metaSEM中导入数据的标签/维度名称修改方法问询

Hey there! Let's walk through how to tweak study labels and dimension/variable names when importing correlation matrices into metaSEM. I’ll use a two-study example to make this super concrete.

1. First, Set Up Example Data #

Let’s start by simulating two correlation matrices (matching your use case) and loading the metaSEM package:

# Load the metaSEM package
library(metaSEM)

# Create correlation matrix for Study 1: Johnson et al (2010)
cor_mat1 <- matrix(c(1.00, 0.65, 0.58,
                     0.65, 1.00, 0.72,
                     0.58, 0.72, 1.00),
                   nrow = 3, ncol = 3)

# Create correlation matrix for Study 2: Smith et al (2012)
cor_mat2 <- matrix(c(1.00, 0.70, 0.62,
                     0.70, 1.00, 0.78,
                     0.62, 0.78, 1.00),
                   nrow = 3, ncol = 3)

# Sample sizes for each study
sample_sizes <- c(250, 320)

2. Import Data (Default Labels First) #

First, let’s import the data using createCor() to see what the default labels look like:

# Initial import with auto-generated labels
meta_obj <- createCor(cor_mat1, cor_mat2, n = sample_sizes)

# Check default output (you'll see Study 1/2 and V1/V2/V3)
print(meta_obj)

3. Rename Your Studies #

To replace the generic "Study 1" labels with your custom study names, directly update the studynames attribute of the metaSEM object:

# Assign custom study names
meta_obj@studynames <- c("Johnson et al (2010)", "Smith et al (2012)")

# Verify the change (optional but helpful)
cat("Updated study names:", meta_obj@studynames, "\n")

4. Rename Dimension/Variable Names #

Next, replace the default V1/V2/V3 labels with your actual variable names. Just update the varnames attribute:

# Assign custom variable/dimension names
meta_obj@varnames <- c("Anxiety", "Depression", "Stress")

# Verify the change
cat("Updated variable names:", meta_obj@varnames, "\n")

5. Confirm the Final Customized Data #

Print the modified object to make sure everything looks right:

# Check the fully customized metaSEM dataset
print(meta_obj)

Quick Notes to Avoid Headaches #

• Make sure the number of custom variable names matches the number of variables in your correlation matrices (e.g., 3 names for 3x3 matrices). Mismatched counts will throw errors.
• If you’re importing data from external files (like CSVs), just read the data into R first, then apply the same @studynames and @varnames updates—this workflow works regardless of how you load your raw data.
• For more than two studies, just extend the studynames vector with all your custom labels (e.g., c("Study A", "Study B", "Study C")).

内容的提问来源于stack exchange，提问作者Chris Wiese