在groupby之后进行前滚求和的更快方法？

import pandas as pd

# 创建示例数据
data = {
    'group': ['A', 'A', 'B', 'B', 'A', 'B'],
    'value': [1, 2, 3, 4, 5, 6]
}
df = pd.DataFrame(data)

# 使用 groupby 和 cumsum 进行前滚求和
result = df.groupby('group')['value'].cumsum()

print(result)

df['cumulative_sum'] = df.groupby('group')['value'].transform(lambda x: x.cumsum())
print(df)

import pandas as pd
from numba import njit

# 创建示例数据
data = {
    'group': ['A', 'A', 'B', 'B', 'A', 'B'],
    'value': [1, 2, 3, 4, 5, 6]
}
df = pd.DataFrame(data)

@njit
def cumsum_numba(arr):
    result = []
    current_sum = 0
    for x in arr:
        current_sum += x
        result.append(current_sum)
    return result

df['cumulative_sum'] = df.groupby('group')['value'].apply(lambda x: cumsum_numba(x.values))
print(df)