Plotting estimates (fixed effects) of a linear mixed-model

Question

Dan Hallatt 2019-7-25

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/473521-plotting-estimates-fixed-effects-of-a-linear-mixed-model

编辑： Dan Hallatt 2022-7-29

I am familiar with standard linear regression (non-nested) where, after employing fitresult = fitlm(domain,range,'poly1'), you can use plot(fitresult) to plot the fitted linear line with also dotted lines for the confidence bounds of that fit over the domain specified within the fit.

I am now interested in fitting the data to a linear mixed-effects model, by using fitresult = fitlme(HLM_Table,'response ~ predictor + (1|cluster)'), but I cannot find any documentation about plotting the fitted linear regression (fixed coefficients). I have found equivalent documentaruon for plotting residuals and the like, but nothing at all about plotting the actual fitted model. I could do this manually by extracting the fitted fixed coefficients and plotting a line with these over the respective domain, but I am wondering if there is a built in function such as the plot(fitresult) that I used with standard linear regression?

2 个评论
显示无隐藏无

Lea Kerciku 2022-6-19

Hi, could you share your solution on how you plotted the fitted model in the end? My statistics are not so strong and I am having the same issue. Thank you.

Dan Hallatt 2022-7-29

编辑：Dan Hallatt 2022-7-29

Hi Lea, I'm quite sorry.. I'm pretty far removed from this and it would honestly take quite a bit of work to dig up this work years ago.

If I have some time in the next while I can post a solution, but I doubt I'll be able to help.

I hope you've found a solution in the meantime!

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

the cyclist 2019-7-25

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/473521-plotting-estimates-fixed-effects-of-a-linear-mixed-model#answer_384861

编辑：the cyclist 2020-1-7

I could find no mention in the documentation of the LinearMixedModel class that there is an equivalent object function or method. I think you'll have to construct it.