0.导语

Scipy是一个用于数学、科学、工程领域的常用软件包,可以处理插值、积分、优化、图像处理、常微分方程数值解的求解、信号处理等问题。它用于有效计算Numpy矩阵,使Numpy和Scipy协同工作,高效解决问题。

Scipy是由针对特定任务的子模块组成:

模块名应用领域
scipy.cluster向量计算/Kmeans
scipy.constants物理和数学常量
scipy.fftpack傅立叶变换
scipy.integrate积分程序
scipy.interpolate插值
scipy.io数据输入输出
scipy.linalg线性代数程序
scipy.ndimagen维图像包
scipy.odr正交距离回归
scipy.optimize优化
scipy.signal信号处理
scipy.sparse稀疏矩阵
scipy.spatial空间数据结构和算法
scipy.special一些特殊的数学函数
scipy.stats统计

备注:本文代码可以在github下载

https://github.com/fengdu78/Data-Science-Notes/tree/master/4.scipy

1.SciPy-数值计算库

import numpy as npimport pylab as pl
import matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']
import scipyscipy.__version__#查看版本
'1.0.0'

常数和特殊函数

from scipy import constants as Cprint (C.c) # 真空中的光速print (C.h) # 普朗克常数
299792458.06.62607004e-34
C.physical_constants["electron mass"]
(9.10938356e-31, 'kg', 1.1e-38)
# 1英里等于多少米, 1英寸等于多少米, 1克等于多少千克, 1磅等于多少千克print(C.mile)print(C.inch)print(C.gram)print(C.pound)
1609.34399999999980.02540.0010.45359236999999997
import scipy.special as S
print (1 + 1e-20)print (np.log(1+1e-20))print (S.log1p(1e-20))
1.00.01e-20
m = np.linspace(0.1, 0.9, 4)u = np.linspace(-10, 10, 200)results = S.ellipj(u[:, None], m[None, :])print([y.shape for y in results])
[(200, 4), (200, 4), (200, 4), (200, 4)]
#%figonly=使用广播计算得到的`ellipj()`返回值fig, axes = pl.subplots(2, 2, figsize=(12, 4))labels = ["$sn$", "$cn$", "$dn$", "$\phi$"]for ax, y, label in zip(axes.ravel(), results, labels):ax.plot(u, y)ax.set_ylabel(label)ax.margins(0, 0.1)axes[1, 1].legend(["$m={:g}$".format(m_) for m_ in m], loc="best", ncol=2);

2.拟合与优化-optimize

非线性方程组求解

import pylab as plimport numpy as np
import matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']
from math import sin, cosfrom scipy import optimizedef f(x): #❶x0, x1, x2 = x.tolist() #❷return [5*x1+3,4*x0*x0 - 2*sin(x1*x2),x1*x2 - 1.5]# f计算方程组的误差,[1,1,1]是未知数的初始值result = optimize.fsolve(f, [1,1,1]) #❸print (result)print (f(result))
[-0.70622057 -0.6-2.5 ][0.0, -9.126033262418787e-14, 5.329070518200751e-15]
def j(x):#❶x0, x1, x2 = x.tolist()return [[0, 5, 0],[8 * x0, -2 * x2 * cos(x1 * x2), -2 * x1 * cos(x1 * x2)],[0, x2, x1]]result = optimize.fsolve(f, [1, 1, 1], fprime=j)#❷print(result)print(f(result))
[-0.70622057 -0.6-2.5 ][0.0, -9.126033262418787e-14, 5.329070518200751e-15]

最小二乘拟合

import numpy as npfrom scipy import optimizeX = np.array([ 8.19,2.72,6.39,8.71,4.7 ,2.66,3.78])Y = np.array([ 7.01,2.78,6.47,6.71,4.1 ,4.23,4.05])def residuals(p): #❶"计算以p为参数的直线和原始数据之间的误差"k, b = preturn Y - (k*X + b)# leastsq使得residuals()的输出数组的平方和最小,参数的初始值为[1,0]r = optimize.leastsq(residuals, [1, 0]) #❷k, b = r[0]print ("k =",k, "b =",b)
k = 0.6134953491930442 b = 1.794092543259387
#%figonly=最小化正方形面积之和(左),误差曲面(右)scale_k = 1.0scale_b = 10.0scale_error = 1000.0def S(k, b):"计算直线y=k*x+b和原始数据X、Y的误差的平方和"error = np.zeros(k.shape)for x, y in zip(X, Y):error += (y - (k * x + b)) ** 2return errorks, bs = np.mgrid[k - scale_k:k + scale_k:40j, b - scale_b:b + scale_b:40j]error = S(ks, bs) / scale_errorfrom mpl_toolkits.mplot3d import Axes3Dfrom matplotlib.patches import Rectanglefig = pl.figure(figsize=(12, 5))ax1 = pl.subplot(121)ax1.plot(X, Y, "o")X0 = np.linspace(2, 10, 3)Y0 = k*X0 + bax1.plot(X0, Y0)for x, y in zip(X, Y):y2 = k*x+brect = Rectangle((x,y), abs(y-y2), y2-y, facecolor="red", alpha=0.2)ax1.add_patch(rect)ax1.set_aspect("equal")ax2 = fig.add_subplot(122, projection='3d')ax2.plot_surface(ks, bs / scale_b, error, rstride=3, cstride=3, cmap="jet", alpha=0.5)ax2.scatter([k], [b / scale_b], [S(k, b) / scale_error], c="r", s=20)ax2.set_xlabel("$k$")ax2.set_ylabel("$b$")ax2.set_zlabel("$error$");

#%fig=带噪声的正弦波拟合def func(x, p):#❶"""数据拟合所用的函数: A*sin(2*pi*k*x + theta)"""A, k, theta = preturn A * np.sin(2 * np.pi * k * x + theta)def residuals(p, y, x):#❷"""实验数据x, y和拟合函数之间的差,p为拟合需要找到的系数"""return y - func(x, p)x = np.linspace(0, 2 * np.pi, 100)A, k, theta = 10, 0.34, np.pi / 6# 真实数据的函数参数y0 = func(x, [A, k, theta])# 真实数据# 加入噪声之后的实验数据np.random.seed(0)y1 = y0 + 2 * np.random.randn(len(x))#❸p0 = [7, 0.40, 0]# 第一次猜测的函数拟合参数# 调用leastsq进行数据拟合# residuals为计算误差的函数# p0为拟合参数的初始值# args为需要拟合的实验数据plsq = optimize.leastsq(residuals, p0, args=(y1, x))#❹print(u"真实参数:", [A, k, theta])print(u"拟合参数", plsq[0])# 实验数据拟合后的参数pl.plot(x, y1, "o", label=u"带噪声的实验数据")pl.plot(x, y0, label=u"真实数据")pl.plot(x, func(x, plsq[0]), label=u"拟合数据")pl.legend(loc="best")
真实参数: [10, 0.34, 0.5235987755982988]拟合参数 [10.252187480.3423992 0.50817423]

def func2(x, A, k, theta):return A*np.sin(2*np.pi*k*x+theta)popt, _ = optimize.curve_fit(func2, x, y1, p0=p0)print (popt)
[10.252187480.3423992 0.50817425]
popt, _ = optimize.curve_fit(func2, x, y1, p0=[10, 1, 0])print(u"真实参数:", [A, k, theta])print(u"拟合参数", popt)
真实参数: [10, 0.34, 0.5235987755982988]拟合参数 [ 0.710934691.02074585 -0.12776742]

计算函数局域最小值

def target_function(x, y):return (1 - x)**2 + 100 * (y - x**2)**2class TargetFunction(object):def __init__(self):self.f_points = []self.fprime_points = []self.fhess_points = []def f(self, p):x, y = p.tolist()z = target_function(x, y)self.f_points.append((x, y))return zdef fprime(self, p):x, y = p.tolist()self.fprime_points.append((x, y))dx = -2 + 2 * x - 400 * x * (y - x**2)dy = 200 * y - 200 * x**2return np.array([dx, dy])def fhess(self, p):x, y = p.tolist()self.fhess_points.append((x, y))return np.array([[2 * (600 * x**2 - 200 * y + 1), -400 * x], [-400 * x, 200]])def fmin_demo(method):target = TargetFunction()init_point = (-1, -1)res = optimize.minimize(target.f,init_point,method=method,jac=target.fprime,hess=target.fhess)return res, [np.array(points) for points in (target.f_points, target.fprime_points,target.fhess_points)]methods = ("Nelder-Mead", "Powell", "CG", "BFGS", "Newton-CG", "L-BFGS-B")for method in methods:res, (f_points, fprime_points, fhess_points) = fmin_demo(method)print("{:12s}: min={:12g}, f count={:3d}, fprime count={:3d}, fhess count={:3d}".format(method, float(res["fun"]), len(f_points), len(fprime_points),len(fhess_points)))
Nelder-Mead : min= 5.30934e-10, f count=125, fprime count=0, fhess count=0Powell: min= 0, f count= 52, fprime count=0, fhess count=0CG: min= 9.63056e-21, f count= 39, fprime count= 39, fhess count=0BFGS: min= 1.84992e-16, f count= 40, fprime count= 40, fhess count=0Newton-CG : min= 5.22666e-10, f count= 60, fprime count= 97, fhess count= 38L-BFGS-B: min=6.5215e-15, f count= 33, fprime count= 33, fhess count=0
#%figonly=各种优化算法的搜索路径def draw_fmin_demo(f_points, fprime_points, ax):xmin, xmax = -3, 3ymin, ymax = -3, 3Y, X = np.ogrid[ymin:ymax:500j,xmin:xmax:500j]Z = np.log10(target_function(X, Y))zmin, zmax = np.min(Z), np.max(Z)ax.imshow(Z, extent=(xmin,xmax,ymin,ymax), origin="bottom", aspect="auto", cmap="gray")ax.plot(f_points[:,0], f_points[:,1], lw=1)ax.scatter(f_points[:,0], f_points[:,1], c=range(len(f_points)), s=50, linewidths=0)if len(fprime_points):ax.scatter(fprime_points[:, 0], fprime_points[:, 1], marker="x", color="w", alpha=0.5)ax.set_xlim(xmin, xmax)ax.set_ylim(ymin, ymax)fig, axes = pl.subplots(2, 3, figsize=(9, 6))methods = ("Nelder-Mead", "Powell", "CG", "BFGS", "Newton-CG", "L-BFGS-B")for ax, method in zip(axes.ravel(), methods):res, (f_points, fprime_points, fhess_points) = fmin_demo(method)draw_fmin_demo(f_points, fprime_points, ax)ax.set_aspect("equal")ax.set_title(method)

