How to merge the values coming from 4 different columns into one fina column-CodePudding

I have a daframe like the one below:

the columns str, mpr, cta, mpt are all the same but for every record I have value in only 1 of these 4 columns.

How can I merge these 4 columns in only 1 like the output below:

id     str     mpr     cta      mpt
 1     10      null    null     null
 2    null     11      null     null
 3    null     null    6        null
 4    null     null    null     1

Output:

as you can see the columns final will always have a value coming from only one of each 4 columns.

Important: sometimes the 4 columns is null then the final column will also be null

CodePudding user response：

# find the max along the row, starting from second columns (skipping over ID)
df['final']=df.iloc[:,1:].max(axis=1, numeric_only=True) #.astype(int)
df[['id','final']]

# if a row can have all Null than don't add astype(int)

   id   final
0   1    10
1   2    11
2   3     6
3   4     1

CodePudding user response：

Lets call your original DataFrame df and lets say you plan on scaling it on a million columns.

Create a list for the final column:

Final_List = []

Create a list of colums

 my_list = df.columns.values.tolist()
    
    for x in range(len(my_list)):
        for y in range(len(df[my_list[x]]):
            if df[my_list[x][y] != 'null'
               final_value = df[my_list[x][y]]
               Final_List.append(final_value) 
               break #breaks the inner loop
            else:
                 final_value = 'nul'
            Final_List.append(final_value)
    
    df["final"] = Final_List