = pd.DataFrame({'word':['two','six']})
# Check if word exists in any txt (1-liner).
exists = [list(d2.word[[word in txt for word in d2.word]])[0] if sum([word in txt for word in d2.word]) == 1 else '***' for txt in d1.txt]
# Resulting output
res = pd.DataFrame(zip(d1.txt,exists), columns = ['text','word'])
Result:
text word
0 WXYnineZAB ***
1 EFGsixHIJ six
2 QRSeightTUV ***
3 GHItwoJKL two
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.