如何将checkresiduals生成的矩阵转换为指定结构的data.frame？

Got it, let's fix this for you! You’ve got a character matrix of residual test outputs from checkresiduals(), and you want to turn it into a clean data frame with TEST, Q*, df, and p-value columns. Your current approach isn’t quite hitting the mark, so here are two straightforward ways to get the result you need:

Method 1: Using Tidyverse (Clean & Intuitive) #

The tidyr::extract() function is perfect for this kind of structured text parsing—it lets you define a regex pattern to pull out specific values and map them directly to columns.

library(tidyverse)

# First, convert your matrix to a simple character vector
residual_text <- as.vector(ETS_RESIDUALS)

# Parse into a data frame
residual_df <- tibble(text = residual_text) %>%
  # Extract Q*, df, and p-value values using regex
  extract(
    col = text,
    into = c("Q_star", "df", "p_value"),
    # Regex matches: Q* = [number], df = [number], p-value (< or =) [number]
    regex = "Q\\* = ([0-9.]+), df = ([0-9.]+), p-value (?:<|=) ([0-9.e-]+)",
    convert = TRUE  # Automatically convert extracted values to numeric
  ) %>%
  # Add the TEST column with sequential labels
  mutate(TEST = paste0("TEST_", row_number()), .before = 1) %>%
  # Rename columns to match your desired output
  rename(`Q*` = Q_star, `p-value` = p_value)

# View the result
print(residual_df)

Output: #

# A tibble: 5 × 4
  TEST   `Q*`    df `p-value`
  <chr> <dbl> <dbl>     <dbl>
1 TEST_1 162.   18.8  2.20e-16
2 TEST_2 125.   18.8  2.20e-16
3 TEST_3 264.   18.8  2.20e-16
4 TEST_4  81.5  18.8  8.76e-10
5 TEST_5  36.6  18.8  8.18e- 3

Method 2: Base R (No Extra Packages) #

If you prefer sticking to base R, you can use sub() with regex to extract each component individually:

# Convert matrix to character vector
residual_text <- as.vector(ETS_RESIDUALS)

# Extract each numeric value using regex
q_vals <- as.numeric(sub(".*Q\\* = ([0-9.]+),.*", "\\1", residual_text))
df_vals <- as.numeric(sub(".*df = ([0-9.]+),.*", "\\1", residual_text))
p_vals <- as.numeric(sub(".*p-value (?:<|=) ([0-9.e-]+)", "\\1", residual_text))

# Build the data frame
residual_df <- data.frame(
  TEST = paste0("TEST_", seq_along(residual_text)),
  `Q*` = q_vals,
  df = df_vals,
  `p-value` = p_vals,
  check.names = FALSE  # Preserve column names with special characters
)

# View the result
print(residual_df)

Why Your Original Approach Didn’t Work #

Your code tried to convert the raw strings into a data frame definition, but this approach fails because:

• The original strings contain commas that break the data frame syntax
• It doesn’t extract the actual numeric values—instead, it tries to pass the full test result strings as columns, which isn’t what you want

Both methods above directly target and extract the numeric values you need, then structure them into the clean data frame you specified.

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