图基公司生产的NaN

Question

我和图基有问题..。我真的不明白为什么..。

我尝试过几个代码，我也遇到了使用as.factor解决的因子变量的问题。但是with NaN产生了我不知道如何管理

我的数据：

> zz2
   id Mean group
1   1 1.17     1
2   2 1.57     2
3   3 1.16     2
4   4 1.14     1
5   5 1.81     4
6   6 1.33     1
7   7 1.27     1
8   8 1.57     1
9   9 1.56     3
10 10 1.36     1
11 11 1.09     1
12 12 1.05     1
13 13 1.43     3
14 14 1.48     2
15 15 1.60     3
16 16 1.78     4
17 17 1.55     1
18 18 1.36     2
19 19 1.06     3
20 20 1.50     2

这是我的代码：

install.packages("multcompView")

library(multcompView)

model=lm( zz2$Mean ~ as.factor(zz2$group))

ANOVA=aov(model)

TUKEY <- TukeyHSD(x=ANOVA, 'zz2$group', conf.level=0.95)

以下是错误消息：

> library(multcompView)
> model=lm( zz2$Mean ~ zz2$group )
> ANOVA=aov(model)
> TUKEY <- TukeyHSD(x=ANOVA, 'zz2$group', conf.level=0.95)
Error in TukeyHSD.aov(x = ANOVA, "zz2$group", conf.level = 0.95) : 
  nessun factor nel modello stimato
Inoltre: Warning message:
In replications(paste("~", xx), data = mf) : non-factors ignored: zz2$group
> model=lm( zz2$Mean ~ as.factor(zz2$group) )
> ANOVA=aov(model)
> TUKEY <- TukeyHSD(x=ANOVA, 'zz2$group', conf.level=0.95)
Warning message:
In qtukey(conf.level, length(means), x$df.residual) : NaNs produced

RStudio版本1.2.5033

我该怎么解决呢？谢谢!

Answer 1

zz2$group <- factor(zz2$group)

model=lm(Mean ~ group, data=zz2)

ANOVA=aov(model)

summary(ANOVA)

            Df Sum Sq Mean Sq F value Pr(>F)  
group        3 0.4396 0.14653   4.098 0.0246 *  # <- p-value (F-test)
Residuals   16 0.5721 0.03576

TUKEY <- TukeyHSD(x=ANOVA, "group", conf.level=0.95)
TUKEY

 Tukey multiple comparisons of means
    95% family-wise confidence level

Fit: aov(formula = model)

$group
          diff         lwr       upr     p adj
2-1  0.1328889 -0.16887514 0.4346529 0.5998626
3-1  0.1313889 -0.19372097 0.4564987 0.6615618
4-1  0.5138889  0.09095761 0.9368202 0.0148342
3-2 -0.0015000 -0.36442423 0.3614242 0.9999994
4-2  0.3810000 -0.07164605 0.8336460 0.1155457
4-3  0.3825000 -0.08603317 0.8510332 0.1312203