循环创建列表写入CSV时行元素数量不一致的技术求助

The Problem #

You're working with a multi-column CSV where each ID has 6 rows of data (20 IDs total, 120 rows). Your goal is to process the data so that for each ID, you calculate the median of columns 22-29, then replace values below the threshold (1e-6) with 0. However, your current code is producing a list with inconsistent row lengths, making it impossible to write to a CSV.

Here's the code you tried:

j=2 
for (each in lst) {   
    i=1   
    column =22   
    while(column<30){
        mlist[[i]]<-median(each[[column]])
        column=column+1
        i=i+1   
    }   
    z=22   
    for (cols in mlist) {
        if (cols>10^-6) {
            DF[[j]][[2]]<-each$SubjectID[[2]]
            DF[[j]][[z]]<-cols

        }else{
            DF[[j]][[2]]<-each$SubjectID[[2]]
            DF[[j]][[z]]<-0
        }
        z=z+1   
    } 
    j=j+1 
}

What's Causing the Issue? #

A few key problems in your code are leading to inconsistent row lengths:

• mlist isn't initialized with a fixed length, so its size can vary between iterations, throwing off the number of columns you're trying to fill
• DF[[j]] isn't pre-defined with a fixed column structure, so you're only populating parts of each row instead of ensuring every row has all required columns
• Manual nested loops are error-prone when trying to maintain consistent data structures

Solution 1: Using Tidyverse (Cleaner, More Reliable) #

The dplyr package makes this kind of grouped data processing straightforward, and it automatically ensures consistent output structures.

First, load the necessary package:

library(dplyr)

Assuming your raw data is stored in a data frame called raw_data, here's how to process it:

processed_data <- raw_data %>%
    # Group the data by SubjectID (each group has 6 rows)
    group_by(SubjectID) %>%
    # Calculate the median for columns 22 through 29
    summarise(across(22:29, ~median(.x)), .groups = "drop") %>%
    # Replace values below the threshold with 0
    mutate(across(22:29, ~ifelse(.x > 1e-6, .x, 0)))

Now you can write this directly to a CSV without any structure issues:

write.csv(processed_data, "your_output_file.csv", row.names = FALSE)

Solution 2: Base R (If You Prefer No External Packages) #

If you want to stick with base R, the key is to pre-initialize your output data frame with a fixed structure to avoid inconsistent row lengths.

# Assuming your raw data is in a data frame called raw_data
num_ids <- length(unique(raw_data$SubjectID))
col_names <- colnames(raw_data)

# Initialize an empty data frame with the correct number of rows and columns
DF <- data.frame(matrix(nrow = num_ids, ncol = length(col_names)))
colnames(DF) <- col_names

# Split the raw data into a list by SubjectID
lst <- split(raw_data, raw_data$SubjectID)

j <- 1
for (each in lst) {
    # Calculate medians for columns 22-29
    med_values <- sapply(each[22:29], median)
    # Apply threshold replacement
    med_values <- ifelse(med_values > 1e-6, med_values, 0)
    
    # Fill the current row in DF
    DF[j, "SubjectID"] <- unique(each$SubjectID)
    DF[j, 22:29] <- med_values
    
    j <- j + 1
}

# Write to CSV
write.csv(DF, "your_output_file.csv", row.names = FALSE)

Both approaches will ensure your output has consistent row lengths, making it easy to save as a CSV.

内容的提问来源于stack exchange，提问作者Zain Gill