变量‘group’已在‘linfct’中指定，但在‘model’中找不到！

Question

我的问题已经在以前的论坛上被问到了，但由于某种原因，答案对我来说并不起作用。我正在使用R对对数转换后的数据进行有计划的比较，并且继续得到错误:mcp2matrix(mcp2matrix，linfct = linfct)中的错误:变量‘group’已在‘linfct’中指定，但在‘model’中找不到！任何帮助都将不胜感激。谢谢!

以下是我的数据：

stress.data = read.table(textConnection(" group    rate   lnrate
1      P  69.169 4.236553
2      P  68.862 4.232105
3      C  84.738 4.439564
4      F  99.692 4.602085
5      C  87.231 4.468560
6      C  84.877 4.441203
7      P  70.169 4.250907
8      P  64.169 4.161520
9      P  58.692 4.072303
10     C  80.369 4.386629
11     C  91.754 4.519111
12     P  79.662 4.377793
13     C  87.446 4.471021
14     C  87.785 4.474891
15     P  69.231 4.237449
16     P  75.985 4.330536
17     F  91.354 4.514742
18     C  73.277 4.294247
19     F  83.400 4.423648
20     F 100.877 4.613902
21     C  84.523 4.437024
22     F 102.154 4.626481
23     C  77.800 4.354141
24     C  70.877 4.260946
25     P  86.446 4.459520
26     P  97.538 4.580242
27     F  89.815 4.497752
28     F  80.277 4.385483
29     P  85.000 4.442651
30     F  98.200 4.587006
31     C  90.015 4.499976
32     F 101.062 4.615734
33     F  76.908 4.342610
34     C  99.046 4.595584
35     F  97.046 4.575185
36     P  69.538 4.241873
37     C  75.477 4.323828
38     C  62.646 4.137500
39     P  70.077 4.249595
40     F  88.015 4.477507
41     F  81.600 4.401829
42     F  86.985 4.465736
43     F  92.492 4.527122
44     P  72.262 4.280298
45     P  65.446 4.181225"), header = TRUE)

library("multcomp")
stress.lm= lm(stress.data$lnrate ~ stress.data$group, data = stress.data)
stressPlanned= glht(stress.lm, linfct=mcp(group=c("C-P=0", "F-P=0")))

组中出错(mcp2matrix，linfct = linfct)：变量‘group’已在‘linfct’中指定，但在‘model’中找不到！

Answer 1

问题来自于在公式中使用$访问数据，而不是像预期的那样从data参数中获取数据。这通常是不鼓励的，因为它会混淆后续的函数，如您的示例中的predict()或glht()。只需使用：

stress.lm <- lm(lnrate ~ group, data = stress.data)
glht(stress.lm, linfct = mcp(group=c("C-P=0", "F-P=0")))
##   General Linear Hypotheses
## 
## Multiple Comparisons of Means: User-defined Contrasts
## 
## Linear Hypotheses:
##            Estimate
## C - P == 0   0.1180
## F - P == 0   0.2215