Ranking with multiple columns in Dataframe

Question

I have a dataframe with 3 columns

Alpha Bravo Charlie
20    30    40
50    10    20
40    60    10

I wish to create 3 new columns with rankings that produces the following that gives the highest among the 3 columns a rank of 3 to 1:

AlphaRank BravoRank CharlieRank
1         2         3
3         1         2
2         3         1

I understand there is dataframe.rank function but I only saw example for 1 column not 3

I tried this with issues:

for newrank in ['Alpha', 'Bravo', 'Charlie']:
    ranksys = df[newrank]

    ranksystem = newrank +'Rank'
    df[ranksystem] = ranksys.rank(axis=1).astype(int)

Answer 1

I think need rank + astype:

cols = ['Alpha', 'Bravo', 'Charlie']
df[cols] = df[cols].rank().astype(int)
print (df)
   Alpha  Bravo  Charlie
0      1      2        3
1      3      1        2
2      2      3        1

Numpy alternative with numpy.argsort:

df[cols] = pd.DataFrame(df[cols].values.argsort(axis=0) + 1,index=df.index,columns=df.columns)