RStudio中Stargazer生成描述统计表格报错：Error in if (is.na(s)) { : ...

错误根源 #

psych::describe()返回的数据框包含变量名、均值、标准差等多维度统计信息，直接转成矩阵传入stargazer()时，函数无法识别这种非标准输入结构，从而触发条件判断的长度错误。

直接解决方法 #

不需要提前用psych::describe()处理数据，直接把原始变量列传入stargazer()，让函数自行计算描述统计量：

# 直接基于原始数据生成Word格式描述统计表格
stargazer(data_corruption[c("innovation_firm", 
            "corruption", "ownership_gov", "competition", "gender_tm", 
            "size_firm", "age_firm", "RandD", "export", 
            "corruption_constraint")], 
          type = "html", 
          title="Descriptive statistics", 
          digits = 2, 
          out="Descriptive_statistics.doc")

自定义统计量的处理方案 #

如果需要psych::describe()提供的偏度、峰度等额外统计指标，先整理结果格式再传入stargazer()：

# 获取psych包的描述统计结果
desc_result <- psych::describe(data_corruption[c("innovation_firm", 
            "corruption", "ownership_gov", "competition", "gender_tm", 
            "size_firm", "age_firm", "RandD", "export", 
            "corruption_constraint")])

# 提取需要的统计列，转换为矩阵并设置行名
custom_desc_matrix <- as.matrix(desc_result[, c("mean", "sd", "median", "min", "max", "skew", "kurtosis")])
rownames(custom_desc_matrix) <- desc_result$vars

# 用整理好的矩阵生成表格，关闭自动统计计算
stargazer(custom_desc_matrix, 
          type = "html", 
          title="Descriptive statistics", 
          digits = 2, 
          out="Descriptive_statistics.doc",
          summary = FALSE)

内容的提问来源于stack exchange，提问作者Carla