import randomimport matplotlibimport matplotlib.pyplot as pltSAMPLE_SIZE = 1000buckets = 100fig = plt.figure()matplotlib.rcParams.update({"font.size": 7})#第一个图形是在[0,1)之间分布的随机变量(normal distributed random variable)。ax = fig.add_subplot(5,2,1)ax.set_xlabel("random.random")res = [random.random() for _ in xrange(1, SAMPLE_SIZE)]ax.hist(res, buckets)#第二个图形是一个均匀分布的随机变量(uniformly distributed random variable)。ax_2 = fig.add_subplot(5,2,2)ax_2.set_xlabel("random.uniform")a = 1b = SAMPLE_SIZEres_2 = [random.uniform(a, b) for _ in xrange(1, SAMPLE_SIZE)]ax_2.hist(res_2, buckets)#第三个图形是一个三角形分布(triangular distribution)。ax_3 = fig.add_subplot(5,2,3)ax_3.set_xlabel("random.triangular")low = 1high = SAMPLE_SIZEres_3 = [random.uniform(low, high) for _ in xrange(1, SAMPLE_SIZE)]ax_3.hist(res_3, buckets)#第四个图形是一个beta分布(beta distribution)。参数的条件是alpha 和 beta 都要大于0, 返回值在0~1之间。plt.subplot(5,2,4)plt.xlabel("random.betavariate")alpha = 1beta = 10res_4 = [random.betavariate(alpha, beta) for _ in xrange(1, SAMPLE_SIZE)]plt.hist(res_4, buckets)#第五个图形是一个指数分布(exponential distribution)。 lambd 的值是 1.0 除以期望的中值,是一个不为零的数(参数应该叫做lambda没但它是python的一个保留字)。如果lambd是整数,返回值的范围是零到正无穷大;如果lambd为负,返回值的范围是负无穷大到零。plt.subplot(5,2,5)plt.xlabel("random.expovariate")lambd = 1.0/ ((SAMPLE_SIZE + 1) / 2.)res_5 = [random.expovariate(lambd) for _ in xrange(1, SAMPLE_SIZE)]plt.hist(res_5, buckets)#第六个图形是gamma分布(gamma distribution), 要求参数alpha 和beta都大于零。plt.subplot(5,2,6)plt.xlabel("random.gammavariate")alpha = 1beta = 10res_6 = [random.gammavariate(alpha, beta) for _ in xrange(1, SAMPLE_SIZE)]plt.hist(res_6, buckets)#第七个图形是对数正态分布(Log normal distribution)。如果取这个分布的自然对数,会得到一个中值为mu,标准差为sigma的正态分布。mu可以取任何值,sigma必须大于零。plt.subplot(5,2,7)plt.xlabel("random.lognormalvariate")mu = 1sigma = 0.5res_7 = [random.lognormvariate(mu, sigma) for _ in xrange(1, SAMPLE_SIZE)]plt.hist(res_7, buckets)#第八个图形是正态分布(normal distribution)。plt.subplot(5,2,8)plt.xlabel("random.normalvariate")mu = 1sigma = 0.5res_8 = [random.normalvariate(mu, sigma) for _ in xrange(1, SAMPLE_SIZE)]plt.hist(res_8, buckets) #最后一个图形是帕累托分布(Pareto distribution), alpha 是形状参数。plt.subplot(5,2,9)plt.xlabel("random.normalvariate")alpha = 1res_9 = [random.paretovariate(alpha) for _ in xrange(1, SAMPLE_SIZE)]plt.hist(res_9, buckets)plt.show()