python数据分析-相关分析

大家好，今天来讲讲数据分析中的相关分析。相关性分析是指对两个或多个具备相关性的变量元素进行分析，从而衡量两个变量因素的相关密切程度。相关性的元素之间需要存在一定的联系或者概率才可以进行相关性分析。

本篇主要介绍相关分析中的典型相关分析。

两个随机变量X、Y的相关性可用它们的相关系数[插图]来度量。但在许多实际问题中，需要研究多个变量间的相关性。比如，在变量组（X1，X2，…，Xp）和（Y1，Y2，…，Yq）中，虽然每个Xi与每个Yj之间的相关性也反映了两组变量中各对数据之间的联系，但不能反映这两组变量整体之间的相关性，而且使用这么多相关系数来整体描述两组变量之间的相关性显得过于烦琐。

典型相关分析由霍特林提出，其基本思想和主成分分析非常相似：首先在每组变量中找出变量的线性组合，使两组的线性组合之间具有最大的相关系数；然后选取和最初挑选的这对线性组合不相关的线性组合，使其配对，并选取相关系数最大的一对，如此继续下去，直到两组变量之间的相关性被提取完毕为止。被选出的线性组合称为典型变量，它们的相关系数称为典型相关系数。典型相关系数度量了两组变量之间相关关系的强度。

使用scikit-learn库中的CCA类可实现典型相关分析：

sklearn.cross_decomposition.CCA(n_components=2,scale=Ture,
max_iter=500,tol=1e-06,copy=Ture)
# n_components:接受int,表示要保留的组分数。默认为2
# scale:接受bool,表示是否对数据进行测量。默认为Ture
# max_iter:接受int,表示指定最大迭代数。默认为500

实例：将scikit-learn库自带的iris数据集划分为两个分别含两个变量的数据集，并计算这两个数据集的典型相关系数。

from sklearn import datasets
from sklearn.cross_decomposition import CCA
iris = datasets.load_iris()  # 导入鸢尾花的数据集
iris_x = iris.data[:,0:2]  # 取样本数据前两个特征
iris_y = iris.data[:,2:4]  # 取样本数据后两个特征
cca = CCA()  # 定义一个典型相关分析对象
# 调用该对象的训练方法，主要接收两个参数：两个不同的数据集
cca.fit(iris_x, iris_y)
print('降维结果为：', cca.transform(iris_x, iris_y))  # 输出降维结果