计算全域最小值

def func(x, p):A, k, theta = preturn A*np.sin(2*np.pi*k*x+theta)def func_error(p, y, x):return np.sum((y - func(x, p))**2)x = np.linspace(0, 2*np.pi, 100)A, k, theta = 10, 0.34, np.pi/6y0 = func(x, [A, k, theta])np.random.seed(0)y1 = y0 + 2 * np.random.randn(len(x))
result = optimize.basinhopping(func_error, (1, 1, 1),niter = 10,minimizer_kwargs={"method":"L-BFGS-B","args":(y1, x)})print (result.x)
[10.25218676 -0.342399092.63341581]
#%figonly=用`basinhopping()`拟合正弦波pl.plot(x, y1, "o", label=u"带噪声的实验数据")pl.plot(x, y0, label=u"真实数据")pl.plot(x, func(x, result.x), label=u"拟合数据")pl.legend(loc="best");

3.线性代数-linalg

解线性方程组

import pylab as plimport numpy as npfrom scipy import linalg
import matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']
import numpy as npfrom scipy import linalgm, n = 500, 50A = np.random.rand(m, m)B = np.random.rand(m, n)X1 = linalg.solve(A, B)X2 = np.dot(linalg.inv(A), B)print (np.allclose(X1, X2))%timeit linalg.solve(A, B)%timeit np.dot(linalg.inv(A), B)
True5.38 ms ± 120 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)8.14 ms ± 994 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)
luf = linalg.lu_factor(A)X3 = linalg.lu_solve(luf, B)np.allclose(X1, X3)
True
M, N = 1000, 100np.random.seed(0)A = np.random.rand(M, M)B = np.random.rand(M, N)Ai = linalg.inv(A)luf = linalg.lu_factor(A)%timeit linalg.inv(A)%timeit np.dot(Ai, B)%timeit linalg.lu_factor(A)%timeit linalg.lu_solve(luf, B)
50.6 ms ± 1.94 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)3.49 ms ± 306 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)20.1 ms ± 1.42 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)4.49 ms ± 65 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

最小二乘解

from numpy.lib.stride_tricks import as_strideddef make_data(m, n, noise_scale):#❶np.random.seed(42)x = np.random.standard_normal(m)h = np.random.standard_normal(n)y = np.convolve(x, h)yn = y + np.random.standard_normal(len(y)) * noise_scale * np.max(y)return x, yn, hdef solve_h(x, y, n):#❷X = as_strided(x, shape=(len(x) - n + 1, n), strides=(x.itemsize, x.itemsize))#❸Y = y[n - 1:len(x)]#❹h = linalg.lstsq(X, Y)#❺return h[0][::-1]#❻
x, yn, h = make_data(1000, 100, 0.4)H1 = solve_h(x, yn, 120)H2 = solve_h(x, yn, 80)print("Average error of H1:", np.mean(np.abs(h[:100] - h)))print("Average error of H2:", np.mean(np.abs(h[:80] - H2)))
Average error of H1: 0.0Average error of H2: 0.2958422158342371
#%figonly=实际的系统参数与最小二乘解的比较fig, (ax1, ax2) = pl.subplots(2, 1, figsize=(6, 4))ax1.plot(h, linewidth=2, label=u"实际的系统参数")ax1.plot(H1, linewidth=2, label=u"最小二乘解H1", alpha=0.7)ax1.legend(loc="best", ncol=2)ax1.set_xlim(0, len(H1))ax2.plot(h, linewidth=2, label=u"实际的系统参数")ax2.plot(H2, linewidth=2, label=u"最小二乘解H2", alpha=0.7)ax2.legend(loc="best", ncol=2)ax2.set_xlim(0, len(H1));

特征值和特征向量

A = np.array([[1, -0.3], [-0.1, 0.9]])evalues, evectors = linalg.eig(A)print(evalues)print(evectors)
[1.13027756+0.j 0.76972244+0.j][[ 0.917245740.79325185] [-0.3983218 0.60889368]]
#%figonly=线性变换将蓝色箭头变换为红色箭头points = np.array([[0, 1.0], [1.0, 0], [1, 1]])def draw_arrows(points, **kw):props = dict(color="blue", arrowstyle="->")props.update(kw)for x, y in points:pl.annotate("",xy=(x, y), xycoords='data',xytext=(0, 0), textcoords='data',arrowprops=props)draw_arrows(points)draw_arrows(np.dot(A, points.T).T, color="red")draw_arrows(evectors.T, alpha=0.7, linewidth=2)draw_arrows(np.dot(A, evectors).T, color="red", alpha=0.7, linewidth=2)ax = pl.gca()ax.set_aspect("equal")ax.set_xlim(-0.5, 1.1)ax.set_ylim(-0.5, 1.1);

np.random.seed(42)t = np.random.uniform(0, 2*np.pi, 60)alpha = 0.4a = 0.5b = 1.0x = 1.0 + a*np.cos(t)*np.cos(alpha) - b*np.sin(t)*np.sin(alpha)y = 1.0 + a*np.cos(t)*np.sin(alpha) - b*np.sin(t)*np.cos(alpha)x += np.random.normal(0, 0.05, size=len(x))y += np.random.normal(0, 0.05, size=len(y))
D = np.c_[x**2, x*y, y**2, x, y, np.ones_like(x)]A = np.dot(D.T, D)C = np.zeros((6, 6))C[[0, 1, 2], [2, 1, 0]] = 2, -1, 2evalues, evectors = linalg.eig(A, C) #❶evectors = np.real(evectors)err = np.mean(np.dot(D, evectors)**2, 0) #❷p = evectors[:, np.argmin(err) ] #❸print (p)
[-0.552142780.5580915-0.238099220.54584559 -0.08350449 -0.14852803]
#%figonly=用广义特征向量计算的拟合椭圆def ellipse(p, x, y):a, b, c, d, e, f = preturn a*x**2 + b*x*y + c*y**2 + d*x + e*y + fX, Y = np.mgrid[0:2:100j, 0:2:100j]Z = ellipse(p, X, Y)pl.plot(x, y, "ro", alpha=0.5)pl.contour(X, Y, Z, levels=[0]);

奇异值分解-SVD

r, g, b = np.rollaxis(pl.imread("vinci_target.png"), 2).astype(float)img = 0.2989 * r + 0.5870 * g + 0.1140 * bimg.shape
(505, 375)
U, s, Vh = linalg.svd(img)print(U.shape)print(s.shape)print(Vh.shape)
(505, 505)(375,)(375, 375)
#%fig=按从大到小排列的奇异值pl.semilogy(s, lw=3);

output_20_1
def composite(U, s, Vh, n):return np.dot(U[:, :n], s[:n, np.newaxis] * Vh[:n, :])print (np.allclose(img, composite(U, s, Vh, len(s))))
True
#%fig=原始图像、使用10、20、50个向量合成的图像(从左到右)img10 = composite(U, s, Vh, 10)img20 = composite(U, s, Vh, 20)img50 = composite(U, s, Vh, 50)
%array_image img; img10; img20; img50
UsageError: Line magic function `%array_image` not found.
pl.imshow(img)

pl.imshow(img10)

pl.imshow(img20)

pl.imshow(img50)

4.统计-stats

import numpy as npimport pylab as plfrom scipy import stats
import matplotlib.pyplot as pltimport matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']

