拉丁超立方抽样的python代码

下面是一个实现拉丁超立方抽样的Python代码：

import numpy as np

def latin_hypercube(n, k):
    """
    Generate a Latin hypercube sample of size n with k variables.
    """
    # Create an empty array to store the sample
    sample = np.zeros((n, k))

    # Generate the marginal quantiles
    quantiles = np.linspace(0, 1, n + 1)

    # Generate the intervals for each variable
    intervals = np.zeros((k, n + 1))
    for i in range(k):
        intervals[i, :] = np.random.permutation(quantiles)

    # Generate the samples
    for i in range(k):
        for j in range(n):
            sample[j, i] = np.random.uniform(intervals[i, j], intervals[i, j + 1])

    return sample

其中，n是采样数量，k是变量数量。函数返回一个n行k列的采样矩阵。