求助：使用R构建含元素出现顺序位置的矩阵

Let's break down how to implement exactly what you're asking for. First, let's recap the requirements to make sure we're on the same page:

• We slide a window of size step=3 over your data (moving one row at a time)
• For each window, if any value in column2 is 1, we mark it as "Yes"; otherwise "No"
• For each unique character in column1, we fill in its position (1,2,3) if it's in the current window, else 0

Here's the complete code to achieve this:

# Load your sample data
sampledata <- data.frame(
  "column1" = c("a","b","c","d","e","a","b","e"), 
  "column2" = c(1,0,0,1,0,0,0,0)
)

# Define parameters
step <- 3
# Get all unique characters from column1 (add "f" here if you need it explicitly)
unique_chars <- unique(sampledata$column1)
# Number of sliding windows
n_windows <- nrow(sampledata) - step + 1

# Function to process each window
process_window <- function(window_start) {
  # Get the rows for this window
  window_rows <- window_start:(window_start + step - 1)
  
  # Check if the window has any 1 in column2
  has_yes <- any(sampledata$column2[window_rows] == 1)
  yes_no_marker <- ifelse(has_yes, "Yes", "No")
  
  # Extract characters in the current window
  window_chars <- sampledata$column1[window_rows]
  
  # Create values for each unique character: position if present, else 0
  position_values <- sapply(unique_chars, function(char) {
    pos <- which(window_chars == char)
    if (length(pos) > 0) pos else 0
  })
  
  # Combine position values with the marker
  c(position_values, Marker = yes_no_marker)
}

# Apply the function to all windows and convert to a data frame
result_matrix <- do.call(rbind, lapply(1:n_windows, process_window))
# Convert position columns to integers (they're characters initially)
result_matrix[, 1:length(unique_chars)] <- lapply(result_matrix[, 1:length(unique_chars)], as.integer)
# Name the rows for clarity
rownames(result_matrix) <- paste0("Window_", 1:n_windows)

# View the final result
print(result_matrix)

What this code does: #

1. Setup: We start by defining your sample data and parameters (window size step=3). We also extract all unique characters from column1 to use as columns in our result.
2. Window Processing Function: The process_window function handles each window:
   • It grabs the rows for the current window
   • Checks if there's any 1 in column2 to set the "Yes/No" marker
   • For each unique character, it finds its position in the window (1,2,3) or uses 0 if it's not present
3. Combine Results: We apply the function to every window, bind the results into a matrix, convert the position columns to integers, and add clear row names for each window.

Example Output: #

When you run this code, you'll get a matrix that looks like this:

a b c d e Marker
Window_1  1 2 3 0 0    Yes
Window_2  0 1 2 3 0    Yes
Window_3  0 0 1 2 3    Yes
Window_4  3 0 0 1 2    Yes
Window_5  2 3 0 0 1     No
Window_6  1 2 0 0 3     No

If you need to include "f" as a column even though it's not in your sample data, just modify the unique_chars line to:

unique_chars <- c(unique(sampledata$column1), "f")

This will add an extra column where all values are 0, as requested in your initial description.

内容的提问来源于stack exchange，提问作者user080391