PDF data analysis and comparison

Question

harley 2013-9-11

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/86850-pdf-data-analysis-and-comparison

i have some theoretical PDF's. I want Matlab to decide which one is the best fit for my experimental PDF. How would i do this.

%x_int, y_int is EXPERIMENTAL PDF.
%NORMAL DISTRIBUTION F(x) VS EXPERIMENTAL PDF P(x)
plot(x,p_nor,x_int,y_int);
%RAYLEIGH PDF F(x) VS EXPERIMENTAL PDF P(x)
plot(x,p_ray,x_int,y_int)
%POISSONS PDF F(x)VS EXPERIMENTAL PDF P(x)
plot(x,p_poi,x_int,y_int)

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2013-9-11

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/86850-pdf-data-analysis-and-comparison#answer_96397

Is it intended that the theoretical distributions already have the correct parameters (mean, standard deviation) for the experimental distribution? If so then I do not know what the recommendation would be.

I know that one approach would be to generate a number of values in each theoretical distribution, and then use kstest2() to compare y_int to each generated vector.

If the parameters are not known for the theoretical distribution, then:

For the specific case of determining whether y_int comes from some normal distribution, see lillietest(); alternately adtest() or jbtest().

I seem to remember having run across a test for poison distribution before, but I do not recall it. I do not recall seeing a test for rayleigh distribution, but I do not work with stats much.