输出结果：较长，略看

降维结果为： (array([[-1.22140847e+00,  4.95152798e-01],
       [-9.98912611e-01, -6.38966306e-01],
       [-1.39955712e+00, -3.40140243e-01],
       [-1.42183927e+00, -6.00140117e-01],
       [-1.42173073e+00,  6.44565829e-01],
       [-1.24358208e+00,  1.47985887e+00],
       [-1.68889943e+00,  1.39792400e-02],
       [-1.24369062e+00,  2.35152925e-01],
       [-1.46640357e+00, -1.12013986e+00],
       [-1.08793266e+00, -4.34259853e-01],
       [-1.06554197e+00,  1.07044597e+00],
       [-1.46629503e+00,  1.24566082e-01],
       [-1.11021481e+00, -6.94259727e-01],
       [-1.66672583e+00, -9.70726833e-01],
       [-8.87393324e-01,  1.90573901e+00],
       [-1.35477574e+00,  2.66927140e+00],
       [-1.24358208e+00,  1.47985887e+00],
       [-1.22140847e+00,  4.95152798e-01],
       [-8.20655420e-01,  1.44103268e+00],
       [-1.48846863e+00,  1.10927215e+00],
       [-7.98481813e-01,  4.56326609e-01],
       [-1.39944858e+00,  9.04565703e-01],
       [-1.86693954e+00,  4.23392145e-01],
       [-1.04336837e+00,  8.57398935e-02],
       [-1.46629503e+00,  1.24566082e-01],
       [-8.87610409e-01, -5.83672885e-01],
       [-1.24369062e+00,  2.35152925e-01],
       [-1.11010627e+00,  5.50446219e-01],
       [-1.02108622e+00,  3.45739767e-01],
       [-1.39955712e+00, -3.40140243e-01],
       [-1.19923487e+00, -4.89553275e-01],
       [-7.98481813e-01,  4.56326609e-01],
       [-1.64422659e+00,  1.77868493e+00],
       [-1.39934004e+00,  2.14927165e+00],
       [-1.08793266e+00, -4.34259853e-01],
       [-1.06565052e+00, -1.74259980e-01],
       [-7.76199664e-01,  7.16326483e-01],
       [-1.53303293e+00,  5.89272408e-01],
       [-1.55542362e+00, -9.15433412e-01],
       [-1.13238842e+00,  2.90446346e-01],
       [-1.33271068e+00,  4.39859377e-01],
       [-8.20981047e-01, -2.29308516e+00],
       [-1.73346373e+00, -5.06020507e-01],
       [-1.33271068e+00,  4.39859377e-01],
       [-1.48846863e+00,  1.10927215e+00],
       [-1.11021481e+00, -6.94259727e-01],
       [-1.48846863e+00,  1.10927215e+00],
       [-1.51085932e+00, -3.95433665e-01],
       [-1.17684418e+00,  1.01515255e+00],
       [-1.15467057e+00,  3.04464723e-02],
       [ 1.16039353e+00,  9.31608443e-01],
       [ 4.92580317e-01,  5.99847916e-01],
       [ 1.13811138e+00,  6.71608570e-01],
       [ 2.92040976e-01, -1.74015094e+00],
       [ 9.59962733e-01, -1.63684472e-01],
       [ 6.95451141e-02, -6.06031841e-01],
       [ 2.92258061e-01,  7.49260948e-01],
       [-4.64792291e-01, -1.86720502e+00],
       [ 9.82244881e-01,  9.63154019e-02],
       [-3.97945844e-01, -1.08720540e+00],
       [ 2.59012502e-03, -2.63073741e+00],
       [ 1.14109412e-01, -8.60320940e-02],
       [ 9.37572041e-01, -1.66839029e+00],
       [ 4.25733870e-01, -1.80151704e-01],
       [-1.30777142e-01, -4.56618810e-01],
       [ 9.15506977e-01,  5.61021728e-01],
       [-2.19797195e-01, -2.51912357e-01],
       [ 2.69867370e-01, -7.55444872e-01],
       [ 1.16017645e+00, -1.55780345e+00],
       [ 2.25303072e-01, -1.27544462e+00],
       [-6.39306949e-02,  3.23380811e-01],
       [ 5.14753923e-01, -3.84858156e-01],
       [ 1.00441849e+00, -8.88390671e-01],
       [ 5.14753923e-01, -3.84858156e-01],
       [ 7.59640477e-01, -1.42714404e-02],
       [ 8.93224828e-01,  3.01021854e-01],
       [ 1.29386934e+00,  2.19579198e-03],
       [ 1.00452703e+00,  3.56315275e-01],
       [ 3.14431668e-01, -2.35445125e-01],
       [ 2.47585221e-01, -1.01544475e+00],
       [ 2.03020923e-01, -1.53544449e+00],
       [ 2.03020923e-01, -1.53544449e+00],
       [ 2.69867370e-01, -7.55444872e-01],
       [ 4.92471774e-01, -6.44858030e-01],
       [-4.42401600e-01, -3.62499200e-01],
       [-1.30668599e-01,  7.88087136e-01],
       [ 9.15506977e-01,  5.61021728e-01],
       [ 1.18245859e+00, -1.29780358e+00],
       [-2.19797195e-01, -2.51912357e-01],
       [ 1.14000870e-01, -1.33073804e+00],
