CODEWITHSUNDEEP
pythonpandasdataframe
Erdnase
Erdnase

Reputation: 790

How to check whether list inside a Pandas data frame is empty

I have a pandas data frame 

list1 = ['A','B', 'C', 'D' ]
list2 = ['P','Q', 'R', 'S' ]
list3 = ['A', 'B', 'C', 'D']
list4 = [[1,3,5], [], [2,4,1,5], []]

df = pd.DataFrame({'col1' : list1,
 'col2' : list2,
 'col3':list3,
 'numlist' : list4

})
print(df)     

col1 col2 col3       numlist
0    A    P    A     [1, 3, 5]
1    B    Q    B            []
2    C    R    C  [2, 4, 1, 5]
3    D    S    D            []

I want to get the index of all rows where the column numlist is not empty and then create a new data frame with rows matching the indexes. I am trying the following code

df[(len(df['numlist']))==0]

But this code is throwing Key Error. How can I achieve the same?

Upvotes: 0

Views: 136

Answers (2)

Dishin H Goyani
Dishin H Goyani

Reputation: 7693

You can use apply and check with x != [] condition that list is empty or not.

>>> df = df[df.numlist.apply(lambda x: x != [])]
>>> df
  col1 col2 col3       numlist
0    A    P    A     [1, 3, 5]
2    C    R    C  [2, 4, 1, 5]

Upvotes: 1

jezrael
jezrael

Reputation: 862641

Empty lists converted to boolen return False is possible filtering this way:

df1 = df[df['numlist'].astype(bool)]

Your solution should be changed by Series.str.len with Series.ne for not equal:

df1 = df[df['numlist'].str.len().ne(0)]

Alternative:

df1 = df[df['numlist'].str.len() != 0]
print(df1)  
  col1 col2 col3       numlist
0    A    P    A     [1, 3, 5]
2    C    R    C  [2, 4, 1, 5]

Upvotes: 3

Related Questions