连续概率分布

from scipy import stats[k for k, v in stats.__dict__.items() if isinstance(v, stats.rv_continuous)]
['ksone', 'kstwobign', 'norm', 'alpha', 'anglit', 'arcsine', 'beta', 'betaprime', 'bradford', 'burr', 'burr12', 'fisk', 'cauchy', 'chi', 'chi2', 'cosine', 'dgamma', 'dweibull', 'expon', 'exponnorm', 'exponweib', 'exponpow', 'fatiguelife', 'foldcauchy', 'f', 'foldnorm', 'weibull_min', 'weibull_max', 'frechet_r', 'frechet_l', 'genlogistic', 'genpareto', 'genexpon', 'genextreme', 'gamma', 'erlang', 'gengamma', 'genhalflogistic', 'gompertz', 'gumbel_r', 'gumbel_l', 'halfcauchy', 'halflogistic', 'halfnorm', 'hypsecant', 'gausshyper', 'invgamma', 'invgauss', 'invweibull', 'johnsonsb', 'johnsonsu', 'laplace', 'levy', 'levy_l', 'levy_stable', 'logistic', 'loggamma', 'loglaplace', 'lognorm', 'gilbrat', 'maxwell', 'mielke', 'kappa4', 'kappa3', 'nakagami', 'ncx2', 'ncf', 't', 'nct', 'pareto', 'lomax', 'pearson3', 'powerlaw', 'powerlognorm', 'powernorm', 'rdist', 'rayleigh', 'reciprocal', 'rice', 'recipinvgauss', 'semicircular', 'skewnorm', 'trapz', 'triang', 'truncexpon', 'truncnorm', 'tukeylambda', 'uniform', 'vonmises', 'vonmises_line', 'wald', 'wrapcauchy', 'gennorm', 'halfgennorm', 'crystalball', 'argus']
stats.norm.stats()
(array(0.), array(1.))
X = stats.norm(loc=1.0, scale=2.0)X.stats()
(array(1.), array(4.))
x = X.rvs(size=10000) # 对随机变量取10000个值np.mean(x), np.var(x) # 期望值和方差
(1.0048352738823323, 3.9372117720073554)
stats.norm.fit(x) # 得到随机序列期望值和标准差
(1.0048352738823323, 1.984240855341749)
pdf, t = np.histogram(x, bins=100, normed=True)#❶t = (t[:-1] + t[1:]) * 0.5#❷cdf = np.cumsum(pdf) * (t[1] - t[0])#❸p_error = pdf - X.pdf(t)c_error = cdf - X.cdf(t)print ("max pdf error: {}, max cdf error: {}".format(np.abs(p_error).max(),np.abs(c_error).max()))
max pdf error: 0.018998755595167102, max cdf error: 0.018503342378306975
#%figonly=正态分布的概率密度函数(左)和累积分布函数(右)fig, (ax1, ax2) = pl.subplots(1, 2, figsize=(7, 2))ax1.plot(t, pdf, label=u"统计值")ax1.plot(t, X.pdf(t), label=u"理论值", alpha=0.6)ax1.legend(loc="best")ax2.plot(t, cdf)ax2.plot(t, X.cdf(t), alpha=0.6);

print(stats.gamma.stats(1.0))print(stats.gamma.stats(2.0))
(array(1.), array(1.))(array(2.), array(2.))
stats.gamma.stats(2.0, scale=2)
(array(4.), array(8.))
x = stats.gamma.rvs(2, scale=2, size=4)x
array([4.40563983, 6.17699951, 3.65503843, 3.28052152])
stats.gamma.pdf(x, 2, scale=2)
array([0.12169605, 0.07037188, 0.14694352, 0.15904745])
X = stats.gamma(2, scale=2)X.pdf(x)
array([0.12169605, 0.07037188, 0.14694352, 0.15904745])

离散概率分布

x = range(1, 7)p = (0.4, 0.2, 0.1, 0.1, 0.1, 0.1)
dice = stats.rv_discrete(values=(x, p))dice.rvs(size=20)
array([2, 5, 2, 6, 1, 6, 6, 5, 3, 1, 5, 2, 1, 1, 1, 1, 1, 2, 1, 6])
np.random.seed(42)samples = dice.rvs(size=(20000, 50))samples_mean = np.mean(samples, axis=1)

核密度估计

#%fig=核密度估计能更准确地表示随机变量的概率密度函数_, bins, step = pl.hist(samples_mean, bins=100, normed=True, histtype="step", label=u"直方图统计")kde = stats.kde.gaussian_kde(samples_mean)x = np.linspace(bins[0], bins[-1], 100)pl.plot(x, kde(x), label=u"核密度估计")mean, std = stats.norm.fit(samples_mean)pl.plot(x, stats.norm(mean, std).pdf(x), alpha=0.8, label=u"正态分布拟合")pl.legend()

#%fig=`bw_method`参数越大核密度估计曲线越平滑for bw in [0.2, 0.3, 0.6, 1.0]:kde = stats.gaussian_kde([-1, 0, 1], bw_method=bw)x = np.linspace(-5, 5, 1000)y = kde(x)pl.plot(x, y, lw=2, label="bw={}".format(bw), alpha=0.6)pl.legend(loc="best");

二项、泊松、伽玛分布

stats.binom.pmf(range(6), 5, 1/6.0)
array([4.01877572e-01, 4.01877572e-01, 1.60751029e-01, 3.21502058e-02, 3.21502058e-03, 1.28600823e-04])
#%fig=当n足够大时二项分布和泊松分布近似相等lambda_ = 10.0x = np.arange(20)n1, n2 = 100, 1000y_binom_n1 = stats.binom.pmf(x, n1, lambda_ / n1)y_binom_n2 = stats.binom.pmf(x, n2, lambda_ / n2)y_poisson = stats.poisson.pmf(x, lambda_)print(np.max(np.abs(y_binom_n1 - y_poisson)))print(np.max(np.abs(y_binom_n2 - y_poisson)))#%hidefig, (ax1, ax2) = pl.subplots(1, 2, figsize=(7.5, 2.5))ax1.plot(x, y_binom_n1, label=u"binom", lw=2)ax1.plot(x, y_poisson, label=u"poisson", lw=2, color="red")ax2.plot(x, y_binom_n2, label=u"binom", lw=2)ax2.plot(x, y_poisson, label=u"poisson", lw=2, color="red")for n, ax in zip((n1, n2), (ax1, ax2)):ax.set_xlabel(u"次数")ax.set_ylabel(u"概率")ax.set_title("n={}".format(n))ax.legend()fig.subplots_adjust(0.1, 0.15, 0.95, 0.90, 0.2, 0.1)
0.006755311103353090.0006301754049777564

#%fig=模拟泊松分布np.random.seed(42)def sim_poisson(lambda_, time):t = np.random.uniform(0, time, size=lambda_ * time)#❶count, time_edges = np.histogram(t, bins=time, range=(0, time))#❷dist, count_edges = np.histogram(count, bins=20, range=(0, 20), density=True)#❸x = count_edges[:-1]poisson = stats.poisson.pmf(x, lambda_)return x, poisson, distlambda_ = 10times = 1000, 50000x1, poisson1, dist1 = sim_poisson(lambda_, times[0])x2, poisson2, dist2 = sim_poisson(lambda_, times[1])max_error1 = np.max(np.abs(dist1 - poisson1))max_error2 = np.max(np.abs(dist2 - poisson2))print("time={}, max_error={}".format(times[0], max_error1))print("time={}, max_error={}".format(times[1], max_error2))#%hidefig, (ax1, ax2) = pl.subplots(1, 2, figsize=(7.5, 2.5))ax1.plot(x1, dist1, "-o", lw=2, label=u"统计结果")ax1.plot(x1, poisson1, "->", lw=2, label=u"泊松分布", color="red", alpha=0.6)ax2.plot(x2, dist2, "-o", lw=2, label=u"统计结果")ax2.plot(x2, poisson2, "->", lw=2, label=u"泊松分布", color="red", alpha=0.6)for ax, time in zip((ax1, ax2), times):ax.set_xlabel(u"次数")ax.set_ylabel(u"概率")ax.set_title(u"time = {}".format(time))ax.legend(loc="lower center")fig.subplots_adjust(0.1, 0.15, 0.95, 0.90, 0.2, 0.1)
time=1000, max_error=0.01964230201602718time=50000, max_error=0.001798012894964722

