如何在R中按两列因子重排DataFrame？reshape包使用遇阻

Hey there! Let’s tackle this DataFrame reshaping problem you’re facing. Since the reshape package didn’t give you the results you wanted, let’s use the more flexible tidyr package (part of the tidyverse ecosystem) which handles most common reshaping scenarios with ease.

First, let’s cover the two most frequent reshaping needs, with concrete examples you can adapt to your data:

1. Converting Long Data to Wide Data (Factor Groups as Columns) #

If your original data is in "long" format (one row per observation with a factor column), and you want to pivot it so each factor becomes its own column:

# Load the tidyverse (includes tidyr)
library(tidyverse)

# Example original long-format DataFrame
original_df <- tibble(
  observation_id = rep(1:3, each = 2),
  factor_group = rep(c("Control", "Treatment"), 3),
  measurement = c(15, 22, 18, 25, 12, 19)
)

# Pivot to wide format
wide_df <- original_df %>%
  pivot_wider(
    names_from = factor_group,  # Which column has your factor labels
    values_from = measurement   # Which column has the values to spread
  )

The resulting wide_df will have observation_id, Control, and Treatment as columns—perfect if you want side-by-side comparisons of each factor group.

2. Converting Wide Data to Long Data (Gather Factor Columns into Rows) #

If your data starts in "wide" format (each factor is a column), and you want to collapse it into a long format with a single factor column:

# Example original wide-format DataFrame
wide_original <- tibble(
  observation_id = 1:3,
  Control = c(15, 18, 12),
  Treatment = c(22, 25, 19)
)

# Pivot to long format
long_df <- wide_original %>%
  pivot_longer(
    cols = c(Control, Treatment),  # Which columns are your factor groups
    names_to = "factor_group",     # Name for the new factor column
    values_to = "measurement"      # Name for the new value column
  )

3. Other Common Reshaping Tasks with Factor Groups #

If you’re looking to sort, split, or aggregate by your factor group before reshaping, you can chain those steps too:

# Sort observations within each factor group
sorted_df <- original_df %>%
  group_by(factor_group) %>%
  arrange(measurement, .by_group = TRUE) %>%  # Sort values within each factor
  ungroup()

# Aggregate values by factor group and reshape
aggregated_wide <- original_df %>%
  group_by(factor_group) %>%
  summarise(avg_measurement = mean(measurement)) %>%
  pivot_wider(names_from = factor_group, values_from = avg_measurement)

If none of these examples match your exact desired output, feel free to share a snippet of your original DataFrame structure (you can use dput(head(your_df)) to get copy-pasteable code) and a description of what you want the final DataFrame to look like. That way we can tailor the solution exactly to your needs!

内容的提问来源于stack exchange，提问作者antecessor