GLM fitting, testing and graphics plotting

1 次查看（过去 30 天）
Ursula Perez-Ramirez
Hello,
I would be very grateful if you could help me with some doubts regarding GLM. My data has two groups (controls and cases) and I want to estimate how the gray matter volume is related to days of drug abstinence, after correcting for age, in both groups (controls, cases). I imagine this like two fitting lines (one for controls, one for cases) y-axis: gray matter, x-axes:days of abstinence (but corrected for age) and testing the slopes.
I have started this simple code, but I dont't know how to code the groups (0, 1 or 1 - 1) or whether I should demean age and abstinence. Also, I would like to plot a graphic with two lines to see how the data is fitted (after correcting for age). How could I do that?
GM =[2;5;8;4]; % Gray matter
group=[0;0;1;1]; % group (0:controls, 1:cases)
age=[45;20;15;58]; % age (years old)
days_abst=[5;8;7;6]; % days of abstinence (behavior value to test)
d = dataset(GM,group,age,days_abst);
d.group=ordinal(d.group);
m = GeneralizedLinearModel.fit(d,'GM~group+age+days_abst');
devianceTest(m);
coefTest(m,[0 0 0 1]) %? Test if the relationship between GM and days_abst is %significantly different in controls and cases
Thank you very much,
Ursula

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Vector Fields 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by