#%fig=模拟伽玛分布def sim_gamma(lambda_, time, k):t = np.random.uniform(0, time, size=lambda_ * time) #❶t.sort()#❷interval = t[k:] - t[:-k] #❸dist, interval_edges = np.histogram(interval, bins=100, density=True) #❹x = (interval_edges[1:] + interval_edges[:-1])/2#❺gamma = stats.gamma.pdf(x, k, scale=1.0/lambda_) #❺return x, gamma, distlambda_ = 10time = 1000ks = 1, 2x1, gamma1, dist1 = sim_gamma(lambda_, time, ks[0])x2, gamma2, dist2 = sim_gamma(lambda_, time, ks[1])#%hidefig, (ax1, ax2) = pl.subplots(1, 2, figsize=(7.5, 2.5))ax1.plot(x1, dist1,lw=2, label=u"统计结果")ax1.plot(x1, gamma1, lw=2, label=u"伽玛分布", color="red", alpha=0.6)ax2.plot(x2, dist2,lw=2, label=u"统计结果")ax2.plot(x2, gamma2, lw=2, label=u"伽玛分布", color="red", alpha=0.6)for ax, k in zip((ax1, ax2), ks):ax.set_xlabel(u"时间间隔")ax.set_ylabel(u"概率密度")ax.set_title(u"k = {}".format(k))ax.legend(loc="upper right")fig.subplots_adjust(0.1, 0.15, 0.95, 0.90, 0.2, 0.1);
png
T = 100000A_count = int(T / 5)B_count = int(T / 10)A_time = np.random.uniform(0, T, A_count) #❶B_time = np.random.uniform(0, T, B_count)bus_time = np.concatenate((A_time, B_time)) #❷bus_time.sort()N = 200000passenger_time = np.random.uniform(bus_time[0], bus_time[-1], N) #❸idx = np.searchsorted(bus_time, passenger_time) #❹np.mean(bus_time[idx] - passenger_time) * 60#❺
202.3388747879705
np.mean(np.diff(bus_time)) * 60
199.99833251643057
#%figonly=观察者偏差import matplotlib.gridspec as gridspecpl.figure(figsize=(7.5, 3))G = gridspec.GridSpec(10, 1)ax1 = pl.subplot(G[:2,0])ax2 = pl.subplot(G[3:, 0])ax1.vlines(bus_time[:10], 0, 1, lw=2, color="blue", label=u"公交车")ptime = np.random.uniform(bus_time[0], bus_time[9], 100)ax1.vlines(ptime, 0, 1, lw=1, color="red", alpha=0.2, label=u"乘客")ax1.legend()count, bins = np.histogram(passenger_time, bins=bus_time)ax2.plot(np.diff(bins), count, ".", alpha=0.3, rasterized=True)ax2.set_xlabel(u"公交车的时间间隔")ax2.set_ylabel(u"等待人数");

from scipy import integratet = 10.0 / 3# 两辆公交车之间的平均时间间隔bus_interval = stats.gamma(1, scale=t)n, _ = integrate.quad(lambda x: 0.5 * x * x * bus_interval.pdf(x), 0, 1000)d, _ = integrate.quad(lambda x: x * bus_interval.pdf(x), 0, 1000)n / d * 60
200.0

学生 t-分布与 t 检验

#%fig=模拟学生t-分布mu = 0.0n = 10samples = stats.norm(mu).rvs(size=(100000, n))#❶t_samples = (np.mean(samples, axis=1) - mu) / np.std(samples, ddof=1, axis=1) * n**0.5#❷sample_dist, x = np.histogram(t_samples, bins=100, density=True)#❸x = 0.5 * (x[:-1] + x[1:])t_dist = stats.t(n - 1).pdf(x)print("max error:", np.max(np.abs(sample_dist - t_dist)))#%hidepl.plot(x, sample_dist, lw=2, label=u"样本分布")pl.plot(x, t_dist, lw=2, alpha=0.6, label=u"t分布")pl.xlim(-5, 5)pl.legend(loc="best")
max error: 0.006832108369761447

#%figonly=当`df`增大,学生t-分布趋向于正态分布fig, (ax1, ax2) = pl.subplots(1, 2, figsize=(7.5, 2.5))ax1.plot(x, stats.t(6-1).pdf(x), label=u"df=5", lw=2)ax1.plot(x, stats.t(40-1).pdf(x), label=u"df=39", lw=2, alpha=0.6)ax1.plot(x, stats.norm.pdf(x), "k--", label=u"norm")ax1.legend()ax2.plot(x, stats.t(6-1).sf(x), label=u"df=5", lw=2)ax2.plot(x, stats.t(40-1).sf(x), label=u"df=39", lw=2, alpha=0.6)ax2.plot(x, stats.norm.sf(x), "k--", label=u"norm")ax2.legend();

n = 30np.random.seed(42)s = stats.norm.rvs(loc=1, scale=0.8, size=n)
t = (np.mean(s) - 0.5) / (np.std(s, ddof=1) / np.sqrt(n))print (t, stats.ttest_1samp(s, 0.5))
2.658584340882224 Ttest_1sampResult(statistic=2.658584340882224, pvalue=0.01263770225709123)
print ((np.mean(s) - 1) / (np.std(s, ddof=1) / np.sqrt(n)))print (stats.ttest_1samp(s, 1), stats.ttest_1samp(s, 0.9))
-1.1450173670383303Ttest_1sampResult(statistic=-1.1450173670383303, pvalue=0.26156414618801477) Ttest_1sampResult(statistic=-0.3842970254542196, pvalue=0.7035619103425202)
#%fig=红色部分为`ttest_1samp()`计算的p值x = np.linspace(-5, 5, 500)y = stats.t(n-1).pdf(x)plt.plot(x, y, lw=2)t, p = stats.ttest_1samp(s, 0.5)mask = x > np.abs(t)plt.fill_between(x[mask], y[mask], color="red", alpha=0.5)mask = x < -np.abs(t)plt.fill_between(x[mask], y[mask], color="red", alpha=0.5)plt.axhline(color="k", lw=0.5)plt.xlim(-5, 5);

from scipy import integratex = np.linspace(-10, 10, 100000)y = stats.t(n-1).pdf(x)mask = x >= np.abs(t)integrate.trapz(y[mask], x[mask])*2
0.012633433707685974
m = 200000mean = 0.5r = stats.norm.rvs(loc=mean, scale=0.8, size=(m, n))ts = (np.mean(s) - mean) / (np.std(s, ddof=1) / np.sqrt(n))tr = (np.mean(r, axis=1) - mean) / (np.std(r, ddof=1, axis=1) / np.sqrt(n))np.mean(np.abs(tr) > np.abs(ts))
0.012695
np.random.seed(42)s1 = stats.norm.rvs(loc=1, scale=1.0, size=20)s2 = stats.norm.rvs(loc=1.5, scale=0.5, size=20)s3 = stats.norm.rvs(loc=1.5, scale=0.5, size=25)print (stats.ttest_ind(s1, s2, equal_var=False)) #❶print (stats.ttest_ind(s2, s3, equal_var=True))#❷
Ttest_indResult(statistic=-2.2391470627176755, pvalue=0.033250866086743665)Ttest_indResult(statistic=-0.5946698521856172, pvalue=0.5551805875810539)

卡方分布和卡方检验

#%fig=使用随机数验证卡方分布a = np.random.normal(size=(300000, 4))cs = np.sum(a**2, axis=1)sample_dist, bins = np.histogram(cs, bins=100, range=(0, 20), density=True)x = 0.5 * (bins[:-1] + bins[1:])chi2_dist = stats.chi2.pdf(x, 4)print("max error:", np.max(np.abs(sample_dist - chi2_dist)))#%hidepl.plot(x, sample_dist, lw=2, label=u"样本分布")pl.plot(x, chi2_dist, lw=2, alpha=0.6, label=u"$\chi ^{2}$分布")pl.legend(loc="best")
max error: 0.0030732520533635066

#%fig=模拟卡方分布repeat_count = 60000n, k = 100, 5np.random.seed(42)ball_ids = np.random.randint(0, k, size=(repeat_count, n)) #❶counts = np.apply_along_axis(np.bincount, 1, ball_ids, minlength=k) #❷cs2 = np.sum((counts - n/k)**2.0/(n/k), axis=1) #❸k = stats.kde.gaussian_kde(cs2) #❹x = np.linspace(0, 10, 200)pl.plot(x, stats.chi2.pdf(x, 4), lw=2, label=u"$\chi ^{2}$分布")pl.plot(x, k(x), lw=2, color="red", alpha=0.6, label=u"样本分布")pl.legend(loc="best")pl.xlim(0, 10);

def choose_balls(probabilities, size):r = stats.rv_discrete(values=(range(len(probabilities)), probabilities))s = r.rvs(size=size)counts = np.bincount(s)return countsnp.random.seed(42)counts1 = choose_balls([0.18, 0.24, 0.25, 0.16, 0.17], 400)counts2 = choose_balls([0.2]*5, 400)print(counts1)print(counts2)
[80 93 97 64 66][89 76 79 71 85]
chi1, p1 = stats.chisquare(counts1)chi2, p2 = stats.chisquare(counts2)print ("chi1 =", chi1, "p1 =", p1)print ("chi2 =", chi2, "p2 =", p2)
chi1 = 11.375 p1 = 0.022657601239769634chi2 = 2.55 p2 = 0.6357054527037017
#%figonly=卡方检验计算的概率为阴影部分的面积x = np.linspace(0, 30, 200)CHI2 = stats.chi2(4)pl.plot(x, CHI2.pdf(x), "k", lw=2)pl.vlines(chi1, 0, CHI2.pdf(chi1))pl.vlines(chi2, 0, CHI2.pdf(chi2))pl.fill_between(x[x>chi1], 0, CHI2.pdf(x[x>chi1]), color="red", alpha=1.0)pl.fill_between(x[x>chi2], 0, CHI2.pdf(x[x>chi2]), color="green", alpha=0.5)pl.text(chi1, 0.015, r"$\chi^2_1$", fontsize=14)pl.text(chi2, 0.015, r"$\chi^2_2$", fontsize=14)pl.ylim(0, 0.2)pl.xlim(0, 20);

table = [[43, 9], [44, 4]]chi2, p, dof, expected = stats.chi2_contingency(table)print(chi2)print(p)
1.07248520710059210.300384770390566
stats.fisher_exact(table)
(0.43434343434343436, 0.23915695682224306)

