python-3.x - 按年重设的累计金额

Question

我有一个熊猫df：

pd.DataFrame({'team': ['ARI','ARI','ARI','BAL','BAL','BAL','BUF','BUF','BUF'],
          'year': [2019, 2019, 2020, 2019, 2019, 2020, 2019, 2019, 2020],
          'week': [16,17,1,16,17,1,16,17,1],
          'hits': [1, 3, 2, 2, 2, 5, 3, 3, 2]})

我想将“命中”列覆盖为累积和，但累积和必须每“年”重置一次。预期输出：

pd.DataFrame({'team': ['ARI','ARI','ARI','BAL','BAL','BAL','BUF','BUF','BUF'],
          'year': [2019, 2019, 2020, 2019, 2019, 2020, 2019, 2019, 2020],
          'week': [16,17,1,16,17,1,16,17,1],
          'cumsum_hits': [1,4,2,2,4,5,3,6,2]})

标签： python-3.xpandascumulative-sum