如何基于另一数据框列名筛选数据框列并解决筛选后列数不一致问题？

Hey there! Let's break down why your column matching isn't getting the two data frames to have the same number of columns, and fix it up step by step.

1. Understand the Root Cause #

Your current code only filters met.kirp.se to keep columns that exist in exp.kirp.log2, but exp.kirp.log2 still retains all its original 290 columns. That's why you're seeing a discrepancy—met.kirp.se has 274 matching columns, but exp.kirp.log2 still has the extra 16 columns that aren't present in met.kirp.se.

2. Fix: Keep Only Common Columns in Both Data Frames #

To make both data frames have identical columns (and thus the same column count), you need to:

• First identify the common column names between the two data frames
• Subset both data frames to only use these common columns

Here's the code to do that:

# Step 1: Find all column names present in both data frames
common_columns <- intersect(colnames(met.kirp.se), colnames(exp.kirp.log2))

# Step 2: Subset both data frames to keep only these common columns
met.kirp.se <- met.kirp.se[, common_columns]
exp.kirp.log2 <- exp.kirp.log2[, common_columns]

# Verify the column counts match
ncol(met.kirp.se) == ncol(exp.kirp.log2) # Should return TRUE

3. Troubleshoot Hidden Mismatches (If the Above Doesn't Work) #

If you still see a mismatch, it's likely due to hidden differences in column names (like case sensitivity, extra spaces, or special characters). For example, "TCGA-12-3456" vs "tcga-12-3456" or "Sample 1" vs "Sample1" won't be matched by intersect() or %in%.

Fix this by standardizing your column names first:

# Standardize column names: convert to lowercase and trim extra spaces
colnames(met.kirp.se) <- tolower(trimws(colnames(met.kirp.se)))
colnames(exp.kirp.log2) <- tolower(trimws(colnames(exp.kirp.log2)))

# Now repeat the common column subsetting
common_columns <- intersect(colnames(met.kirp.se), colnames(exp.kirp.log2))
met.kirp.se <- met.kirp.se[, common_columns]
exp.kirp.log2 <- exp.kirp.log2[, common_columns]

4. Check Which Columns Are Missing (Optional) #

If you want to see exactly which columns are unique to exp.kirp.log2 (the 16 that aren't in met.kirp.se), run this:

exp_unique_cols <- setdiff(colnames(exp.kirp.log2), colnames(met.kirp.se))
print(exp_unique_cols) # Lists the 16 columns only present in exp.kirp.log2

内容的提问来源于stack exchange，提问作者melolilili