5.数值积分-integrate

import pylab as plimport numpy as npfrom scipy import integratefrom scipy.integrate import odeintimport matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']

球的体积

def half_circle(x):return (1-x**2)**0.5
N = 10000x = np.linspace(-1, 1, N)dx = x[1] - x[0]y = half_circle(x)2 * dx * np.sum(y) # 面积的两倍
3.1415893269307373
np.trapz(y, x) * 2 # 面积的两倍
3.1415893269315975
from scipy import integratepi_half, err = integrate.quad(half_circle, -1, 1)pi_half * 2
3.141592653589797
def half_sphere(x, y):return (1-x**2-y**2)**0.5
volume, error = integrate.dblquad(half_sphere, -1, 1,lambda x:-half_circle(x),lambda x:half_circle(x))print (volume, error, np.pi*4/3/2)
2.094395102393199 1.0002356720661965e-09 2.0943951023931953

解常微分方程组

#%fig=洛伦茨吸引子:微小的初值差别也会显著地影响运动轨迹from scipy.integrate import odeintimport numpy as npdef lorenz(w, t, p, r, b): #❶# 给出位置矢量w,和三个参数p, r, b计算出# dx/dt, dy/dt, dz/dt的值x, y, z = w.tolist()# 直接与lorenz的计算公式对应return p*(y-x), x*(r-z)-y, x*y-b*zt = np.arange(0, 30, 0.02) # 创建时间点# 调用ode对lorenz进行求解, 用两个不同的初始值track1 = odeint(lorenz, (0.0, 1.00, 0.0), t, args=(10.0, 28.0, 3.0)) #❷track2 = odeint(lorenz, (0.0, 1.01, 0.0), t, args=(10.0, 28.0, 3.0)) #❸#%hidefrom mpl_toolkits.mplot3d import Axes3Dfig = pl.figure()ax = Axes3D(fig)ax.plot(track1[:,0], track1[:,1], track1[:,2], lw=1)ax.plot(track2[:,0], track2[:,1], track2[:,2], lw=1);

ode 类

def mass_spring_damper(xu, t, m, k, b, F):x, u = xu.tolist()dx = udu = (F - k*x - b*u)/mreturn dx, du
#%fig=滑块的速度和位移曲线m, b, k, F = 1.0, 10.0, 20.0, 1.0init_status = 0.0, 0.0args = m, k, b, Ft = np.arange(0, 2, 0.01)result = odeint(mass_spring_damper, init_status, t, args)#%hidefig, (ax1, ax2) = pl.subplots(2, 1)ax1.plot(t, result[:, 0], label=u"位移")ax1.legend()ax2.plot(t, result[:, 1], label=u"速度")ax2.legend();

from scipy.integrate import odeclass MassSpringDamper(object): #❶def __init__(self, m, k, b, F):self.m, self.k, self.b, self.F = m, k, b, Fdef f(self, t, xu):x, u = xu.tolist()dx = udu = (self.F - self.k*x - self.b*u)/self.mreturn [dx, du]system = MassSpringDamper(m=m, k=k, b=b, F=F)init_status = 0.0, 0.0dt = 0.01r = ode(system.f) #❷r.set_integrator('vode', method='bdf')r.set_initial_value(init_status, 0)t = []result2 = [init_status]while r.successful() and r.t + dt < 2: #❸r.integrate(r.t + dt)t.append(r.t)result2.append(r.y)result2 = np.array(result2)np.allclose(result, result2)
True
class PID(object):def __init__(self, kp, ki, kd, dt):self.kp, self.ki, self.kd, self.dt = kp, ki, kd, dtself.last_error = Noneself.status = 0.0def update(self, error):p = self.kp * errori = self.ki * self.statusif self.last_error is None:d = 0.0else:d = self.kd * (error - self.last_error) / self.dtself.status += error * self.dtself.last_error = errorreturn p + i + d
#%fig=使用PID控制器让滑块停在位移为1.0处def pid_control_system(kp, ki, kd, dt, target=1.0):system = MassSpringDamper(m=m, k=k, b=b, F=0.0)pid = PID(kp, ki, kd, dt)init_status = 0.0, 0.0r = ode(system.f)r.set_integrator('vode', method='bdf')r.set_initial_value(init_status, 0)t = [0]result = [init_status]F_arr = [0]while r.successful() and r.t + dt < 2.0:r.integrate(r.t + dt)err = target - r.y[0]#❶F = pid.update(err)#❷system.F = F#❸t.append(r.t)result.append(r.y)F_arr.append(F)result = np.array(result)t = np.array(t)F_arr = np.array(F_arr)return t, F_arr, resultt, F_arr, result = pid_control_system(50.0, 100.0, 10.0, 0.001)print(u"控制力的终值:", F_arr[-1])#%hidefig, (ax1, ax2, ax3) = pl.subplots(3, 1, figsize=(6, 6))ax1.plot(t, result[:, 0], label=u"位移")ax1.legend(loc="best")ax2.plot(t, result[:, 1], label=u"速度")ax2.legend(loc="best")ax3.plot(t, F_arr, label=u"控制力")ax3.legend(loc="best")
控制力的终值: 19.943404699515057

%%timefrom scipy import optimizedef eval_func(k):kp, ki, kd = kt, F_arr, result = pid_control_system(kp, ki, kd, 0.01)return np.sum(np.abs(result[:, 0] - 1.0))kwargs = {"method": "L-BFGS-B","bounds": [(10, 200), (10, 100), (1, 100)],"options": {"approx_grad": True}}opt_k = optimize.basinhopping(eval_func, (10, 10, 10), niter=10, minimizer_kwargs=kwargs)print(opt_k.x)
[56.67106149 99.974347571.33963577]Wall time: 55.1 s
#%fig=优化PID的参数降低控制响应时间kp, ki, kd = opt_k.xt, F_arr, result = pid_control_system(kp, ki, kd, 0.01)idx = np.argmin(np.abs(t - 0.5))x, u = result[idx]print ("t={}, x={:g}, u={:g}".format(t[idx], x, u))#%hidefig, (ax1, ax2, ax3) = pl.subplots(3, 1, figsize=(6, 6))ax1.plot(t, result[:, 0], label=u"位移")ax1.legend(loc="best")ax2.plot(t, result[:, 1], label=u"速度")ax2.legend(loc="best")ax3.plot(t, F_arr, label=u"控制力")ax3.legend(loc="best");
t=0.5000000000000002, x=1.07098, u=0.315352

6.信号处理-signal

import pylab as plimport numpy as npfrom scipy import signal
import matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']

中值滤波

#%fig=使用中值滤波剔除瞬间噪声t = np.arange(0, 20, 0.1)x = np.sin(t)x[np.random.randint(0, len(t), 20)] += np.random.standard_normal(20)*0.6 #❶x2 = signal.medfilt(x, 5) #❷x3 = signal.order_filter(x, np.ones(5), 2)print (np.all(x2 == x3))pl.plot(t, x, label=u"带噪声的信号")pl.plot(t, x2 + 0.5, alpha=0.6, label=u"中值滤波之后的信号")pl.legend(loc="best");
True

output_4_1

滤波器设计

sampling_rate = 8000.0# 设计一个带通滤波器:# 通带为0.2*4000 - 0.5*4000# 阻带为0.6*4000# 通带增益的最大衰减值为2dB# 阻带的最小衰减值为40dBb, a = signal.iirdesign([0.2, 0.5], [0.1, 0.6], 2, 40) #❶# 使用freq计算滤波器的频率响应w, h = signal.freqz(b, a) #❷# 计算增益power = 20*np.log10(np.clip(np.abs(h), 1e-8, 1e100)) #❸freq = w / np.pi * sampling_rate / 2
#%fig=用频率扫描波测量的频率响应# 产生2秒钟的取样频率为sampling_rate Hz的频率扫描信号# 开始频率为0, 结束频率为sampling_rate/2t = np.arange(0, 2, 1/sampling_rate) #❶sweep = signal.chirp(t, f0=0, t1=2, f1=sampling_rate/2) #❷# 对频率扫描信号进行滤波out = signal.lfilter(b, a, sweep) #❸# 将波形转换为能量out = 20*np.log10(np.abs(out)) #❹# 找到所有局部最大值的下标index = signal.argrelmax(out, order=3)#❺# 绘制滤波之后的波形的增益pl.figure(figsize=(8, 2.5))pl.plot(freq, power, label=u"带通IIR滤波器的频率响应")pl.plot(t[index]/2.0*4000, out[index], label=u"频率扫描波测量的频谱", alpha=0.6) #❻pl.legend(loc="best")#%hidepl.title(u"频率扫描波测量的滤波器频谱")pl.ylim(-100,20)pl.ylabel(u"增益(dB)")pl.xlabel(u"频率(Hz)");

