TypeError: unsupported operand type(s) for |: 'str' and 'bool'

问题

I have dataframe users with different columns. My goal is to add the column [uses_name] which should be True when a password is the same as each users first or last name.

For example, [user_name] in twelve row contain milford.hubbard. Then in [uses_name] will be True, because the [password] and [last_name] are the same.

To do this, I create two columns [first_name] and [last_name] with regular expressions. When create [uses_name] I have trouble with | operator. I am read more in pandas doc about Boolean indexing but not find an answer.

My code:

import pandas as pd

users = pd.read_csv('datasets/users.csv')

# Extracting first and last names into their own columns

users['first_name'] = users['user_name'].str.extract(r'(^\w+)', expand=False)

users['last_name'] = users['user_name'].str.extract(r'(\w+$)', expand=False)

# Flagging the users with passwords that matches their names

users['uses_name'] = users['password'].isin(users['first_name'] | users['last_name'])

# Counting and printing the number of users using names as passwords

print(users['uses_name'].count())

# Taking a look at the 12 first rows

print(users.head(12))

When I try to compile this, I give an error:

TypeError: unsupported operand type(s) for |: 'str' and 'bool'

First 12 rows in users dataframe with created first_name and last_name columns:

id          user_name            password   first_name  last_name
0    1    vance.jennings          joobheco      vance    jennings
1    2    consuelo.eaton        0869347314   consuelo       eaton
2    3   mitchel.perkins        fabypotter    mitchel     perkins
3    4    odessa.vaughan         aharney88     odessa     vaughan
2    3   mitchel.perkins        fabypotter    mitchel     perkins
3    4    odessa.vaughan         aharney88     odessa     vaughan
4    5    araceli.wilder        acecdn3000    araceli      wilder
5    6  shawn.harrington           5278049      shawn  harrington
6    7        evelyn.gay            master     evelyn         gay
7    8       noreen.hale            murphy     noreen        hale
8    9       gladys.ward           lwsves2     gladys        ward
9   10   brant.zimmerman  1190KAREN5572497      brant   zimmerman
10  11     leanna.abbott          aivlys24     leanna      abbott
11  12   milford.hubbard           hubbard    milford     hubbard

回答1:

This works:

users['uses_name']= (users['password']==users['first_name'] )| (users['password']==users['last_name'])

回答2:

You can concat , since both of then are Series

users['password'].isin(pd.concat([users['first_name'],users['last_name']]))

Since you change the question , Update one

df[['first_name','last_name']].eq(df.password,axis=0).any(1)

回答3:

Use numpy.union1d:

val = np.union1d(users['first_name'], users['last_name'])
users['uses_name'] = users['password'].isin(val)
print (users)
    id         user_name          password first_name   last_name  uses_name
0    1    vance.jennings          joobheco      vance    jennings      False
1    2    consuelo.eaton        0869347314   consuelo       eaton      False
2    3   mitchel.perkins        fabypotter    mitchel     perkins      False
3    4    odessa.vaughan         aharney88     odessa     vaughan      False
2    3   mitchel.perkins        fabypotter    mitchel     perkins      False
3    4    odessa.vaughan         aharney88     odessa     vaughan      False
4    5    araceli.wilder        acecdn3000    araceli      wilder      False
5    6  shawn.harrington           5278049      shawn  harrington      False
6    7        evelyn.gay            master     evelyn         gay      False
7    8       noreen.hale            murphy     noreen        hale      False
8    9       gladys.ward           lwsves2     gladys        ward      False
9   10   brant.zimmerman  1190KAREN5572497      brant   zimmerman      False
10  11     leanna.abbott          aivlys24     leanna      abbott      False
11  12   milford.hubbard           hubbard    milford     hubbard       True

回答4:

I think the best would be to perform a set union and pass that to isin:

users['uses_name'] = users['password'].isin(
   set(users['first_name']).union(users['last_name'])
)

users 

    id         user_name          password first_name   last_name  uses_name
0    1    vance.jennings          joobheco      vance    jennings      False
1    2    consuelo.eaton        0869347314   consuelo       eaton      False
2    3   mitchel.perkins        fabypotter    mitchel     perkins      False
3    4    odessa.vaughan         aharney88     odessa     vaughan      False
2    3   mitchel.perkins        fabypotter    mitchel     perkins      False
3    4    odessa.vaughan         aharney88     odessa     vaughan      False
4    5    araceli.wilder        acecdn3000    araceli      wilder      False
5    6  shawn.harrington           5278049      shawn  harrington      False
6    7        evelyn.gay            master     evelyn         gay      False
7    8       noreen.hale            murphy     noreen        hale      False
8    9       gladys.ward           lwsves2     gladys        ward      False
9   10   brant.zimmerman  1190KAREN5572497      brant   zimmerman      False
10  11     leanna.abbott          aivlys24     leanna      abbott      False
11  12   milford.hubbard           hubbard    milford     hubbard       True

Note that | is the logical OR, it has no meaning for string columns in pandas.

来源：https://stackoverflow.com/questions/49364654/typeerror-unsupported-operand-types-for-str-and-bool

标签

python

pandas

dataframe

boolean