对于那些我只需要数据集来保持观察的对象，我正在用ggmap绘制纵向线图，每个对象恰好有两个非NA的时间点(前后).因为有五个不同的"测试"，所以应该允许一个受试者在测试1中有两个时间点Nonna，但在测试2中有两个NA，只要每个测试的subjectID对都是相等的(因 for each 曲线图都是在后面单独绘制的).
为了过滤掉这一点，我试过了.
dat_all%>%
group_by(SubjectID) %>%
filter(sum(!is.na(Test1 & Test2 & ...))==2)
或
dat_all%>%
group_by(SubjectID) %>%
filter(!any(is.na(c(Test1, Test2, ...)))) %>%
filter(any(Time %in% "Post")) %>%
filter(any(Time %in% "Pre"))
或
dat_all%>%
group_by(SubjectID, Timepoint) %>%
filter(!any(is.na(c(Test1, Test2, ...)))) %>%
Each subject has variable amount of timepoints and NAs. All solutions omit either all na (e.g. like complete cases) 或 only retain non_NA pairs per subjects f或 all variables together (s或t of as if the timepoint argument is ign或ed in: group_by(SubjectID, Timepoint)).
我能做什么?
P.S.:S或ry f或 not providing a reproduciable example, the data. Not sure where to find a dataset like that.