连续时间线性系统

#%fig=系统的阶跃响应和正弦波响应m, b, k = 1.0, 10, 20numerator = [1]denominator = [m, b, k]plant = signal.lti(numerator, denominator)#❶t = np.arange(0, 2, 0.01)_, x_step = plant.step(T=t)#❷_, x_sin, _ = signal.lsim(plant, U=np.sin(np.pi * t), T=t)#❸#%hidepl.plot(t, x_step, label=u"阶跃响应")pl.plot(t, x_sin, label=u"正弦波响应")pl.legend(loc="best")pl.xlabel(u"时间(秒)")pl.ylabel(u"位移(米)")
Text(0,0.5,'位移(米)')

7.插值-interpolate

import numpy as npimport pylab as plfrom scipy import interpolate
import matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']

一维插值

WARNING:高次interp1d()插值的运算量很大,因此对于点数较多的数据,建议使用后面介绍的UnivariateSpline()

#%fig=`interp1d`的各阶插值from scipy import interpolatex = np.linspace(0, 10, 11)y = np.sin(x)xnew = np.linspace(0, 10, 101)pl.plot(x, y, 'ro')for kind in ['nearest', 'zero', 'slinear', 'quadratic']:f = interpolate.interp1d(x, y, kind=kind)#❶ynew = f(xnew)#❷pl.plot(xnew, ynew, label=str(kind))pl.legend(loc='lower right')

output_5_1
外推和 Spline 拟合
#%fig=使用UnivariateSpline进行插值:外推(上),数据拟合(下)x1 = np.linspace(0, 10, 20)y1 = np.sin(x1)sx1 = np.linspace(0, 12, 100)sy1 = interpolate.UnivariateSpline(x1, y1, s=0)(sx1)#❶x2 = np.linspace(0, 20, 200)y2 = np.sin(x2) + np.random.standard_normal(len(x2)) * 0.2sx2 = np.linspace(0, 20, 2000)spline2 = interpolate.UnivariateSpline(x2, y2, s=8)#❷sy2 = spline2(sx2)pl.figure(figsize=(8, 5))pl.subplot(211)pl.plot(x1, y1, ".", label=u"数据点")pl.plot(sx1, sy1, label=u"spline曲线")pl.legend()pl.subplot(212)pl.plot(x2, y2, ".", label=u"数据点")pl.plot(sx2, sy2, linewidth=2, label=u"spline曲线")pl.plot(x2, np.sin(x2), label=u"无噪声曲线")pl.legend()

output_7_1
print(np.array_str(spline2.roots(), precision=3))
[ 0.0533.1516.36 9.386 12.603 15.619 18.929]
#%fig=计算Spline与水平线的交点def roots_at(self, v):#❶coeff = self.get_coeffs()coeff -= vtry:root = self.roots()return rootfinally:coeff += vinterpolate.UnivariateSpline.roots_at = roots_at#❷pl.plot(sx2, sy2, linewidth=2, label=u"spline曲线")ax = pl.gca()for level in [0.5, 0.75, -0.5, -0.75]:ax.axhline(level, ls=":", color="k")xr = spline2.roots_at(level)#❸pl.plot(xr, spline2(xr), "ro")

参数插值
#%fig=使用参数插值连接二维平面上的点x = [4.913, 4.913, 4.918, 4.938, 4.955, 4.949, 4.911, 4.848, 4.864, 4.893,4.935, 4.981, 5.01, 5.021]y = [5.2785, 5.2875, 5.291, 5.289, 5.28, 5.26, 5.245, 5.245, 5.2615, 5.278,5.2775, 5.261, 5.245, 5.241]pl.plot(x, y, "o")for s in (0, 1e-4):tck, t = interpolate.splprep([x, y], s=s)#❶xi, yi = interpolate.splev(np.linspace(t[0], t[-1], 200), tck)#❷pl.plot(xi, yi, lw=2, label=u"s=%g" % s)pl.legend()

单调插值
import numpy as npimport matplotlib.pyplot as pltfrom scipy import interpolatex = np.arange(0, 2 * np.pi + np.pi / 4, 2 * np.pi / 8)y = np.sin(x)tck = interpolate.splrep(x, y, s=0)xnew = np.arange(0, 2 * np.pi, np.pi / 50)ynew = interpolate.splev(xnew, tck, der=0)plt.figure()plt.plot(x, y, 'x', xnew, ynew, xnew, np.sin(xnew), x, y, 'b')plt.legend(['Linear', 'Cubic Spline', 'True'])plt.axis([-0.05, 6.33, -1.05, 1.05])plt.title('三次样条插值')plt.show()

多维插值

#%fig=使用interp2d类进行二维插值def func(x, y):#❶return (x + y) * np.exp(-5.0 * (x**2 + y**2))# X-Y轴分为15*15的网格y, x = np.mgrid[-1:1:15j, -1:1:15j]#❷fvals = func(x, y)# 计算每个网格点上的函数值# 二维插值newfunc = interpolate.interp2d(x, y, fvals, kind='cubic')#❸# 计算100*100的网格上的插值xnew = np.linspace(-1, 1, 100)ynew = np.linspace(-1, 1, 100)fnew = newfunc(xnew, ynew)#❹#%hidepl.subplot(121)pl.imshow(fvals,extent=[-1, 1, -1, 1],cmap=pl.cm.jet,interpolation='nearest',origin="lower")pl.title("fvals")pl.subplot(122)pl.imshow(fnew,extent=[-1, 1, -1, 1],cmap=pl.cm.jet,interpolation='nearest',origin="lower")pl.title("fnew")pl.show()

griddata

WARNING

griddata()使用欧几里得距离计算插值。如果 K 维空间中每个维度的取值范围相差较大,则应先将数据正规化,然后使用griddata()进行插值运算。

#%fig=使用gridata进行二维插值# 计算随机N个点的坐标,以及这些点对应的函数值N = 200np.random.seed(42)x = np.random.uniform(-1, 1, N)y = np.random.uniform(-1, 1, N)z = func(x, y)yg, xg = np.mgrid[-1:1:100j, -1:1:100j]xi = np.c_[xg.ravel(), yg.ravel()]methods = 'nearest', 'linear', 'cubic'zgs = [interpolate.griddata((x, y), z, xi, method=method).reshape(100, 100)for method in methods]#%hidefig, axes = pl.subplots(1, 3, figsize=(11.5, 3.5))for ax, method, zg in zip(axes, methods, zgs):ax.imshow(zg,extent=[-1, 1, -1, 1],cmap=pl.cm.jet,interpolation='nearest',origin="lower")ax.set_xlabel(method)ax.scatter(x, y, c=z)

径向基函数插值
#%fig=一维RBF插值from scipy.interpolate import Rbfx1 = np.array([-1, 0, 2.0, 1.0])y1 = np.array([1.0, 0.3, -0.5, 0.8])funcs = ['multiquadric', 'gaussian', 'linear']nx = np.linspace(-3, 4, 100)rbfs = [Rbf(x1, y1, function=fname) for fname in funcs]#❶rbf_ys = [rbf(nx) for rbf in rbfs]#❷#%hidepl.plot(x1, y1, "o")for fname, ny in zip(funcs, rbf_ys):pl.plot(nx, ny, label=fname, lw=2)pl.ylim(-1.0, 1.5)pl.legend()

output_20_1
for fname, rbf in zip(funcs, rbfs):print (fname, rbf.nodes)
multiquadric [-0.888228852.176545131.42877511 -2.67919021]gaussian [ 1.00321945 -0.02345964 -0.654417160.91375159]linear [-0.266666670.6 0.73333333 -0.9 ]
#%fig=二维径向基函数插值rbfs = [Rbf(x, y, z, function=fname) for fname in funcs]rbf_zg = [rbf(xg, yg).reshape(xg.shape) for rbf in rbfs]#%hidefig, axes = pl.subplots(1, 3, figsize=(11.5, 3.5))for ax, fname, zg in zip(axes, funcs, rbf_zg):ax.imshow(zg,extent=[-1, 1, -1, 1],cmap=pl.cm.jet,interpolation='nearest',origin="lower")ax.set_xlabel(fname)ax.scatter(x, y, c=z)

#%fig=`epsilon`参数指定径向基函数中数据点的作用范围epsilons = 0.1, 0.15, 0.3rbfs = [Rbf(x, y, z, function="gaussian", epsilon=eps) for eps in epsilons]zgs = [rbf(xg, yg).reshape(xg.shape) for rbf in rbfs]#%hidefig, axes = pl.subplots(1, 3, figsize=(11.5, 3.5))for ax, eps, zg in zip(axes, epsilons, zgs):ax.imshow(zg,extent=[-1, 1, -1, 1],cmap=pl.cm.jet,interpolation='nearest',origin="lower")ax.set_xlabel("eps=%g" % eps)ax.scatter(x, y, c=z)

8.稀疏矩阵-sparse

%matplotlib inlineimport numpy as npimport pylab as plfrom scipy import sparsefrom scipy.sparse import csgraph

稀疏矩阵的储存形式

from scipy import sparsea = sparse.dok_matrix((10, 5))a[2, 3] = 1.0a[3, 3] = 2.0a[4, 3] = 3.0print(a.keys())print(a.values())
dict_keys([(2, 3), (3, 3), (4, 3)])dict_values([1.0, 2.0, 3.0])
b = sparse.lil_matrix((10, 5))b[2, 3] = 1.0b[3, 4] = 2.0b[3, 2] = 3.0print(b.data)print(b.rows)
[list([]) list([]) list([1.0]) list([3.0, 2.0]) list([]) list([]) list([]) list([]) list([]) list([])][list([]) list([]) list([3]) list([2, 4]) list([]) list([]) list([]) list([]) list([]) list([])]
row = [2, 3, 3, 2]col = [3, 4, 2, 3]data = [1, 2, 3, 10]c = sparse.coo_matrix((data, (row, col)), shape=(5, 6))print (c.col, c.row, c.data)print (c.toarray())
[3 4 2 3] [2 3 3 2] [ 123 10][[ 000000] [ 000000] [ 000 1100] [ 003020] [ 000000]]

矩阵向量相乘

import numpy as npfrom scipy.sparse import csr_matrixA = csr_matrix([[1, 2, 0], [0, 0, 3], [4, 0, 5]])v = np.array([1, 0, -1])A.dot(v)
array([ 1, -3, -1], dtype=int32)
示例1

构造一个1000×1000 lil_matrix并添加值:

from scipy.sparse import lil_matrixfrom scipy.sparse.linalg import spsolvefrom numpy.linalg import solve, normfrom numpy.random import rand
A = lil_matrix((1000, 1000))A[0, :100] = rand(100)A[1, 100:200] = A[0, :100]A.setdiag(rand(1000))

现在将其转换为CSR格式,并求解的:

A = A.tocsr()b = rand(1000)x = spsolve(A, b)

将其转换为密集矩阵并求解,并检查结果是否相同:

x_ = solve(A.toarray(), b)

现在我们可以使用以下公式计算错误的范数:

err = norm(x-x_)err < 1e-10
True
示例2

构造COO格式的矩阵:

from scipy import sparsefrom numpy import arrayI = array([0,3,1,0])J = array([0,3,1,2])V = array([4,5,7,9])A = sparse.coo_matrix((V,(I,J)),shape=(4,4))

注意,索引不需要排序。

转换为CSR或CSC时,将对重复的(i,j)条目进行求和。

I = array([0,0,1,3,1,0,0])J = array([0,2,1,3,1,0,0])V = array([1,1,1,1,1,1,1])B = sparse.coo_matrix((V,(I,J)),shape=(4,4)).tocsr()

这对于构造有限元刚度矩阵和质量矩阵是有用的。

9.图像处理-ndimage

import numpy as npimport pylab as pl

形态学图像处理

import numpy as npdef expand_image(img, value, out=None, size = 10):if out is None:w, h = img.shapeout = np.zeros((w*size, h*size),dtype=np.uint8)tmp = np.repeat(np.repeat(img,size,0),size,1)out[:,:] = np.where(tmp, value, out)out[::size,:] = 0out[:,::size] = 0return outdef show_image(*imgs): for idx, img in enumerate(imgs, 1):ax = pl.subplot(1, len(imgs), idx)pl.imshow(img, cmap="gray")ax.set_axis_off()pl.subplots_adjust(0.02, 0, 0.98, 1, 0.02, 0)
膨胀和腐蚀
#%fig=四连通和八连通的膨胀运算from scipy.ndimage import morphologydef dilation_demo(a, structure=None):b = morphology.binary_dilation(a, structure)img = expand_image(a, 255)return expand_image(np.logical_xor(a,b), 150, out=img)a = pl.imread("scipy_morphology_demo.png")[:,:,0].astype(np.uint8)img1 = expand_image(a, 255)img2 = dilation_demo(a)img3 = dilation_demo(a, [[1,1,1],[1,1,1],[1,1,1]])show_image(img1, img2, img3)

#%fig=不同结构元素的膨胀效果img4 = dilation_demo(a, [[0,0,0],[1,1,1],[0,0,0]])img5 = dilation_demo(a, [[0,1,0],[0,1,0],[0,1,0]])img6 = dilation_demo(a, [[0,1,0],[0,1,0],[0,0,0]])show_image(img4, img5, img6)

#%fig=四连通和八连通的腐蚀运算def erosion_demo(a, structure=None):b = morphology.binary_erosion(a, structure)img = expand_image(a, 255)return expand_image(np.logical_xor(a,b), 100, out=img)img1 = expand_image(a, 255)img2 = erosion_demo(a)img3 = erosion_demo(a, [[1,1,1],[1,1,1],[1,1,1]])show_image(img1, img2, img3)

Hit和Miss
#%fig=Hit和Miss运算def hitmiss_demo(a, structure1, structure2):b = morphology.binary_hit_or_miss(a, structure1, structure2)img = expand_image(a, 100)return expand_image(b, 255, out=img)img1 = expand_image(a, 255)img2 = hitmiss_demo(a, [[0,0,0],[0,1,0],[1,1,1]], [[1,0,0],[0,0,0],[0,0,0]])img3 = hitmiss_demo(a, [[0,0,0],[0,0,0],[1,1,1]], [[1,0,0],[0,1,0],[0,0,0]])show_image(img1, img2, img3)

#%fig=使用Hit和Miss进行细线化运算def skeletonize(img):h1 = np.array([[0, 0, 0],[0, 1, 0],[1, 1, 1]]) #❶m1 = np.array([[1, 1, 1],[0, 0, 0],[0, 0, 0]])h2 = np.array([[0, 0, 0],[1, 1, 0],[0, 1, 0]])m2 = np.array([[0, 1, 1],[0, 0, 1],[0, 0, 0]])hit_list = []miss_list = []for k in range(4): #❷hit_list.append(np.rot90(h1, k))hit_list.append(np.rot90(h2, k))miss_list.append(np.rot90(m1, k))miss_list.append(np.rot90(m2, k))img = img.copy()while True:last = imgfor hit, miss in zip(hit_list, miss_list):hm = morphology.binary_hit_or_miss(img, hit, miss) #❸# 从图像中删除hit_or_miss所得到的白色点img = np.logical_and(img, np.logical_not(hm)) #❹# 如果处理之后的图像和处理前的图像相同,则结束处理if np.all(img == last): #❺breakreturn imga = pl.imread("scipy_morphology_demo2.png")[:,:,0].astype(np.uint8)b = skeletonize(a)#%hide_, (ax1, ax2) = pl.subplots(1, 2, figsize=(9, 3))ax1.imshow(a, cmap="gray", interpolation="nearest")ax2.imshow(b, cmap="gray", interpolation="nearest")ax1.set_axis_off()ax2.set_axis_off()pl.subplots_adjust(0.02, 0, 0.98, 1, 0.02, 0)

图像分割

squares = pl.imread("suqares.jpg")squares = (squares[:,:,0] < 200).astype(np.uint8)
from scipy.ndimage import morphologysquares_dt = morphology.distance_transform_cdt(squares)print ("各种距离值", np.unique(squares_dt))
各种距离值 [ 0123456789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27]
squares_core = (squares_dt > 8).astype(np.uint8)
from scipy.ndimage.measurements import label, center_of_massdef random_palette(labels, count, seed=1):np.random.seed(seed)palette = np.random.rand(count+1, 3)palette[0,:] = 0return palette[labels]labels, count = label(squares_core)h, w = labels.shapecenters = np.array(center_of_mass(labels, labels, index=range(1, count+1)), np.int)cores = random_palette(labels, count)
index = morphology.distance_transform_cdt(1-squares_core,return_distances=False,return_indices=True) #❶near_labels = labels[index[0], index[1]] #❷mask = (squares - squares_core).astype(bool)labels2 = labels.copy()labels2[mask] = near_labels[mask] #❸separated = random_palette(labels2, count)
#%figonly=矩形区域分割算法各个步骤的输出图像fig, axes = pl.subplots(2, 3, figsize=(7.5, 5.0), )fig.delaxes(axes[1, 2])axes[0, 0].imshow(squares, cmap="gray");axes[0, 1].imshow(squares_dt)axes[0, 2].imshow(squares_core, cmap="gray")ax = axes[1, 0]ax.imshow(cores)center_y, center_x = centers.Tax.plot(center_x, center_y, "o", color="white")ax.set_xlim(0, w)ax.set_ylim(h, 0)axes[1, 1].imshow(separated)for ax in axes.ravel():ax.axis("off")fig.subplots_adjust(wspace=0.01, hspace=0.01)

10.空间算法库-spatial

import numpy as npimport pylab as plfrom scipy import spatial
import matplotlib as mplmpl.rcParams['font.sans-serif'] = ['SimHei']

