R语言数据处理那些事（十八）函数列名封装

1.如何使用purrr按照行操作？

library(purrr)
 
df <- data.frame(
	height = c(1.7, 1.8, 1.65),
	weight = c(65, 72, 58)
)
 
# 计算每行的BMI
bmi <- pmap_dbl(df, function(height, weight, ...) {
	weight / (height^2)
}
)

2.如何传递列名封装为类似dplyr的函数？

library(dplyr)
library(rlang)
 
filter_by_col <- function(data, col, value) {
	col <- enquo(col)
	data %>% filter(!!col == value)
}
 
filter_by_col(mtcars, cyl, 6)
 

3.按物种分组保存为单独CSV文件？

library(dplyr)
iris %>%
	group_by(Species) %>%
	group_walk(~ write.csv(.x, file = paste0("iris_", .y$Species, ".csv")))

4.如何获取从某一列到最后一列的列名？

 
library(tidyverse)
candidate_df <- tibble::tibble(
	mapped_mass_1 = c(100.1, 200.2),
	MF = c("H2O", "C6H12O6"),
	raw = c("H2O", "C6H12O6"),
	added_K = c("H2OK", "C6H12O6K"),
	added_Na = c("H2ONa", "C6H12O6Na")
)
#方法1
mol_name <- colnames(candidate_df)[which(colnames(candidate_df) %in% "raw"):ncol(candidate_df)]
#方法2
mol_name <- candidate_df %>% select(raw:last_col()) %>% colnames()

5.如何在purrr中使用自定义的函数？

library(dplyr)
library(purrr)
 
# 每个数据框都有一列叫 letter
datas2 <- list(
	data.frame(id = 1:3, letter = c("A", "K", "C")),
	data.frame(id = 4:6, letter = c("K", "B", "K"))
)
 
get_site_to_query <- function(df, site) {
	df %>% filter(letter == site)
}
 
# 方法一：匿名函数
res <- map_dfr(datas2, function(x) { get_site_to_query(x, "K") })
 
# 方法二：参数传递（推荐）
res <- map_dfr(datas2, get_site_to_query, site = "K")