       [ 2.49808162e-02, -1.12603159e+00],
       [ 3.36713817e-01,  2.45547484e-02],
       [ 3.58887423e-01, -9.60151324e-01],
       [-2.64470035e-01, -2.01661805e+00],
       [ 4.72629652e-02, -8.66031714e-01],
       [-1.08494993e-01, -1.96618936e-01],
       [-1.94749393e-02, -4.01325389e-01],
       [ 5.37036072e-01, -1.24858283e-01],
       [-3.31207940e-01, -1.55191172e+00],
       [ 6.95451141e-02, -6.06031841e-01],
       [ 2.92258061e-01,  7.49260948e-01],
       [ 2.69867370e-01, -7.55444872e-01],
       [ 1.44973584e+00,  5.77488960e-01],
       [ 6.48338275e-01, -6.95648616e-02],
       [ 7.81922626e-01,  2.45728433e-01],
       [ 2.00624685e+00,  8.53956066e-01],
       [-5.53812344e-01, -1.66249857e+00],
       [ 1.76136030e+00,  4.83369350e-01],
       [ 1.44962730e+00, -6.67216986e-01],
       [ 1.02691772e+00,  1.86102109e+00],
       [ 6.03882519e-01,  6.55141338e-01],
       [ 9.37680584e-01, -4.23684345e-01],
       [ 1.11582923e+00,  4.11608697e-01],
       [ 3.36605274e-01, -1.22015120e+00],
       [ 1.80847316e-01, -5.50738420e-01],
       [ 4.92580317e-01,  5.99847916e-01],
       [ 7.81922626e-01,  2.45728433e-01],
       [ 1.40538863e+00,  2.54690111e+00],
       [ 2.47362927e+00,  9.04236779e-02],
       [ 9.37572041e-01, -1.66839029e+00],
       [ 1.04909133e+00,  8.76315022e-01],
       [-4.17570882e-02, -6.61325262e-01],
       [ 2.29558916e+00,  4.99836583e-01],
       [ 8.26378381e-01, -4.78977766e-01],
       [ 7.37466870e-01,  9.70434632e-01],
       [ 1.38299794e+00,  1.04219529e+00],
       [ 6.26056126e-01, -3.29564735e-01],
       [ 3.36713817e-01,  2.45547484e-02],
       [ 8.48660530e-01, -2.18977893e-01],
       [ 1.56103804e+00,  6.32782381e-01],
       [ 1.96168255e+00,  3.33956319e-01],
       [ 1.62799303e+00,  2.65748795e+00],
       [ 8.48660530e-01, -2.18977893e-01],
       [ 7.37358328e-01, -2.74271314e-01],
       [ 6.92794030e-01, -7.94271061e-01],
       [ 2.11754905e+00,  9.09249487e-01],
       [ 2.03238008e-01,  9.53967400e-01],
       [ 5.81600370e-01,  3.95141464e-01],
       [ 2.25411614e-01, -3.07386728e-02],
       [ 1.13811138e+00,  6.71608570e-01],
       [ 9.15506977e-01,  5.61021728e-01],
       [ 1.13811138e+00,  6.71608570e-01],
       [ 2.69867370e-01, -7.55444872e-01],
       [ 9.37789126e-01,  8.21021601e-01],
       [ 7.37466870e-01,  9.70434632e-01],
       [ 1.00452703e+00,  3.56315275e-01],
       [ 1.00441849e+00, -8.88390671e-01],
       [ 7.81922626e-01,  2.45728433e-01],
       [ 9.19358053e-02,  8.98673979e-01],
       [ 1.14109412e-01, -8.60320940e-02]]), 
       array([[-0.82345966, -0.12102114],
       [-0.82345966, -0.12102114],
       [-0.87688013, -0.03868193],
       [-0.7700392 , -0.20336036],
       [-0.82345966, -0.12102114],
       [-0.75049545,  0.04475237],
       [-0.86710826,  0.08537443],
       [-0.7700392 , -0.20336036],
       [-0.82345966, -0.12102114],
       [-0.7263906 , -0.40975593],
       [-0.7700392 , -0.20336036],
       [-0.71661873, -0.28569957],
       [-0.77981107, -0.32741672],
       [-0.94007247, -0.08039908],
       [-0.9303006 ,  0.04365728],
       [-0.85733639,  0.20943079],
       [-0.96417733,  0.37410921],
       [-0.86710826,  0.08537443],
       [-0.70684686, -0.16164321],
       [-0.81368779,  0.00303522],
       [-0.66319826, -0.36803878],
       [-0.85733639,  0.20943079],
       [-1.03714154,  0.2083357 ],
       [-0.79414405,  0.25114794],
       [-0.55635732, -0.5327172 ],
       [-0.71661873, -0.28569957],
       [-0.80391592,  0.12709158],
       [-0.7700392 , -0.20336036],
       [-0.82345966, -0.12102114],
       [-0.71661873, -0.28569957],
       [-0.71661873, -0.28569957],
       [-0.85733639,  0.20943079],
       [-0.7263906 , -0.40975593],
       [-0.82345966, -0.12102114],
       [-0.7700392 , -0.20336036],
       [-0.9303006 ,  0.04365728],
       [-0.87688013, -0.03868193],
       [-0.77981107, -0.32741672],
