TypeError: '<' not supported between instances of 'int' and 'list' in numpy intersect1d

Question

I want to apply intersect between two list,phone & phone_office.So I wrote following code

phone=df_final_naFill.iloc[0,4]
type(phone) # List
s=pd.Series(phone)
type(s) #pandas.core.series.Series
a=pd.Series(s.apply(pd.Series).stack().astype(int).groupby(level=0).apply(list))
phone_office=df_final_naFill.iloc[0,6]
# type(phone_office) #List
h=pd.Series(phone_office)
phone_comb=np.intersect1d(a,h)

But after running the code, I'm getting following error message

  File "<ipython-input-206-3512341621de>", line 1, in <module>
    phone_comb=np.intersect1d(a,h)

  File "C:\ProgramData\Anaconda3\lib\site-packages\numpy\lib\arraysetops.py", line 337, in intersect1d
    aux.sort()

TypeError: '<' not supported between instances of 'int' and 'list'

The first 5 rows of phone looks like

0    [735015372, 72151508105, 7217511580, 721150431...
1                                                   []
2    [735152771, 7351515043, 7115380870, 7115427...
3    [7111332015, 73140214, 737443075, 7110815115...
4    [718218718, 718221342, 73551401, 71811507...

my h looks like

and a looks like

although type of a & h are showing as pandas.core.series.Series. I'm guessing this type of format for a is coming in some iterations.Most of the time my code is running without any error.

Can you suggest me what changes I need to do in python 3.x?

Could you show how a and h look like and what is the expected output? — Mr Tarsa
– Mr Tarsa, Commented Jan 19, 2018 at 22:36
I have given the base data from which a & h came.Expected outcome is intersection between these two.Any suggestions please? — Tanvi Mirza
– Tanvi Mirza, Commented Jan 20, 2018 at 5:07
Thanks @jezrael, it works. Yes my data is problematic.In some iterations it works with a=pd.Series(s.apply(pd.Series).stack().astype(int).groupby(l‌evel=0).apply(list)) & some iteration it doesn't. Will your solution work if my a will have structure like h? — Tanvi Mirza
– Tanvi Mirza, Commented Jan 20, 2018 at 19:25

jezrael · Accepted Answer · 2018-01-20 07:46:32Z

2

If both lists are not nested you can simplify your code a lot:

phone_comb=np.intersect1d(phone, phone_office)

And if need compare each row in original DataFrame:

f = lambda x: np.intersect1d(x.iat[4], x.iat[6])
df_final_naFill['inter'] = df_final_naFill.apply(f, axis=1).values.tolist()

edited Jan 20, 2018 at 7:46

answered Jan 20, 2018 at 7:35

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

12 Comments

cs95 Over a year ago

The more upvotes you get, the more envy, and so more downvotes... it becomes a problem when they downvote 0 answers so it actually looks very bad (-1).

Tanvi Mirza Over a year ago

@jezrael, Previously you made some changes in my a=pd.Series(s.apply(pd.Series).stack().astype(int).groupby(level=0).apply(list)) and added it in comment section.I'm not seeing that one. Can you please add it?

Tanvi Mirza Over a year ago

@jezrael,can you please suggest some alternative options of my a=pd.Series(s.apply(pd.Series).stack().astype(int).groupby(level=0).apply(list)). Actually my data is problematic.Using your phone_comb=np.intersect1d(phone, phone_office) I'm not getting the desired output.

Tanvi Mirza Over a year ago

Can you still share that code @jezrael which you think was wrong.I think it was working for me

jezrael Over a year ago

@TanviMirza - there is problem I forget it ;) But data are confidental? Because if it does not working I think some data problem.

|

Collectives™ on Stack Overflow

TypeError: '<' not supported between instances of 'int' and 'list' in numpy intersect1d

1 Answer 1

12 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

12 Comments

Your Answer

Sign up or log in

Post as a guest

Related