在Alteryx中使用R时遇tibble列必须为向量错误求解决

我在Alteryx中使用R进行统计分析时，运行代码出现如下错误：

! All Columns in a tibble must be vectors.

完整代码如下：

library("tibble")

# Calling Data from Connection #1
data <- read.Alteryx("#1")

average_wilcox <- c("1","1","1","1","1","1","1")

# Creating data frame for in case it comes an empty table
df <- data.frame(average_wilcox)

#Verify if p-value is empty
# In case is different that empty, executes the steps for the Hypothesis Test for non-normal data
if (length(data$p.value) == 0) {
write.Alteryx(df, 1)
} else if (data$p.value != '') {
Week1 <- read.Alteryx("#2", mode="data.frame")
"&"
Week2 <- read.Alteryx("#3", mode="data.frame")

# MANN WHITNEY TEST (AVERAGE TEST FOR NON NORMAL)

Week1_data <- Week1$Wk1_feature_value
Week2_data <- Week2$Wk2_feature_value

# DEFINE VECTORS
week1 <- c(Week1_data)
week2 <- c(Week2_data)

merge(cbind(Week1, X=1:length(week1)),
      cbind(Week2, X=1:length(week2)), all.y =T) [-1]

# MANN WHITNEY TEST (MEAN TEST FOR NON NORMAL)
average_wilcox <- wilcox.test(week1,week2, alternative='two.sided', conf.level=.95)

average_test <- tibble(average_wilcox) 

average_test[] <- lapply(average_test, as.character)

write.Alteryx(average_test, 1)
}

#### NORMAL HYPOTHESIS TEST ####

# Calling Data from Connection #4
data1 <- read.Alteryx("#4")
df1 <- data.frame(Date=as.Date(character()),"p.value"=character(),User=character(),stringsAsFactors=FALSE)

# Verify if p-value is empty
# In case if different than empty, executes the steps for the Hypothesis Test for normal data
if(length(data1$p.value) == 0) { 
write.Alteryx(df1, 3)
} else if (data1$p.value != '') { 
Week1 <- read.Alteryx("#2", mode="data.frame") 
"&"
Week2 <- read.Alteryx("#3", mode="data.frame") 

# T TEST (MEAN TEST FOR NORMAL)

Week1_data <- Week1$Wk1_feature_value 
Week2_data <- Week2$Wk2_feature_value 

# DEFINE VECTORS
week1 <- c(Week1_data) 
week2 <- c(Week2_data) 

# T TEST (MEAN TEST FOR NORMAL)
t_test <- t.test(week1,week2, alternative='two.sided',conf.level=.95) 

write.Alteryx(t_test,3)
}

错误原因及解决方法 #

核心错误点

错误出在average_test <- tibble(average_wilcox)这一行：wilcox.test()返回的是一个列表对象，包含统计量、p值、置信区间等多种结构的数据，而tibble()要求每一列必须是向量，直接将整个列表传入就会触发报错。

同时代码里还有两处无效/冗余内容：

1. 单独的"&"语句无任何实际作用，建议删除；
2. merge(cbind(Week1, X=1:length(week1)), cbind(Week2, X=1:length(week2)), all.y =T) [-1]计算后未赋值给任何变量，属于无效执行，可根据需求删除或保存结果。

修正后的关键代码

1. Wilcoxon检验部分修正

将检验结果提取为单独字段后再构建tibble：

# MANN WHITNEY TEST (MEAN TEST FOR NON NORMAL)
average_wilcox <- wilcox.test(week1,week2, alternative='two.sided', conf.level=.95)

# 提取测试结果中的关键字段，构建规范的tibble
average_test <- tibble(
  statistic = average_wilcox$statistic,
  p_value = average_wilcox$p.value,
  conf_interval_low = average_wilcox$conf.int[1],
  conf_interval_high = average_wilcox$conf.int[2],
  alternative_hypothesis = average_wilcox$alternative,
  test_method = average_wilcox$method,
  data_source = average_wilcox$data.name
)

# 按需转换为字符型
average_test[] <- lapply(average_test, as.character)

write.Alteryx(average_test, 1)

2. t检验部分修正

t.test()同样返回列表，直接传入write.Alteryx可能出问题，建议转为数据框后输出：

# T TEST (MEAN TEST FOR NORMAL)
t_test <- t.test(week1,week2, alternative='two.sided',conf.level=.95) 

# 将列表结果转换为数据框
t_test_df <- as.data.frame(unlist(t_test))
colnames(t_test_df) <- "value"
write.Alteryx(t_test_df,3)

额外优化建议

• data$p.value != ''的判断逻辑有问题：如果p.value是数值型，和空字符串比较会引发类型错误，建议改为!is.na(data$p.value) && length(data$p.value) > 0；
• week1 <- c(Week1_data)如果Week1_data已经是向量，此行冗余，可直接使用Week1_data。

内容的提问来源于stack exchange，提问作者Wil