如何反转 pandas 字符串列的顺序

Question

我有一个类似于以下内容的网球数据集：

import pandas as pd

player_name = ['Novak Djokovic','Rafael Nadal','Roger Federer','Andy Murray']
match_id = [202, 202, 203, 203]
score = ['6-3 5-7 6-4', '6-3 5-7 6-4', '6-4 7-6', '6-4 7-6']
outcome = [1,0,1,0]

df = pd.DataFrame(
    {'player_name': player_name,
        'match_id': match_id,
         'score': score,
         'outcome':outcome
    })

df["sets"]= df["score"].str.split(" ", n = 3, expand = False)
df['set1_score'] = df['sets'].str[0]
df['set2_score'] = df['sets'].str[1]
df['set3_score'] = df['sets'].str[2]

结果变量的值为 1 表示获胜，值为 0 表示失败。目前，分数是从获胜者的角度书写的。我希望能够切换以 0 值为条件的 set1_score、set2_score、set3_score 列的顺序（即：对于比赛的失败者）。

所以，我希望数据框看起来像这样：

player_name    match_id    score         outcome    sets            set1_score    set2_score   set3_score
Novak Djokovic 202         6-3 5-7 6-4   1          [6-3, 5-7, 6-4] 6-3             5-7          6-4
Rafel Nadal    202         6-3 5-7 6-4   0          [6-3, 5-7, 6-4] 3-6             7-5          4-6
Roger Federer  203         6-4 7-6       1          [6-4, 7-6]      6-4             7-6          NaN
Andy Murray    203         6-4 7-6       0          [6-4, 7-6]      4-6             6-7          NaN

我应该如何进行这个过程？

谢谢！

Answer 1

要根据您所期望的结果反转 pandas DataFrame 中的分数顺序，您可以定义一个函数，在结果为 0（损失）时反转每个分数。然后可以使用 apply 方法和 axis=1 将该函数应用于每一行，以对行而不是列进行操作。尝试使用下面的代码

import pandas as pd

# Your original DataFrame
player_name = ['Novak Djokovic', 'Rafael Nadal', 'Roger Federer', 'Andy Murray']
match_id = [202, 202, 203, 203]
score = ['6-3 5-7 6-4', '6-3 5-7 6-4', '6-4 7-6', '6-4 7-6']
outcome = [1, 0, 1, 0]

df = pd.DataFrame({
    'player_name': player_name,
    'match_id': match_id,
    'score': score,
    'outcome': outcome
})

# Function to reverse scores
def reverse_score(row):
    if row['outcome'] == 0:
        new_scores = []
        for set_score in row['score'].split():
            a, b = set_score.split('-')
            new_scores.append(f"{b}-{a}")
        return ' '.join(new_scores)
    else:
        return row['score']

# Apply the function to each row
df['score'] = df.apply(reverse_score, axis=1)

# Split the scores into individual sets
df['sets'] = df['score'].str.split(' ')
df['set1_score'] = df['sets'].str[0]
df['set2_score'] = df['sets'].str[1]
df['set3_score'] = df['sets'].str[2]

print(df)