make each 2 rows in one row with pandas-CodePudding

player    probability_of_win
player1    35%
player2    65%
player3    20%
player4    80%
player5    90%
player6    10%

every 2 rows represent a match of tennis, I want to make a data frame with 1 row per match. like this:

playerA    playerB    probability_of_win_A    probability_of_win_B
player1    player2    35%                     65%
player3    player4    20%                     80%
player5    player6    90%                     10%

CodePudding user response：

Create new columns by integer and modulo 2 in DataFrame.assign, pass to DataFrame.pivot with rename, last flatten MultiIndex in columns:

df = (df.assign(A = np.arange(len(df.index)) // 2,
                B = np.arange(len(df.index)) % 2)
        .pivot(index='A', columns='B')
        .rename(columns={0:'A', 1:'B'}))
df.columns = df.columns.map(lambda x: f'{x[0]}_{x[1]}')
print (df)
  player_A player_B probability_of_win_A probability_of_win_B
A                                                            
0  player1  player2                  35%                  65%
1  player3  player4                  20%                  80%
2  player5  player6                  90%                  10%