Problem when plotting the figure

Question

yusra Ch 2020-9-6

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/589711-problem-when-plotting-the-figure

评论： Star Strider 2020-9-6

I have a vector called K [622x1] double I plot the PDF of K . I fitted my K vector to the Gaussian mixture distribution and ploted the PDF with a solid line :

>> GMModel = fitgmdist(k,2);
>> gm=gmdistribution(GMModel.mu,GMModel.Sigma);
>> pdfk=pdf(gm,k);
>> figure
>> plot(k,pdfk)

The plot has several lines and I dont know what is the problem and how to resolve it. Ca anyone help me with this? thank youuu

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

Star Strider 2020-9-6

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/589711-problem-when-plotting-the-figure#answer_490624

在 MATLAB Online 中打开

I do not have ‘k’, so I created my own, and was able to reproduce essentially the result you got.

The multiple lines disappear of you sort ‘k’ first:

k = sum([randn(622,1)+5; randn(622,1)+1],2);
k = sort(k);                                                % <— ADD ‘sort’ CALL
GMModel = fitgmdist(k,2);
gm=gmdistribution(GMModel.mu,GMModel.Sigma);
pdfk=pdf(gm,k);
figure
plot(k,pdfk)

.

4 个评论
显示 2更早的评论隐藏 2更早的评论

yusra Ch 2020-9-6

I have one more question, How to draw the PDF of the vector K ?

I have plotted the PDF fitted to the mixture Gaussian distribution and I want to compare the two plots.

Star Strider 2020-9-6

在 MATLAB Online 中打开

I would do something like this (starting with my original code):

k = sum([randn(622,1)+5; 0.5*randn(622,1)+1],2);
k = sort(k);                                                % <— ADD ‘sort’ CALL
GMModel = fitgmdist(k,2);
gm=gmdistribution(GMModel.mu,GMModel.Sigma);
pdfk=pdf(gm,k);
[f,x] = ecdf(k);                                            % Empirical Cumulative Distribution Function
[fr,xr] = resample(f, x);                                   % Resample To Constant Sampling Interval
df = gradient(fr)./gradient(xr);                            % Calculate Derivative To Get PDF
dfs = smoothdata(df, 'gaussian', 150);                      % Smooth To Eliminate Noise
figure
plot(k,pdfk)
hold on
plot(xr, dfs)
hold off

This appears to produce a reasonable approximation. Experiment with the smoothdata function (introduced in R2017a) with your data to get different results. (I do not kinow what MATLAB version you are using. Other options, such as movmedian to do the smoothing were introduced in R2016a, and of course there are still other options.)

.

请先登录，再进行评论。