使用R语言intsvy包合并PISA 2015数据时遇重复标签错误求助

问题描述 #

我用intsvy包分析PISA数据，尝试通过pisa.select.merge()合并2015年学生文件与学校文件时，突然出现「Duplicate labels」错误。这段代码过去两个月一直正常运行，如今却报错。我原本以为合并功能可识别并合并带有相似标签的两个数据集，已尝试删除并重新下载原始PISA数据文件，也卸载重装了intsvy包与RStudio，但问题仍未解决。

代码示例 #

library(intsvy)

PISA2015 <- pisa.select.merge(folder = "/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets",
                              school.file = "2015_SCHQ.sav",
                              student.file = "2015_STUQ1.sav",
                              student = c("ESCS", "PARED"), 
                              school = c("CLSIZE", "SCHSIZE"),
                              countries = c("PRT"))

报错信息 #

File character set is 'WINDOWS-1252'.
Converting character set to UTF-8.
File character set is 'WINDOWS-1252'.
Converting character set to UTF-8.
Error in as.factor(x) : Duplicate labels
In addition: Warning messages:
  1: 11 variables have duplicated labels:
  CNTRYID, Region, STRATUM, SUBNATIO, ST011D17TA, ST011D18TA,
ST011D19TA, PROGN, OCOD1, OCOD2, OCOD3 
2: 4 variables have duplicated labels:
  CNTRYID, Region, STRATUM, SUBNATIO 

解决方案 #

1. 手动读取并合并数据，绕开intsvy的自动标签处理 #

从警告可知，学生和学校数据集存在重复标签的变量（如CNTRYID、Region），这些变量可能在两个文件中标签文本重复，导致intsvy的合并函数触发错误。可以手动读取数据并筛选变量后合并：

library(foreign)

# 分别读取学生和学校数据，不自动转换为带标签的因子
student_data <- read.spss(
  "/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets/2015_STUQ1.sav",
  use.value.labels = FALSE,
  to.data.frame = TRUE,
  reencode = "UTF-8"
)
school_data <- read.spss(
  "/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets/2015_SCHQ.sav",
  use.value.labels = FALSE,
  to.data.frame = TRUE,
  reencode = "UTF-8"
)

# 筛选需要的变量，保留合并键（CNTRYID、SCHOOLID）
student_subset <- student_data[, c("CNTRYID", "SCHOOLID", "ESCS", "PARED")]
school_subset <- school_data[, c("CNTRYID", "SCHOOLID", "CLSIZE", "SCHSIZE")]

# 手动合并数据集
PISA2015 <- merge(student_subset, school_subset, by = c("CNTRYID", "SCHOOLID"), all.x = TRUE)

合并完成后，可继续用intsvy的其他函数处理数据（如设置分析权重）。

2. 修复数据中的重复标签后再调用intsvy合并函数 #

用haven包读取SPSS文件，修复重复的因子标签后保存，再用pisa.select.merge()读取合并：

library(haven)

# 读取原始数据
student_data <- read_sav("/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets/2015_STUQ1.sav")
school_data <- read_sav("/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets/2015_SCHQ.sav")

# 修复学生数据中的重复因子标签
for (var in names(student_data)) {
  if (is.factor(student_data[[var]])) {
    levels(student_data[[var]]) <- make.unique(levels(student_data[[var]]))
  }
}

# 修复学校数据中的重复因子标签
for (var in names(school_data)) {
  if (is.factor(school_data[[var]])) {
    levels(school_data[[var]]) <- make.unique(levels(school_data[[var]]))
  }
}

# 保存处理后的文件
write_sav(student_data, "/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets/2015_STUQ1_clean.sav")
write_sav(school_data, "/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets/2015_SCHQ_clean.sav")

# 调用合并函数
PISA2015 <- pisa.select.merge(folder = "/Users/x/Desktop/x/EPICER/Analysis/R Script and Supporting Datasets",
                              school.file = "2015_SCHQ_clean.sav",
                              student.file = "2015_STUQ1_clean.sav",
                              student = c("ESCS", "PARED"), 
                              school = c("CLSIZE", "SCHSIZE"),
                              countries = c("PRT"))

3. 回退到intsvy旧版本 #

若问题是新版本intsvy对标签处理更严格导致的，可尝试安装旧版本的包：

# 先安装devtools（如果没装过）
install.packages("devtools")
library(devtools)

# 安装指定版本的intsvy，这里以2.3为例（根据你之前正常运行时的版本调整）
install_version("intsvy", version = "2.3")

内容的提问来源于stack exchange，提问作者Daniel