在Python矩阵中，考虑到选择列的成本，如何选择每行一次并获得最低总和？

def min_cost(matrix):
    rows = len(matrix)
    cols = len(matrix[0])

    # 创建dp数组并初始化第1行
    dp = [[0] * cols for _ in range(rows)]
    dp[0] = matrix[0]

    # 动态规划，计算最低总和
    for i in range(1, rows):
        for j in range(cols):
            dp[i][j] = matrix[i][j] + min(dp[i-1][:j] + dp[i-1][j+1:])

    # 获取最低总和
    min_sum = min(dp[rows-1])

    # 回溯，选择每行的列
    selected_cols = []
    min_index = dp[rows-1].index(min_sum)
    selected_cols.append(min_index)

    for i in range(rows-2, -1, -1):
        for j in range(cols):
            if j != min_index and dp[i][j] == dp[i+1][min_index] - matrix[i+1][min_index]:
                selected_cols.append(j)
                min_index = j
                break

    selected_cols.reverse()
    return min_sum, selected_cols

matrix = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
min_sum, selected_cols = min_cost(matrix)
print("最低总和：", min_sum)
print("选择的列：", selected_cols)

最低总和： 12
选择的列： [0, 1, 0]