每个键的字典调用函数

Question

我有下面的代码，其中我试图根据分布的类型和所述分布的参数构建一个随机数对象。该代码的工作原理是生成一个包含10000个均匀分布点的对象

import numpy as np

rngSeed = 654  # Random seed for reproducibility
numSimulations = 10_000


def get_rand_unif(min_value, max_value, n_samples=1_000):
    """
    This function generates a random number [nd]array of size [n_samples]
    from a uniform distribution between the two input values
    [min_value, max_value)
    Args:
    - min_value (float)
    - max_value (float)
    - n_samples (int)
    Return:
    - Random number [nd]array of [n_samples] between this range (float)
    """
    return np.random.default_rng(rngSeed).uniform(min_value, max_value, n_samples)


def get_rand_norm(mean, std_dev, n_samples=1_000):
    """
    This function generates a random number [nd]array of size [n_samples]
    from a normal distribution with a mean of [mean] and a standard deviation
    of [std_dev].
    Args:
    - mean(float)
    - std_dev (float)
    - n_samples (int)
    Return:
    - Random number [nd]array of [n_samples] from a normal distribution (float)
    """
    return np.random.default_rng(rngSeed).normal(mean, std_dev, n_samples)


class InVar:
    def __init__(self, *parms):
        self.rndData = self.get_disto(*parms)

    def get_disto(self, *parms):
        distR = {
            "unif": get_rand_unif(parms[1], parms[2], numSimulations),
            "norm": get_rand_norm(parms[1], parms[2], numSimulations),
        }
        return distR.get(parms[0])

    def rand_data(self):
        return self.rndData

p1 = InVar("unif", 0, 1)

然而，当我调试代码时，我注意到正在为均匀和正态分布计算随机值。我想知道如何更改代码，以便只执行与被调用键对应的函数。虽然只是一个轶事，但我担心如果我引入需要两个以上参数的额外概率分布，我会遇到麻烦

Answer 1

我认为您正在寻找的简单修复方法是在构建dict之后调用函数

def get_disto(self, *parms):
    distR = {
        "unif": get_rand_unif,
        "norm": get_rand_norm,
    }
    return distR.get(parms[0])(parms[1], parms[2], numSimulations)

更新：如果要像以前一样单独调用函数，请使用If/else构造

def get_disto(self, *parms):
    if parms[0] == "unif":
        return get_rand_unif(parms[1], parms[2], numSimulations)
    elif parms[0] == "norm":
        return get_rand_norm(parms[1], parms[2], numSimulations)
    else:
        raise Exception(f"Unrecognized distribution type: {parms[0]}")