Cfrgtkky

I am trying to convert 4 colunms from my DataFrame to a unique column.

I have the following DataFrame:

    doggo   floofer pupper  puppo

0   None    None    None    None

1   None    None    None    None

2   None    None    None    None

3   None    None    None    None

4   None    None    None    None

5   None    None    None    None

6   None    None    None    None

7   None    None    None    None

8   None    None    None    None

9   doggo   None    None    None

10  None    None    None    None

11  None    None    None    None

12  None    None    None    puppo

13  None    None    None    None

14  None    None    None    puppo

I want a unique column filled with the values 'None', 'doggo', 'floofer', 'pupper', 'puppo'.

I have tried using the Melt function with no success.

my actual code:

melt = pd.melt(melt, id_vars=['doggo', 'floofer', 'pupper', 'puppo'], var_name='classification')

Any help?

EDIT

Below the complete solution(comments are in portuguese):

#criar uma cópia do DataFrame para não comprometer o DataFrame original

df = twitter_archive.copy()



#Apagar os valores None 

df = df.replace('None', '')



#criar e preencher a coluna classification com as informações das colunas doggo, floofer, pupper e puppo

df['classification'] = (df['doggo'].fillna('') + df['floofer'].fillna('') + df['pupper'].fillna('') + df['puppo'].fillna('')).replace('', np.nan)



#Dropar todas as colunas e deixar somente a classification

df = df.drop(columns=['in_reply_to_status_id', 'in_reply_to_user_id', 'timestamp', 'source', 'text', 'retweeted_status_id', 'retweeted_status_user_id', 'retweeted_status_timestamp', 'expanded_urls', 'rating_numerator', 'rating_denominator', 'name', 'doggo','floofer', 'pupper', 'puppo'])



#Acrescentar a coluna classification no DataFrame twitter_archive e remover as colunas doggo, floofer, pupper e puppo

twitter_archive = pd.merge(twitter_archive, df, on= 'tweet_id')

twitter_archive = twitter_archive.drop(columns=['doggo', 'floofer', 'pupper', 'puppo'])

A quick and dirty way to do it :

df['classification'] = (df['doggo'].fillna('') + df['floofer'].fillna('') + df['pupper'].fillna('') + df['puppo'].fillna('')).replace('', np.nan)

just make take the maximum. Every string is larger than None. Assuming your entries(per line) are unique. The following should work

d = {"col1": [None, "x", None], "col2": ["y",None, None]}

x = pd.DataFrame(d)

x["col3"] = x[["col1", "col2"]].max(axis=1)

Output:

   col1  col2  col3

0  None     y     y

1     x  None     x

2  None  None  None