Question

这是对this thread的跟进。像那里一样，我需要定义组列的游程类型ID（忽略NA），并附加列seq_break指示序列应在seq_break = TRUE时结束。但是，该主题中提供的解决方案在seq_break = TRUE行中开始一个新序列，而实际上应将它作为先前序列的最后一个事件包括在内。示例数据附在下面。可以在行46中观察到这种差异-先前的解决方案将从此处的序列元素13开始，而我需要将其包含在序列12中。

df <- structure(list(group = c(NA, NA, "home", "home", "home", "home", 
"home", "home", "away", NA, NA, "home", "home", "home", NA, NA, 
NA, "home", "away", "away", NA, "away", "away", "away", "home", 
"away", "away", "away", NA, "home", "home", NA, NA, "away", NA, 
NA, "home", NA, NA, "home", "home", "home", "home", "home", "home", 
"home", "away", "away", NA, NA), seq_break = c(FALSE, FALSE, 
FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, TRUE, TRUE, 
FALSE, FALSE, FALSE, TRUE, TRUE, TRUE, FALSE, FALSE, FALSE, TRUE, 
FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, FALSE, TRUE, FALSE, 
FALSE, TRUE, TRUE, FALSE, TRUE, FALSE, FALSE, TRUE, TRUE, FALSE, 
FALSE, FALSE, FALSE, FALSE, FALSE, TRUE, FALSE, FALSE, TRUE, 
TRUE), expected_output = c(NA, NA, 1, 1, 1, 1, 1, 1, 2, NA, NA, 
3, 3, 3, NA, NA, NA, 4, 5, 5, NA, 6, 6, 6, 7, 8, 8, 8, NA, 9, 
9, NA, NA, 10, NA, NA, 11, NA, NA, 12, 12, 12, 12, 12, 12, 12, 
13, 13, NA, NA)), .Names = c("group", "seq_break", "expected_output"
), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
-50L))

有什么想法可以用tidyverse完成吗？我不知道如何在这里替换cumsum ...

Answer 1

使用rleid并从data.table中转移...

library(data.table)
setDT(df)

# make groups
df[, v := rleid(group, shift(cumsum(seq_break)))]

# drop if group is NA
df[is.na(group), v := NA]

# renumber the others
df[!is.na(group), v := .GRP, by=v]

# check
stopifnot( df[, all.equal(v, expected_output)] )

在示例中，seq_break列实际上是无关紧要的，因此我不确定是否正确使用了它：

df[, v2 := rleid(group)][is.na(group), v2 := NA][!is.na(group), v2 := .GRP, by=v2]

# check
stopifnot( df[, all.equal(v2, expected_output)] )

由于OP想要一个'反向答案，所以这里是一个翻译（仍然使用rleid）：

library(dplyr)
res = df  %>% mutate(
  v2 = data.table::rleid(group) %>% replace(is.na(group), NA),
  v2 = match(v2, na.omit(unique(v2)))
) 

# check
stopifnot( with(res, all.equal(v2, expected_output)) )

Answer 2

我们可以创建一个新列来调用seq_break2并添加到管道中，如下所示。这样会产生与预期输出相同的结果。

library(tidyverse)
library(data.table)

df2 <- df %>% 
  select(-expected_output) %>%
  rowid_to_column() 

df3 <- df2 %>%
  mutate(seq_break2 = ifelse(seq_break & !is.na(group), FALSE, seq_break)) %>%
  mutate(ID = rleid(group, seq_break2)) %>%
  group_by(group, seq_break2, ID) %>%
  filter(!(is.na(group) & seq_break2 & row_number() > 1)) %>%
  ungroup() %>%
  mutate(ID2 = cumsum(seq_break2)) %>%
  drop_na(group) %>%
  mutate(expected_output = rleid(group, ID2)) %>%
  select(rowid, expected_output) %>%
  left_join(df2, ., by = "rowid") %>%
  select(-rowid)

如何使用其他列条件定义游程序列

2 个答案: