Data Frame Operations in R

In the R language's data analysis ecosystem, the Data Frame is the most fundamental and widely used data structure. Essentially, it is a two-dimensional table object composed of rows (observations) and columns (variables), where each column must consist of data of the same type. Unlike Lists, which do not enforce a shared indexing system across elements, Data Frames require all columns to share the same index system. This gives them a natural advantage when handling statistical data and performing matrix operations.

There are two standard methods for creating Data Frames. The first is using the data.frame() function by passing multiple vectors. For example, we can combine three numeric vectors and one character vector into a table containing age, salary, and job information:

age <- c(25, 30, 35)
salary <- c(40000, 50000, 60000)
job <- c("工程师", "经理", "总监")
df <- data.frame(age, salary, job)

The second method involves using the read.table() or read.csv() functions to read data from external files, which is the most common scenario in practical data analysis.
Mastering Data Frame operations relies heavily on efficient querying capabilities. R provides flexible syntax that allows users to quickly extract desired subsets through logical conditions. The most basic filtering method is based on logical judgments using column names. For instance, to find employee records with a salary higher than 50000, you can use the following code:

high_salary <- df[df$salary > 50000, ]

Here, the empty space after the comma indicates retaining all rows, while the condition before the comma is used to filter specific columns.

In addition to numeric comparisons, logical operators & (AND), | (OR), and ! (NOT) are also key in constructing complex filtering conditions. If we need data that satisfies both "age greater than 30" AND "job is Engineer", it can be written as:

target_group <- df[age > 30 & job == "工程师", ]

Furthermore, R supports using the which() function in conjunction with logical vectors to locate row indices. This approach is often clearer and more efficient than directly using conditions when dealing with large-scale data.

Column Operations and Attribute Management of Data

Columns in a Data Frame are independent data objects, so they can be modified, deleted, or renamed independently. Deleting a column is very intuitive; you simply specify the column name and set remove = TRUE (though this parameter is usually omitted in newer versions of R):

df <- df[, c("age", "salary")] # Directly select via column indexing, automatically excluding the job column
# Or explicitly delete
df$job <- NULL

Renaming columns uses the names() function combined with an assignment operation. If the original data labels "Salary" as "薪资", we can uniformly change it to Chinese:

names(df)[2] <- "薪资"
names(df)[3] <- "职业"

Regarding data types, although R infers them automatically, manual conversion is often necessary after importing data to ensure calculation precision. For example, converting character-type age to numeric type:

df$age <- as.numeric(df$age)

Advanced Techniques for Merging and Splitting Datasets

In real-world projects, data is often scattered across different files or variables. R provides the powerful merge() function to implement connection operations based on keys. This is similar to the JOIN concept in SQL and is commonly used to integrate multi-table data. Suppose we have two Data Frames: df1 (containing Employee ID and Name) and df2 (containing Employee ID and Department). We merge them through the common "ID" column:

merged_df <- merge(df1, df2, by = "ID")

By default, merge() performs an Inner Join, meaning it retains only records where keys exist in both tables. If you need to retain all records and fill in missing values, you can specify all = TRUE.

Conversely, when you need to split a large table into multiple sub-tables, you can use the split() function. This function splits a Data Frame into list-form subsets based on a specified factor column:

dept_list <- split(df, df$部门)
# dept_list is now a list containing sub-data frames named "工程师", "经理", etc.

This structure is particularly suitable for subsequent parallel processing or group-wise statistical calculations.

Performance Optimization and Best Practice Recommendations

When handling data at the level of millions or tens of millions of records, operational efficiency is crucial. First, always ensure that columns involved in comparisons and calculations are numeric rather than character, because R performs element-wise operations on character vectors, which often leads to unnecessary overhead. Second, avoid repeatedly creating temporary objects within loops; instead, make use of vectorized functions that return results directly. Finally, after performing large-scale filtering, remember to clean up variables no longer needed to free up memory, for example, using the rm() function to delete intermediate results.

By proficiently applying the above operations, Data Frames in R can become a solid foundation for building complex analysis models. From basic reading and cleaning to advanced merging and aggregation, mastering these skills is the必经之路 (inevitable path) to entering the field of R language data science.