计算最近旁点

x = np.sort(np.random.rand(100))idx = np.searchsorted(x, 0.5)print (x[idx], x[idx - 1]) #距离0.5最近的数是这两个数中的一个
0.5244435681885733 0.4982156075770372
from scipy import spatialnp.random.seed(42)N = 100points = np.random.uniform(-1, 1, (N, 2))kd = spatial.cKDTree(points)targets = np.array([(0, 0), (0.5, 0.5), (-0.5, 0.5), (0.5, -0.5), (-0.5, -0.5)])dist, idx = kd.query(targets, 3)
r = 0.2idx2 = kd.query_ball_point(targets, r)idx2
array([list([48]), list([37, 78]), list([22, 79, 92]), list([6, 35, 58]), list([7, 42, 55, 83])], dtype=object)
idx3 = kd.query_pairs(0.1) - kd.query_pairs(0.08)idx3
{(1, 46), (3, 21), (3, 82), (3, 95), (5, 16), (9, 30), (10, 87), (11, 42), (11, 97), (18, 41), (29, 74), (32, 51), (37, 78), (39, 61), (41, 61), (50, 84), (55, 83), (73, 81)}
#%figonly=用cKDTree寻找近旁点x, y = points.Tcolors = "r", "b", "g", "y", "k"fig, (ax1, ax2, ax3) = pl.subplots(1, 3, figsize=(12, 4))for ax in ax1, ax2, ax3:ax.set_aspect("equal")ax.plot(x, y, "o", markersize=4)for ax in ax1, ax2:for i in range(len(targets)):c = colors[i]tx, ty = targets[i]ax.plot([tx], [ty], "*", markersize=10, color=c)for i in range(len(targets)):nx, ny = points[idx[i]].Tax1.plot(nx, ny, "o", markersize=10, markerfacecolor="None", markeredgecolor=colors[i], markeredgewidth=1)nx, ny = points[idx2[i]].Tax2.plot(nx, ny, "o", markersize=10, markerfacecolor="None", markeredgecolor=colors[i], markeredgewidth=1)ax2.add_artist(pl.Circle(targets[i], r, fill=None, linestyle="dashed"))for pidx1, pidx2 in idx3:sx, sy = points[pidx1]ex, ey = points[pidx2]ax3.plot([sx, ex], [sy, ey], "r", linewidth=2, alpha=0.6)ax1.set_xlabel(u"搜索最近的3个近旁点")ax2.set_xlabel(u"搜索距离在0.2之内的所有近旁点")ax3.set_xlabel(u"搜索所有距离在0.08到0.1之间的点对");

from scipy.spatial import distancedist1 = distance.squareform(distance.pdist(points))dist2 = distance.cdist(points, targets)print(dist1.shape)print(dist2.shape)
(100, 100)(100, 5)
print (dist[:, 0]) # cKDTree.query()返回的与targets最近的距离print (np.min(dist2, axis=0))
[0.15188266 0.09595807 0.05009422 0.11180181 0.19015485][0.15188266 0.09595807 0.05009422 0.11180181 0.19015485]
dist1[np.diag_indices(len(points))] = np.infnearest_pair = np.unravel_index(np.argmin(dist1), dist1.shape)print (nearest_pair, dist1[nearest_pair])
(22, 92) 0.005346210248158245
dist, idx = kd.query(points, 2)print (idx[np.argmin(dist[:, 1])], np.min(dist[:, 1]))
[22 92] 0.005346210248158245
N = 1000000start = np.random.uniform(0, 100, N)span = np.random.uniform(0.01, 1, N)span = np.clip(span, 2, 100)end = start + span
def naive_count_at(start, end, time):mask = (start  time)return np.sum(mask)
#%figonly=使用二维K-d树搜索指定区间的在线用户def _():N = 100start = np.random.uniform(0, 100, N)span = np.random.normal(40, 10, N)span = np.clip(span, 2, 100)end = start + spantime = 40fig, ax = pl.subplots(figsize=(8, 6))ax.scatter(start, end)mask = (start  time)start2, end2 = start[mask], end[mask]ax.scatter(start2, end2, marker="x", color="red")rect = pl.Rectangle((-20, 40), 60, 120, alpha=0.3)ax.add_patch(rect)ax.axhline(time, color="k", ls="--")ax.axvline(time, color="k", ls="--")ax.set_xlabel("Start")ax.set_ylabel("End")ax.set_xlim(-20, 120)ax.set_ylim(-20, 160)ax.plot([0, 120], [0, 120])_()

class KdSearch(object):def __init__(self, start, end, leafsize=10):self.tree = spatial.cKDTree(np.c_[start, end], leafsize=leafsize)self.max_time = np.max(end)def count_at(self, time):max_time = self.max_timeto_search = spatial.cKDTree([[time - max_time, time + max_time]])return self.tree.count_neighbors(to_search, max_time, p=np.inf)naive_count_at(start, end, 40) == KdSearch(start, end).count_at(40)
True

QUESTION

请读者研究点数Nleafsize参数与创建 K-d 树和搜索时间之间的关系。

凸包

np.random.seed(42)points2d = np.random.rand(10, 2)ch2d = spatial.ConvexHull(points2d)print(ch2d.simplices)print(ch2d.vertices)
[[2 5] [2 6] [0 5] [1 6] [1 0]][5 2 6 1 0]
#%fig=二维平面上的凸包poly = pl.Polygon(points2d[ch2d.vertices], fill=None, lw=2, color="r", alpha=0.5)ax = pl.subplot(aspect="equal")pl.plot(points2d[:, 0], points2d[:, 1], "go")for i, pos in enumerate(points2d):pl.text(pos[0], pos[1], str(i), color="blue")ax.add_artist(poly);

np.random.seed(42)points3d = np.random.rand(40, 3)ch3d = spatial.ConvexHull(points3d)ch3d.simplices.shape

(38, 3)

沃罗诺伊图

points2d = np.array([[0.2, 0.1], [0.5, 0.5], [0.8, 0.1], [0.5, 0.8], [0.3, 0.6], [0.7, 0.6], [0.5, 0.35]])vo = spatial.Voronoi(points2d)
print(vo.vertices); print(vo.regions); print(vo.ridge_vertices)
[[0.50.045 ] [0.2450.351 ] [0.7550.351 ] [0.3375 0.425 ] [0.6625 0.425 ] [0.45 0.65] [0.55 0.65]][[-1, 0, 1], [-1, 0, 2], [], [6, 4, 3, 5], [5, -1, 1, 3], [4, 2, 0, 1, 3], [6, -1, 2, 4], [6, -1, 5]][[-1, 0], [0, 1], [-1, 1], [0, 2], [-1, 2], [3, 5], [3, 4], [4, 6], [5, 6], [1, 3], [-1, 5], [2, 4], [-1, 6]]
bound = np.array([[-100, -100], [-100,100],[ 100,100], [ 100, -100]])vo2 = spatial.Voronoi(np.vstack((points2d, bound)))
#%figonly=沃罗诺伊图将空间分割为多个区域fig, (ax1, ax2) = pl.subplots(1, 2, figsize=(9, 4.5))ax1.set_aspect("equal")ax2.set_aspect("equal")spatial.voronoi_plot_2d(vo, ax=ax1)for i, v in enumerate(vo.vertices):ax1.text(v[0], v[1], str(i), color="red")for i, p in enumerate(points2d):ax1.text(p[0], p[1], str(i), color="blue")n = len(points2d)color = pl.cm.rainbow(np.linspace(0, 1, n))for i in range(n):idx = vo2.point_region[i]region = vo2.regions[idx]poly = pl.Polygon(vo2.vertices[region], facecolor=color[i], alpha=0.5, zorder=0)ax2.add_artist(poly)ax2.scatter(points2d[:, 0], points2d[:, 1], s=40, c=color, linewidths=2, edgecolors="k")ax2.plot(vo2.vertices[:, 0], vo2.vertices[:, 1], "ro", ms=6)for ax in ax1, ax2:ax.set_xlim(0, 1)ax.set_ylim(0, 1)

output_26_1
print (vo.point_region)print (vo.regions[6])
[0 3 1 7 4 6 5][6, -1, 2, 4]

德劳内三角化

x = np.array([46.445, 263.251, 174.176, 280.899, 280.899,189.358, 135.521, 29.638, 101.907, 226.665])y = np.array([287.865, 250.891, 287.865, 160.975, 54.252,160.975, 232.404, 179.187, 35.765, 71.361])points2d = np.c_[x, y]dy = spatial.Delaunay(points2d)vo = spatial.Voronoi(points2d)print(dy.simplices)print(vo.vertices)
[[8 5 7] [1 5 3] [5 6 7] [6 0 7] [0 6 2] [6 1 2] [1 6 5] [9 5 8] [4 9 8] [5 9 3] [9 4 3]][[104.58977484 127.03566055] [235.1285 198.68143374] [107.83960707 155.53682482] [ 71.22104881 228.39479887] [110.3105 291.17642838] [201.40695449 227.68436282] [201.61895891 226.21958623] [152.9623186493.25060083] [205.40381294 -90.5480267 ] [235.1285 127.45701644] [267.91709907 107.6135]]
#%fig=德劳内三角形的外接圆与圆心cx, cy = vo.vertices.Tax = pl.subplot(aspect="equal")spatial.delaunay_plot_2d(dy, ax=ax)ax.plot(cx, cy, "r.")for i, (cx, cy) in enumerate(vo.vertices):px, py = points2d[dy.simplices[i, 0]]radius = np.hypot(cx - px, cy - py)circle = pl.Circle((cx, cy), radius, fill=False, ls="dotted")ax.add_artist(circle)ax.set_xlim(0, 300)ax.set_ylim(0, 300);

往期精彩回顾适合初学者入门人工智能的路线及资料下载(图文+视频)机器学习入门系列下载机器学习及深度学习笔记等资料打印《统计学习方法》的代码复现专辑机器学习交流qq群955171419,加入微信群请扫码