R语言新手求助：如何将df1行中参数值填充到df2对应列

Hey there! I get that working with unstructured key-value pairs in data frames can feel tricky when you're new to R—let's walk through a straightforward solution to get your data into the format you want.

Step 1: First, let's simulate your data (so you can follow along) #

I'll create sample versions of df1 and df2 to match your description:

# Sample df1: each row has key-value pairs as strings
df1 <- data.frame(params = c("A=5,C=1", "B=3,A=2", "C=7"), stringsAsFactors = FALSE)

# Sample df2: empty single-row structure with all parameter columns
df2 <- data.frame(A = NA, B = NA, C = NA)

Step 2: Use tidyverse for an intuitive solution #

The tidyverse package set makes parsing and reshaping data super straightforward. If you haven't installed it yet, run install.packages("tidyverse") first.

Here's the code to parse df1 and fill into the structure of df2:

library(tidyverse)

# Parse df1's key-value pairs into a structured data frame
parsed_df <- df1 %>%
  # Add a row ID to keep track of which original row each pair comes from
  mutate(row_id = row_number()) %>%
  # Split each row's comma-separated pairs into individual rows
  separate_rows(params, sep = ",") %>%
  # Split each "key=value" string into two columns
  separate(params, into = c("key", "value"), sep = "=", convert = TRUE) %>%
  # Reshape from long to wide format (one column per parameter)
  pivot_wider(names_from = key, values_from = value, id_cols = row_id) %>%
  # Remove the row ID column since we don't need it anymore
  select(-row_id)

# Align with df2's columns and fill missing values with NA (R's equivalent of NULL for data frames)
result_df <- parsed_df %>%
  # Reorder columns to match df2 exactly
  select(all_of(colnames(df2))) %>%
  # Replace any missing values with NA
  mutate(across(everything(), ~replace_na(., NA)))

# Check the result
print(result_df)

This will output exactly what you're looking for:

A  B  C
1  5 NA  1
2  2  3 NA
3 NA NA  7

Step 3: What if you prefer base R (no extra packages)? #

If you don't want to load the tidyverse, here's a base R alternative:

# Function to parse a single parameter string into a named list
parse_param_string <- function(param_str) {
  # Split string into individual key-value pairs
  pairs <- strsplit(param_str, ",")[[1]]
  # Split each pair into key and value
  key_value_pairs <- strsplit(pairs, "=")
  # Extract keys and values
  keys <- sapply(key_value_pairs, `[`, 1)
  values <- as.numeric(sapply(key_value_pairs, `[`, 2))
  # Return as a named list
  setNames(as.list(values), keys)
}

# Parse all rows in df1
parsed_list <- lapply(df1$params, parse_param_string)

# Convert the list to a data frame matching df2's structure
result_df <- do.call(rbind, lapply(parsed_list, function(row_data) {
  # For each column in df2, get the value or NA if missing
  data.frame(sapply(colnames(df2), function(col) row_data[[col]] %||% NA))
}))

# Note: The %||% operator is from the purrr package. If you don't have it, replace with:
# function(col) ifelse(is.null(row_data[[col]]), NA, row_data[[col]]))

Quick notes to clarify #

• In R, data frames use NA to represent missing values (what you referred to as NULL)—NULL is typically used for empty list elements, not data frame cells.
• Both methods will ensure that any parameters not present in a row of df1 get filled with NA in the corresponding column of the result.

内容的提问来源于stack exchange，提问作者rjn_bnf