如何基于含Frequency列的数据框创建计数列联表？

Got it, let's figure out how to build that exact cross tab you need using your data frame with a Frequency column. I know you tried table(), xtabs(), and aggregate() but couldn't get it right—let's fix that with base R tools that work perfectly for this scenario.

Core Approach: Use xtabs() to Sum Frequency by Groups #

The key here is telling R to sum the Frequency values across combinations of Hair and Eye categories, not just count rows. Here's the step-by-step code:

# Assume your data frame is named `df` (replace with your actual data frame name)
# Create the cross tabulation using Frequency as the value to aggregate
cross_tab <- xtabs(Frequency ~ Eye + Hair, data = df)

# Convert the contingency table to your desired data frame format
final_table <- as.data.frame.matrix(cross_tab)

Let's Test It with Your Example Data #

If we use the HairEyeColor converted to a data frame:

HEC <- as.data.frame(HairEyeColor)
cross_tab <- xtabs(Frequency ~ Eye + Hair, data = HEC)
final_table <- as.data.frame.matrix(cross_tab)

# Optional: Reorder columns to match your exact desired output
final_table <- final_table[, c("Black", "Blond", "Brown", "Red")]
print(final_table)

This will output exactly the table you showed:

Black Blond Brown Red
Blue     20    94    84  17
Brown    68     7   119  26
Green     5    16    29  14
Hazel    15    10    54  14

Why This Works (And What You Might Have Missed) #

• xtabs() is designed for cross-tabulating weighted data (which is exactly what your Frequency column is). The formula Frequency ~ Eye + Hair tells R: "sum Frequency for every combination of Eye (rows) and Hair (columns)".
• as.data.frame.matrix() converts the contingency table object from xtabs() into a standard data frame with columns as Hair categories and rows as Eye categories—just like your target table.

Alternative Base R Method: tapply() #

If you prefer another base R approach, tapply() works too:

final_table <- as.data.frame(with(df, tapply(Frequency, list(Eye, Hair), sum)))
# Again, reorder columns if needed
final_table <- final_table[, c("Black", "Blond", "Brown", "Red")]

Both methods avoid plyr entirely and leverage base R tools to properly use your Frequency column instead of just counting rows.

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