二项式重复数据中的glmmadaptive

发布于 2025-02-05 05:58:22 字数 789 浏览 4 评论 0原文

我正在使用R中的GLMMADAPTIVE进行重复的二项式数据

。 。

​所以我运行了这个模型。我的模型是:固定效果(时间*治疗+持续时间),随机效应仅具有随机截距。

fm <- mixed_model(fixed = Outcome ~ Time * Treatment + Duration,
                  random = ~ 1 | ID,
                  data = data2,
                  family = binomial(),
                  nAGQ = 15,
                  iter_EM=0,
                  max_coef_value = 50,
                  initial_values = list(betas = rep(0, 5)),
                  na.action=na.exclude)

这样做有警告:

> Hessian matrix at convergence is not positive value: unstable solution.

然后我使用摘要(FM),它在我的SE和P值中显示nas

有人知道如何解决吗?我被卡住了1个月以上.......

I am doing repeated binomial data with GLMMadaptive in R. Outcomes were collected at 3 different time points(collected at baseline, 12 weeks, 24 weeks)and data looks like:

original data set1

I want to calculate the marginal probability of this outcome, and I have missing data (MAR). So I run this model. My model is: fixed effect(time*Treatment+Duration), Random effect only has random intercept.

fm <- mixed_model(fixed = Outcome ~ Time * Treatment + Duration,
                  random = ~ 1 | ID,
                  data = data2,
                  family = binomial(),
                  nAGQ = 15,
                  iter_EM=0,
                  max_coef_value = 50,
                  initial_values = list(betas = rep(0, 5)),
                  na.action=na.exclude)

There is a warning by doing this:

> Hessian matrix at convergence is not positive value: unstable solution.

And then I use summary(fm), which showed NAs in my SE and P-value.

Is anyone know how to fix it? I have been stuck for more than 1 month.......

如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

扫码二维码加入Web技术交流群

发布评论

需要 登录 才能够评论, 你可以免费 注册 一个本站的账号。
列表为空,暂无数据
我们使用 Cookies 和其他技术来定制您的体验包括您的登录状态等。通过阅读我们的 隐私政策 了解更多相关信息。 单击 接受 或继续使用网站,即表示您同意使用 Cookies 和您的相关数据。
原文