运行泊松问题 model

Question

I am trying to run a Poisson model in R. I checked a number of web sites to do this and I believe the code below should work.我正在尝试在 R 中运行 Poisson model。我检查了许多 web 站点来执行此操作，我相信下面的代码应该有效。 However, I am getting an error message that indicates that there is an unused argument.但是，我收到一条错误消息，指出存在未使用的参数。 I don't understand what the issue is.我不明白这是什么问题。

library(lme4)
m1<-lmer(sum ~ Lunch + intervention_y_n + (1 | prcid), 
         data=data_long_1, family = "poisson"(link = "log"))

Error in lmer(sum ~ Lunch + intervention_y_n + (1 | prcid), data = data_long_1, : unused argument (family = poisson(link = "log")) lmer 错误（sum ~ Lunch + intervention_y_n + (1 | prcid), data = data_long_1, : unused argument (family = poisson(link = "log"))

Answer 1

You need glmer() for generalized linear mixed models (eg a Poisson model).对于广义线性混合模型（例如泊松模型），您需要glmer() ) 。 lmer() is for linear mixed models only (Gaussian response with an identity link), so it doesn't have a family argument. lmer()仅适用于线性混合模型（具有身份链接的高斯响应），因此它没有family参数。

This code worked until fairly recently (4 months ago).此代码直到最近（4 个月前）才有效。 If you check the NEWS file for lme4 you'll see that version 1.1-22, which was released in April 2020, disabled the automatic forwarding of lmer(..., family=...) to glmer .如果您查看lme4 的新闻文件，您会看到 2020 年 4 月发布的版本 1.1-22 禁用了lmer(..., family=...)到glmer的自动转发。 However, this usage had been deprecated (ie, gave a warning message telling you not to use it) starting in 2013...但是，从 2013 年开始，这种用法已被弃用（即发出警告消息告诉您不要使用它）...

运行泊松问题 model

问题描述

1 个解决方案

解决方案1
2 已采纳 2020-09-01 23:29:43

运行泊松问题 model

问题描述

1 个解决方案

解决方案1 2 已采纳 2020-09-01 23:29:43

解决方案1
2 已采纳 2020-09-01 23:29:43