       [-0.87688013, -0.03868193],
       [-0.7700392 , -0.20336036],
       [-0.92052873,  0.16771364],
       [-0.92052873,  0.16771364],
       [-0.87688013, -0.03868193],
       [-0.89121312,  0.53988272],
       [-0.64365452, -0.11992606],
       [-0.86710826,  0.08537443],
       [-0.71661873, -0.28569957],
       [-0.82345966, -0.12102114],
       [-0.7700392 , -0.20336036],
       [-0.82345966, -0.12102114],
       [ 0.41563263, -0.36146826],
       [ 0.26514309,  0.00960574],
       [ 0.47882497, -0.31975111],
       [ 0.08533795,  0.00851065],
       [ 0.31856356, -0.07273347],
       [ 0.35244029, -0.40318541],
       [ 0.32833543,  0.05132289],
       [-0.15765954, -0.03430159],
       [ 0.40586076, -0.48552462],
       [-0.01173112,  0.29724544],
       [-0.0508186 , -0.19898001],
       [ 0.10488169,  0.25662337],
       [ 0.21628374, -0.61067607],
       [ 0.41563263, -0.36146826],
       [-0.12834393,  0.3378675 ],
       [ 0.25537122, -0.11445062],
       [ 0.26514309,  0.00960574],
       [ 0.26970421, -0.69301528],
       [ 0.26514309,  0.00960574],
       [ 0.11921467, -0.32194128],
       [ 0.29445871,  0.38177482],
       [ 0.08533795,  0.00851065],
       [ 0.47882497, -0.31975111],
       [ 0.50292982, -0.7742594 ],
       [ 0.24559935, -0.23850698],
       [ 0.25537122, -0.11445062],
       [ 0.4690531 , -0.44380747],
       [ 0.44494824,  0.01070082],
       [ 0.26514309,  0.00960574],
       [-0.0508186 , -0.19898001],
       [ 0.0657942 , -0.23960207],
       [ 0.05602233, -0.36365843],
       [ 0.07556607, -0.11554571],
       [ 0.54201731, -0.27803396],
       [ 0.26514309,  0.00960574],
       [ 0.2214945 ,  0.21600131],
       [ 0.37198403, -0.15507269],
       [ 0.29901982, -0.3208462 ],
       [ 0.13875841, -0.07382856],
       [ 0.08533795,  0.00851065],
       [ 0.34266842, -0.52724177],
       [ 0.36221216, -0.27912905],
       [ 0.12898654, -0.19788492],
       [-0.15765954, -0.03430159],
       [ 0.19217888, -0.15616777],
       [ 0.23582748, -0.36256335],
       [ 0.19217888, -0.15616777],
       [ 0.24559935, -0.23850698],
       [-0.36156954,  0.41911162],
       [ 0.13875841, -0.07382856],
       [ 0.62996415,  0.83847329],
       [ 0.41107152,  0.34115276],
       [ 0.75113807,  0.09523021],
       [ 0.72182245, -0.27693887],
       [ 0.654069  ,  0.38396499],
       [ 1.12508135, -0.48114427],
       [ 0.1778459 ,  0.42239688],
       [ 1.09576573, -0.85331336],
       [ 0.82866339, -0.4416173 ],
       [ 0.68338462,  0.75613408],
       [ 0.36742292,  0.54754833],
       [ 0.51791245,  0.17647433],
       [ 0.53745619,  0.42458706],
       [ 0.31400245,  0.62988754],
       [ 0.19282853,  1.37313062],
       [ 0.34331806,  1.00205663],
       [ 0.66840199, -0.19459966],
       [ 1.13485322, -0.35708791],
       [ 1.19804556, -0.31537076],
       [ 0.53224544, -0.40209032],
       [ 0.55699994,  0.67269978],
       [ 0.26058198,  0.71222675],
       [ 1.22215041, -0.76987906],
       [ 0.34787918,  0.29943561],
       [ 0.64429713,  0.25990863],
       [ 0.93550433, -0.60629572],
       [ 0.29445871,  0.38177482],
       [ 0.34787918,  0.29943561],
       [ 0.59087666,  0.34224784],
       [ 0.91596059, -0.85440844],
       [ 0.9452762 , -0.48223936],
       [ 1.06188901, -0.52286142],
       [ 0.54722807,  0.54864342],
       [ 0.5856659 , -0.48442953],
       [ 0.89641684, -1.10252117],
       [ 0.77068181,  0.34334293],
       [ 0.45993087,  0.96143456],
       [ 0.66840199, -0.19459966],
       [ 0.29445871,  0.38177482],
       [ 0.48403573,  0.50692627],
       [ 0.45993087,  0.96143456],
       [ 0.23647713,  1.16673505],
       [ 0.41107152,  0.34115276],
       [ 0.66384087,  0.50802135],
       [ 0.46970274,  1.08549092],
       [ 0.28989759,  1.08439584],
       [ 0.35765105,  0.42349197],
       [ 0.42084339,  0.46520912],
       [ 0.39673853,  0.91971741],
       [ 0.45472011,  0.13475719]]))