Longest Strings in List in Dataframe-CodePudding

I want to return the longest string in list in a dataframe. The original data will look like:

items
['pants','hat']
['clothes']
['mouse']

My desired output is:

items               longest_str
['pants','hat']     ['pants']
['clothes']         ['clothes']
['mouse','xx']      ['mouse']

My code is :

def longest(i):
    return max(i,key=len)

for i in range(0:2):
    print(df['items'][i])
    print(longest(df['items'][i]))

but it does not print out what I want. My ultimate goal is to find the longest string in the whole dataframe. Please suggest any other optimal way to approach it

CodePudding user response：

Just call your longest function within df.apply:

In [492]: df['longest_str'] = df['items'].apply(lambda x: [longest(x)])

In [493]: df
Out[493]: 
          items longest_str
0  [pants, hat]     [pants]
1     [clothes]   [clothes]
2       [mouse]     [mouse]

CodePudding user response：

Try:

df = pd.DataFrame({'items':[['pants', 'hat'],['clothes'],['mouse', 'xx']]})

df['longest_str'] = [[max(x, key=len)] for x in df['items']]

df

Output:

          items longest_str
0  [pants, hat]     [pants]
1     [clothes]   [clothes]
2   [mouse, xx]     [mouse]

To finish up your total question:

df = pd.DataFrame({'items':[['pants', 'hat'],['clothes'],['mouse', 'xx']]})

df['longest_str'] = [max(x, key=len) for x in df['items']]

max((x for x in df['longest_str']), key=len)

Output:

'clothes'

Getting really cute:

df['items'].explode().loc[lambda x: x.str.len().max() == x.str.len()].to_numpy()[0]

Output:

'clothes'

CodePudding user response：

Use apply, maybe is better without your function.

df['longest_str'] = df['items'].apply(lambda x: max(x, key=len))