如何在 r 中将 enquo 与向量元素一起使用

Question

处理自定义函数时出现错误。“错误：Quosures 只能在 quasiquotation 上下文中不被引用。”很多类似的问答，但我在技术上还不足以解释。我所能辨别的是，这是我通过自定义函数调用 time_sec[event=='A'] 的尝试。违规行在下面的代码中注明。

library(dplyr)

# the data
set.seed(123)
event_df <- tibble(time_sec = c(1:120)) %>% 
  sample_n(100) %>%
  mutate(period = sample(c(1,2,3),
                         size = 100,
                         replace = TRUE),
         event = sample(c("A","B"), 
                        size = 100, 
                        replace = TRUE, 
                        prob = c(0.1,0.9))) %>% 
  select(period, time_sec, event) %>% 
  arrange(period, time_sec)

# the function
filter_within_timeframe <- function(.data, condition, time, lead_time = 0, lag_time = 0){
  condition <- enquo(condition)
  time <- enquo(time)
  filtered_lead <- c()
  filtered_lag <- c()
  
  if(lead_time > 0){
  # filter lead events
  filtered_lead <- .data %>%
    slice(., 1:max(which(!!condition))) %>%
    group_by(., grp =  lag(cumsum(!!condition), default = 0)) %>%
    filter(.,(last(!!time) - !!time) <= lead_time & (last(!!time) - !!time) >= 0) %>% 
    select(-grp)
  }

  if(lag_time > 0){
  #filter lag events
  filtered_lag <- .data %>%
    slice(min(which(!!condition)):nrow(.)) %>% # remove rows before first occurance
    group_by(grp =  lead(rev(cumsum(rev(!!condition))), default = 0), period) %>%
    filter((!!time - first(!!time[!!condition])) <= lag_time) # something in here doesn't work !!!!!!!!!!!!!
  }

  filtered <- rbind(filtered_lead, filtered_lag)
  return(filtered)
  
}

利用lead_time 的调用按预期工作......

event_df %>% filter_within_timeframe(event == 'A', time_sec, 10, 0)

...但是对 lag_time 的调用会引发错误..

event_df %>% filter_within_timeframe(event == 'A', time_sec, 0, 10)

Answer 1

您可以使用{{ }} 在函数中使用不带引号的列名。

library(dplyr)

filter_within_timeframe <- function(.data, condition, time, lead_time = 0, lag_time = 0){
  condition <- enquo(condition)
  filtered_lead <- c()
  filtered_lag <- c()
  
  if(lead_time > 0){
    # filter lead events
    filtered_lead <- .data %>%
      slice(., 1:max(which(!!condition))) %>%
      group_by(., grp =  lag(cumsum(!!condition), default = 0)) %>%
      filter(.,(last({{time}}) - {{time}}) <= lead_time & (last({{time}}) - {{time}}) >= 0) 
  }
  
  if(lag_time > 0){
    #filter lag events
    filtered_lag <- .data %>%
      slice(min(which(!!condition)):nrow(.)) %>% # remove rows before first occurance
      group_by(grp =  lead(rev(cumsum(rev(!!condition))), default = 0), period) %>%
      filter(({{time}} - first({{time}}[match(!!condition, TRUE)])) <= lag_time)
  }
  
  filtered <- rbind(filtered_lead, filtered_lag)
  return(filtered)
  
}

使用原始函数将变量替换为{{}} 后出现错误。我用match解决了。

我们可以这样称呼它：

event_df %>% filter_within_timeframe(event == 'A', time_sec, 10, 0)
#     grp period time_sec event
#   <dbl>  <dbl>    <int> <chr>
# 1     0      1       43 B    
# 2     0      1       45 B    
# 3     0      1       52 A    
# 4     1      1       54 A    
# 5     2      1       61 B    
# 6     2      1       64 B    
# 7     2      1       70 A    
# 8     3      1       92 B    
# 9     3      1       95 A    
#10     4      2       75 B    
# … with 13 more rows

event_df %>% filter_within_timeframe(event == 'A', time_sec, 0, 10)

#  period time_sec event   grp
#    <dbl>    <int> <chr> <dbl>
# 1      1       52 A         7
# 2      1       54 A         6
# 3      1       58 B         6
# 4      1       59 B         6
# 5      1       61 B         6
# 6      1       64 B         6
# 7      1       70 A         5
# 8      1       72 B         5
# 9      1       74 B         5
#10      1       76 B         5
# … with 17 more rows