整洁运算 tidy evaluation

数据处理
Author

hcl

Published

September 27, 2022

整洁计算

tidyverse 代码之所以 “整洁、优雅”，是因为它内部采用了一套整洁计算框架。

  1. 数据屏蔽与整洁选择

  • 整洁计算的两种基本形式是：

    • 数据屏蔽：使得可以不用带数据框（环境变量）名字，就能使用数据框内的变量（数据变量），便于在数据集内计算值
    • 整洁选择：即各种选择列语法，便于使用数据集中的列

  • 数据屏蔽内在的机制是先冻结表达式，然后注入函数，再恢复其计算。 整洁计算已经为此做好了两种封装：

  1. {{ }}（curly-curly 算符） : 若只是传递，可将 “冻结 + 注入” 合成一步
  2. enquo() 和!!（引用与反引用）：不只是传递，冻结后还需要访问

{{}}运算符

  • 自定义函数时，想要像 tidyverse 那样整洁传递变量名，需要用到{{ }}, 即用两个大括号将变量括起来：
library(tidyverse)
var_summary = function(data, var){
    data %>% 
        summarise(n = n(), mean = mean({{var}}))
}

mtcars %>%
    group_by(cyl) %>%
    var_summary(mpg)
# A tibble: 3 × 3
    cyl     n  mean
  <dbl> <int> <dbl>
1     4    11  26.7
2     6     7  19.7
3     8    14  15.1
  • 若想用字符串形式传递变量名，在访问数据时需要借助.data[[var]]，这里.data 是代替数据集
var_summary = function(data, var){
    data %>% 
        summarise(n = n(), mean = mean(.data[[var]]))
}

mtcars %>%
    group_by(cyl) %>%
    var_summary("mpg")
# A tibble: 3 × 3
    cyl     n  mean
  <dbl> <int> <dbl>
1     4    11  26.7
2     6     7  19.7
3     8    14  15.1
  • 还可用于对列名向量做循环迭代，比如对因子型各列计算各水平值频数：
mtcars[, 9:10] %>% 
    names() %>%
    map(~ count(mtcars, .data[[.x]]))
[[1]]
  am  n
1  0 19
2  1 13

[[2]]
  gear  n
1    3 15
2    4 12
3    5  5
  • 将整洁选择作为函数参数传递，也需要用到 {{ }}
summarise_mean = function(data, vars){
    data %>%
        summarise(n = n(), across({{vars}}, mean))
}

mtcars %>%
    group_by(cyl) %>%
    summarise_mean(where(is.numeric))
# A tibble: 3 × 12
    cyl     n   mpg  disp    hp  drat    wt  qsec    vs    am  gear  carb
  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
1     4    11  26.7  105.  82.6  4.07  2.29  19.1 0.909 0.727  4.09  1.55
2     6     7  19.7  183. 122.   3.59  3.12  18.0 0.571 0.429  3.86  3.43
3     8    14  15.1  353. 209.   3.23  4.00  16.8 0     0.143  3.29  3.5
  • 若传递是多个列名构成的字符向量，则需要借助函数 all_of() 或 any_of()
vars = c("mpg", "vs")
mtcars %>% select(all_of(vars))
mtcars %>% select(!all_of(vars))
  • 使用 {{ }} 或整洁选择同时修改列名
my_summarise = function(data, mean_var, sd_var) {
    data %>%
        summarise("mean_{{mean_var}}" := mean({{mean_var}}),
                 "sd_{{sd_var}}" := mean({{sd_var}}))
}   

mtcars %>%
    group_by(cyl) %>%
    my_summarise(mpg, disp)
# A tibble: 3 × 3
    cyl mean_mpg sd_disp
  <dbl>    <dbl>   <dbl>
1     4     26.7    105.
2     6     19.7    183.
3     8     15.1    353.
my_summarise <- function(data, group_var, summarise_var) {
     data %>%
         group_by(across({{ group_var }})) %>%
         summarise(across({{ summarise_var }}, mean,
             .names = "mean_{.col}"
         ))
 }
mtcars %>%
     my_summarise(c(am, cyl), where(is.numeric))
`summarise()` has grouped output by 'am'. You can override using the `.groups`
argument.
# A tibble: 6 × 11
# Groups:   am [2]
     am   cyl mean_mpg mean_disp mean_hp mean_…¹ mean_wt mean_…² mean_vs mean_…³
  <dbl> <dbl>    <dbl>     <dbl>   <dbl>   <dbl>   <dbl>   <dbl>   <dbl>   <dbl>
1     0     4     22.9     136.     84.7    3.77    2.94    21.0   1        3.67
2     0     6     19.1     205.    115.     3.42    3.39    19.2   1        3.5 
3     0     8     15.0     358.    194.     3.12    4.10    17.1   0        3   
4     1     4     28.1      93.6    81.9    4.18    2.04    18.4   0.875    4.25
5     1     6     20.6     155     132.     3.81    2.76    16.3   0        4.33
6     1     8     15.4     326     300.     3.88    3.37    14.6   0        5   
# … with 1 more variable: mean_carb <dbl>, and abbreviated variable names
#   ¹​mean_drat, ²​mean_qsec, ³​mean_gear
  • 传递列名向量，同时修改列名
var_summary <- function(data, var) {
    data %>%
        summarise(
            n = n(),
            !!enquo(var) := mean(.data[[var]])
        )
}
mtcars %>%
    group_by(cyl) %>%
    var_summary("mpg")
# A tibble: 3 × 3
    cyl     n   mpg
  <dbl> <int> <dbl>
1     4    11  26.7
2     6     7  19.7
3     8    14  15.1
var_summary <- function(data, var) {
    data %>%
        summarise(
            n = n(),
            !!str_c("mean_", var) := mean(.data[[var]])
        )
}
mtcars %>%
    group_by(cyl) %>%
    var_summary("mpg")
# A tibble: 3 × 3
    cyl     n mean_mpg
  <dbl> <int>    <dbl>
1     4    11     26.7
2     6     7     19.7
3     8    14     15.1

