多元线性回归程序程序.docx

程序 1： data=[398,138,56.2058,12.1124,49.895,76.8624,228.80,98.205,5540.39; ...369,118,59.0433,9.3304,16.595,88.8056,177.45,224.953,5439.04; ...268,129,56.7236,28.7481,89.182,51.2972,166.40,263.032,4290.00; ...484,111,57.8627,12.8916,106.738,39.6473,258.05,320.928,5502.34; ...394,146,59.1178,13.3815,142.552,51.6517,209.30,406.989,4871.77; ...332,140,60.1113,11.0859,61.287,20.5476,180.05,246.996,4708.08; ...336,136,59.8398,24.9579,-30.385,40.1534,213.20,328.436,4627.81; ...383,104,60.0523,20.8096,-44.586,31.6456,200.85,298.456,4100.24; ...285,105,63.1415,8.4853,-28.373,12.4570,176.15,218.110,4122.69; ...277,135,62.3026,10.7301,75.723,68.3076,174.85,410.467,4842.25; ...456,128,64.9220,21.8743,144.030,52.4536,252.85,93.006,5740.65; ...355,131,64.8577,23.5062,112.904,76.6778,208.00,307.226,5094.10; ...364,120,63.5919,13.8940,128.347,96.0677,195.00,106.792,5383.20; ...32,147,65.6145,14.8659,10.097,47.9795,154.05,304.921,4888.17; ...311,143,67.0228,22.4940,-24.760,27.2319,180.70,59.612,4033.13; ...362,145,66.9049,23.3698,116.748,72.6681,219.70,238.986,4941.96; ...408,131,66.1843,13.0345,120.406,62.3129,234.65,141.074,5312.80; ...433,124,67.8651,8.0330,121.823,24.7122,258.05,290.832,5139.87; ...359,106,68.8892,27.0486,71.055,73.9123,196.30,413.636,4397.36; ...476,138,71.4177,18.2208,4.186,63.2737,278.85,206.454,5149.47; ...415,148,69.2775,7.7422,46.935,28.6762,207.35,79.566,5150.83; ...420,136,69.7334,10.1361,7.621,91.3633,213.20,428.982,4989.02; ...536,111,73.1628,27.3709,127.509,74.0169,296.40,273.072,5926.86; ...432,152,73.3650,15.5291,-49.574,16.1628,245.05,309.422,4703.88; ...436,123,73.0500,32.4918,100.098,42.9984,275.60,280.139,5356.59; ...415,119,74.9102,19.7127,-40.185,41.1346,211.25,314.548,4630.09; ...462,112,73.2007,14.8358,68.153,92.5180,282.75,212.058,5711.86; ...429,125,74.1615,11.3694,87.963,83.2870,217.75,118.065,5059.48; ...517,142,74.2838,26.7510,27.088,74.8921,306.80,344.553,6124.37; ...328,123,77.1409,19.6038,59.343,87.5103,210.60,140.872,4787.34; ...418,135,78.5940,34.6881,141.969,74.4712,269.75,82.855,5053.62; ...515,120,77.0938,23.2020,126.420,21.2711,328.25,398.425,5288.01; ...412,149,78.2313,35.7396,29.558,26.4941,258.05,124.027,4647.01; ...455,126,77.9296,21.5891,18.007,94.6311,232.70,117.911,5315.63; ...554,138,81.0394,19.5692,42.352,92.5448,323.70,161.250,6180.06; ...441,120,79.8485,15.5031,-21.558,50.0480,267.15,405.088,4800.97; ...417,120,80.6394,34.9238,148.450,83.1803,257.40,110.740,5512.13; ...461,132,82.2843,26.5496,-17.584,91.2214,266.50,170.392,5272.21]; % data(14,:)=[];% data(28,:)=[];% data(8,:)=[];n=38;m=8;alpha=0.05;y=data(:,9);x1=data(:,1);x2=data(:,2);x3=data(:,3);x4=data(:,4);x5=data(:,5);x6=data(:,6);x7=data(:,7);x8=data(:,8);X=[ones(n,1),x1,x2,x3,x4,x5,x6,x7,x8];[b,bint,r,rint,s]=regress(y,X,alpha);% b 回归系数% bint 回归系数的区间估计% r 残差% rint 残差置信区间% stats用于检验回归模型的统计量，有三个数值：相关系数R2、F值、与F对应的概率p, 相关系数 R2 越接近 1，说明回归方程越显著；%F > F1- a (k, n-k-1)时拒绝HO, F越大，说明回归方程越显著；与F对应的概率p时拒 绝H0,回归模型成立。

Y 为 n*1 的矩阵；% X 为(ones(n,1),x1xm)的矩阵；% alpha 显著性水平s2=sum(r42)/(n-m-1);b,bint,s,s2rcoplot(r,rint);%用这个图来来做参差及其置信区间的图，如果数据的置信区间不包含零点，则可认为这 个数据是异常的，应把它剔除调试结果：1) b =1.0e+003 *3.4130 -0.0009bint =1.0e+003 *2.06894.7571-0.00280.0010-0.00320.0108-0.0364-0.0014-0.0299-0.0043-0.00030.00290.00580.01320.00680.0162-0.00140.0003s =1.0e+004 *6.7454data(14,:)=[];'0.0001 0.0015 0.0000 s2 =6.7454e+004图形见(fulul)2)把语句'％ data(14,:)=[];',变为 n=38，变为 n=37结果：b =1.0e+003 *3.19340.0030bint =1.0e+003 *2.02304.36380.00000.0059-0.00290.0093-0.0352-0.0050-0.02230.00130.00010.00300.00520.01180.00070.0113-0.00130.0001s =1.0e+004 *0.0001 0.0022 0.0000 5.0297 s2 =5.0297e+004图形显示见(fulu2)3)把语句‘% data(14,:)=[];% data(28,:)=[];% data(8,:)=[]; '，变为‘ data(14,:)=[]; data(28,:)=[]; data(8,:)=[];' n=38，变为 n=35结果：b=1.0e+003 *3.96730.0034bint =1.0e+003 *2.81255.12210.00080.0060-0.00650.0053-0.0352-0.0071-0.02000.00180.00000.00270.00440.01030.00000.0094-0.0015-0.0001s =1.0e+004 *0.0001 0.0023 0.0000 3.8458 s2 =3.8458e+004图形显示见（fulu3）程序 2 data=[398,138,56.2058,12.1124,49.895,76.8624,228.80,98.205,5540.39; ...369,118,59.0433,9.3304,16.595,88.8056,177.45,224.953,5439.04; ...268,129,56.7236,28.7481,89.182,51.2972,166.40,263.032,4290.00; ...484,111,57.8627,12.8916,106.738,39.6473,258.05,320.928,5502.34; ...394,146,59.1178,13.3815,142.552,51.6517,209.30,406.989,4871.77; ...332,140,60.1113,11.0859,61.287,20.5476,180.05,246.996,4708.08; ...336,136,59.8398,24.9579,-30.385,40.1534,213.20,328.436,4627.81; ...383,104,60.0523,20.8096,-44.586,31.6456,200.85,298.456,4100.24; ...285,105,63.1415,8.4853,-28.373,12.4570,176.15,218.110,4122.69; ...277,135,62.3026,10.7301,75.723,68.3076,174.85,410.467,4842.25; ...456,128,64.9220,。