Fitting Bimodal Data to Histogram

Question

Sophie Lis 2018-6-5

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/404170-fitting-bimodal-data-to-histogram

I am having trouble fitting the below bimodal data, which I plotted on a histogram. I am using the suggested MathWorks code to fit bivariate distributions (below), but for some reason, when I plot probability density on the yaxis, it appear to the 10^4 power, while the maximum value of the original histogram is 14, so I cannot overlay the two.

min_pts =

1.0e-04 *

0.0341 -0.1578 -0.1203

0.0198 -0.1459 -0.1284

0.0014 -0.1314 -0.1343

-0.0106 -0.1208 -0.1378

-0.0157 -0.1243 -0.1385

-0.0146 -0.1235 -0.1360

-0.0085 -0.1184 -0.1302

0.0015 -0.1092 -0.1216

0.0140 -0.0964 -0.1355

0.0278 -0.0807 -0.1465

0.0417 -0.0698 -0.1551

0.0554 -0.0655 -0.1618

0.0687 -0.0601 -0.1673

0.0816 -0.0550 -0.1721

0.0937 -0.0535 -0.1764

0.0860 -0.0726 -0.1803

0.0746 -0.1059 -0.1838

0.0443 -0.1343 -0.2418

0.0175 -0.1572 -0.2958

-0.0059 -0.1753 -0.3404

-0.0267 -0.1894 -0.3754

-0.0455 -0.2006 -0.4009

-0.0646 -0.2097 -0.4175

-0.0860 -0.2176 -0.4259

-0.1122 -0.2252 -0.4271

-0.1635 -0.2325 -0.4222

-0.2402 -0.2406 -0.4124

-0.3529 -0.2804 -0.3989

-0.4898 -0.3394 -0.3829

-0.6217 -0.4144 -0.4013

-0.7379 -0.4785 -0.4637

-0.8333 -0.5280 -0.5109

-0.9040 -0.5600 -0.5454

-0.9467 -0.5744 -0.5877

-0.9629 -0.5723 -0.6217

-0.9552 -0.5548 -0.6455

-0.9265 -0.5251 -0.6580

-0.9125 -0.4869 -0.6580

-0.8938 -0.4436 -0.6452

-0.8624 -0.3977 -0.6207

-0.8205 -0.3534 -0.5859

-0.7651 -0.3244 -0.5425

-0.7324 -0.2901 -0.4927

-0.7186 -0.2511 -0.4389

-0.6929 -0.2083 -0.3833

-0.6558 -0.1624 -0.3274

-0.6081 -0.1434 -0.3061

-0.5509 -0.1350 -0.3289

-0.4862 -0.1283 -0.3494

-0.4163 -0.1228 -0.3659

-0.4981 -0.1175 -0.3768

-0.6118 -0.1146 -0.3808

-0.7117 -0.1273 -0.3766

-0.7933 -0.1395 -0.3633

-0.8527 -0.1499 -0.3409

-0.8893 -0.1573 -0.3095

-0.9035 -0.1608 -0.2699

-0.8962 -0.1593 -0.2241

-0.8706 -0.1520 -0.1857

-0.8303 -0.1394 -0.2045

-0.7791 -0.1221 -0.2232

-0.7207 -0.1010 -0.2405

-0.6610 -0.0774 -0.2552

-0.6147 -0.0663 -0.2662

-0.5666 -0.0562 -0.2724

-0.5184 -0.0455 -0.2727

-0.4712 -0.0417 -0.2670

-0.4261 -0.0796 -0.2551

-0.3832 -0.1217 -0.2371

-0.3425 -0.1664 -0.2138

-0.3041 -0.2118 -0.1923

-0.2677 -0.2561 -0.2120

-0.2332 -0.2973 -0.2276

-0.2003 -0.3335 -0.2378

-0.1686 -0.3632 -0.2417

-0.1377 -0.3855 -0.2384

-0.1071 -0.3998 -0.2282

-0.0769 -0.4057 -0.2114

-0.0469 -0.4033 -0.1889

-0.0175 -0.3931 -0.1620

0.0106 -0.3755 -0.1446

0.0364 -0.3510 -0.1289

0.0252 -0.3204 -0.1119

-0.0195 -0.2842 -0.1305

-0.0644 -0.2435 -0.1471

-0.1089 -0.1992 -0.1612

-0.1521 -0.1523 -0.1726

-0.1932 -0.1620 -0.1811

-0.2311 -0.1738 -0.1865

-0.2646 -0.1840 -0.1892

-0.2927 -0.1924 -0.1897

-0.3141 -0.1986 -0.1885

-0.3280 -0.2023 -0.1864

-0.3335 -0.2031 -0.1900

-0.3300 -0.2010 -0.2451

-0.3174 -0.2063 -0.2989

-0.2960 -0.2274 -0.3493

        x = min_pts(:);
        figure;
        hist(x,100);
        pdf_normmixture = @(x,p,mu1,mu2,sigma1,sigma2) ...
                         p*normpdf(x,mu1,sigma1) + (1-p)*normpdf(x,mu2,sigma2);
        pStart = .5;
        muStart = quantile(x,[.25 .75]);
        sigmaStart = sqrt(var(x) - .25*diff(muStart).^2);
        start = [pStart muStart sigmaStart sigmaStart];
        lb = [0 -Inf -Inf 0 0];
        ub = [1 Inf Inf Inf Inf];
        paramEsts = mle(x, 'pdf',pdf_normmixture, 'start',start, ...
                          'lower',lb, 'upper',ub);
       bins = 100;