引用与反引用

引用与反引用将冻结和注入分成两步，在使用上更加灵活：

  • 用 enquo() 让函数自动引用其参数

  • 用!! 反引用该参数

  • 以自定义计算分组均值函数为例：

grouped_mean <- function(data, summary_var, group_var) {
    summary_var <- enquo(summary_var)
    group_var <- enquo(group_var)
    data %>%
        group_by(!!group_var) %>%
        summarise(mean = mean(!!summary_var))
}
grouped_mean(mtcars, mpg, cyl)
# A tibble: 3 × 2
    cyl  mean
  <dbl> <dbl>
1     4  26.7
2     6  19.7
3     8  15.1
  • 要想修改结果列名，可借助 as_label() 从引用中提取名字
grouped_mean <- function(data, summary_var, group_var) {
    summary_var <- enquo(summary_var)
    group_var <- enquo(group_var)
    summary_nm <- str_c("mean_", as_label(summary_var))
    group_nm <- str_c("group_", as_label(group_var))
    data %>%
        group_by(!!group_nm := !!group_var) %>%
        summarise(!!summary_nm := mean(!!summary_var))
}

grouped_mean(mtcars, mpg, cyl)
# A tibble: 3 × 2
  group_cyl mean_mpg
      <dbl>    <dbl>
1         4     26.7
2         6     19.7
3         8     15.1
  • 要传递多个参数可以用特殊参数…。比如，还想让用于计算分组均值的group_var 可以是任意多个
grouped_mean <- function(.data, .summary_var, ...) {
    summary_var <- enquo(.summary_var)
    .data %>%
        group_by(...) %>%
        summarise(mean = mean(!!summary_var))
}
grouped_mean(mtcars, disp, cyl, am)
`summarise()` has grouped output by 'cyl'. You can override using the `.groups`
argument.
# A tibble: 6 × 3
# Groups:   cyl [3]
    cyl    am  mean
  <dbl> <dbl> <dbl>
1     4     0 136. 
2     4     1  93.6
3     6     0 205. 
4     6     1 155  
5     8     0 358. 
6     8     1 326
  • … 参数不需要做引用和反引用就能正确工作，但若要修改结果列名仍需要借助引用和反引用，但是要用 enques() 和!!!
grouped_mean <- function(.data, .summary_var, ...) {
    summary_var <- enquo(.summary_var)
    group_vars <- enquos(..., .named = TRUE)
    summary_nm <- str_c("avg_", as_label(summary_var))
    names(group_vars) <- str_c("groups_", names(group_vars))
    .data %>%
        group_by(!!!group_vars) %>%
        summarise(!!summary_nm := mean(!!summary_var))
}
grouped_mean(mtcars, disp, cyl, am)
`summarise()` has grouped output by 'groups_cyl'. You can override using the
`.groups` argument.
# A tibble: 6 × 3
# Groups:   groups_cyl [3]
  groups_cyl groups_am avg_disp
       <dbl>     <dbl>    <dbl>
1          4         0    136. 
2          4         1     93.6
3          6         0    205. 
4          6         1    155  
5          8         0    358. 
6          8         1    326
  • 参数… 还可以传递表达式
filter_fun = function(df, ...){
    filter(df, ...)
}

mtcars %>% 
    filter_fun(mpg > 25 & disp > 90)
               mpg cyl  disp  hp drat    wt qsec vs am gear carb
Porsche 914-2 26.0   4 120.3  91 4.43 2.140 16.7  0  1    5    2
Lotus Europa  30.4   4  95.1 113 3.77 1.513 16.9  1  1    5    2
  • 最后看一个自定义绘制散点图的模板函数
scatter_plot = function(df, x_var,y_var) {
x_var = enquo(x_var)
y_var = enquo(y_var)
ggplot(data = df, aes(x = !!x_var, y = !!y_var)) +
    geom_point() +
    theme_bw() +
    # theme(plot.title = element_text(lineheight = 1, face = "b
    geom_smooth() +
    ggtitle(str_c(as_label(y_var), " vs. ", as_label(x_var)))
}

scatter_plot(mtcars, disp, hp)

鄂ICP备2022016232号-1