我正在尝试从小标题中的嵌套列表中提取第n个值。我正在尝试使用tidyverse。
我的原始数据是这样的
> x %>%
+ select( .,
+ meta_event,
+ meta_aktivitet,
+ meta_subject_id,
+ time,
+ time_cum,
+ vo2) ->
+ x
> dput(head(x))
structure(list(meta_event = c("001", "001", "001", "001", "001",
"001"), meta_aktivitet = c("001", "001", "001", "001", "001",
"001"), meta_subject_id = c("100001", "100001", "100001", "100001",
"100001", "100001"), time = c("0:10 min", "0:15 min", "0:20 min",
"0:25 min", "0:30 min", "0:35 min"), time_cum = c(10, 15, 20,
25, 30, 35), vo2 = c(1.665, 2.515, 2.641, 2.677, 2.66, 2.712)), row.names = c(NA,
-6L), class = c("tbl_df", "tbl", "data.frame"))
并嵌套
> x %>%
+ group_by( .,
+ meta_event,
+ meta_aktivitet,
+ meta_subject_id) %>%
+ nest() ->
+ x_n
> head(x_n)
# A tibble: 6 x 4
meta_event meta_aktivitet meta_subject_id data
<chr> <chr> <chr> <list>
1 001 001 100001 <tibble [~
2 001 001 100002 <tibble [~
3 001 001 100003 <tibble [~
4 001 001 100004 <tibble [~
5 001 001 100005 <tibble [~
6 001 001 100006 <tibble [~
在这里，我尝试通过mutate和map_dbl提取一些信息。我尝试了不同的方式，没有任何运气。
> x_n %>%
+ mutate( vald_end = data %>% map_dbl( ., function(x) max(x$time_cum)),
+ h1 = vald_end - 15*60,
+ index_1 = data %>% map_dbl( ., function(x) which.max(x$time_cum > h1)),
+ index_2 = index_1 - 1,
+ vald_start = map_dbl( data, ~ nth( "time_cum",
+ n = index_1,
+ order_by = "time_cum"))) %>%
+ unnest() %>% View()
Error in mutate_impl(.data, dots) :
Evaluation error: length(n) == 1 is not TRUE.
In addition: There were 20 warnings (use warnings() to see them)
似乎在最后一个突变上失败了。如果我不带“ vald_start”运行代码，则会产生以下结果
> x_n %>%
+ mutate( vald_end = data %>% map_dbl( ., function(x) max(x$time_cum)),
+ h1 = vald_end - 15*60,
+ index_1 = data %>% map_dbl( ., function(x) which.max(x$time_cum > h1)),
+ index_2 = index_1 - 1) %>%
+ unnest() ->
+ x_r
There were 20 warnings (use warnings() to see them)
> x_r
# A tibble: 10,639 x 10
meta_event meta_aktivitet meta_subject_id vald_end h1 index_1 index_2
<chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
1 001 001 100001 4015 3115 491 490
2 001 001 100001 4015 3115 491 490
3 001 001 100001 4015 3115 491 490
4 001 001 100001 4015 3115 491 490
5 001 001 100001 4015 3115 491 490
6 001 001 100001 4015 3115 491 490
7 001 001 100001 4015 3115 491 490
8 001 001 100001 4015 3115 491 490
9 001 001 100001 4015 3115 491 490
10 001 001 100001 4015 3115 491 490
# ... with 10,629 more rows, and 3 more variables: time <chr>,
# time_cum <dbl>, vo2 <dbl>
> dput(head(x_r))
structure(list(meta_event = c("001", "001", "001", "001", "001",
"001"), meta_aktivitet = c("001", "001", "001", "001", "001",
"001"), meta_subject_id = c("100001", "100001", "100001", "100001",
"100001", "100001"), vald_end = c(4015, 4015, 4015, 4015, 4015,
4015), h1 = c(3115, 3115, 3115, 3115, 3115, 3115), index_1 = c(491,
491, 491, 491, 491, 491), index_2 = c(490, 490, 490, 490, 490,
490), time = c("0:10 min", "0:15 min", "0:20 min", "0:25 min",
"0:30 min", "0:35 min"), time_cum = c(10, 15, 20, 25, 30, 35),
vo2 = c(1.665, 2.515, 2.641, 2.677, 2.66, 2.712)), row.names = c(NA,
-6L), class = c("tbl_df", "tbl", "data.frame"))
有人可以用nth和map_dbl指向我正确的方向
根据@ Moody_Mudskipper的建议，我尝试了以下代码，没有遇到任何麻烦。我已将第n个函数中的两个字符串“ time_cum”更改为。$ time_cum
> x_n %>%
+ mutate( vald_end = data %>% map_dbl( ., function(x) max(x$time_cum)),
+ h1 = vald_end - 15*60,
+ index_1 = data %>% map_dbl( ., function(x) which.max(x$time_cum > h1)),
+ index_2 = index_1 - 1,
+ vald_start = data %>% map_dbl( ., ~ nth( .$time_cum,
+ n = index_1,
+ order_by = .$time_cum))) %>%
+ unnest() %>% View()
Error in mutate_impl(.data, dots) :
Evaluation error: length(n) == 1 is not TRUE.
In addition: There were 20 warnings (use warnings() to see them)
如果我更改代码，所以我引用使用由index_1插入的数字，那么R提供了预期的结果。因此，当我引用index_1列时，我一定做错了。
> x_n %>%
+ mutate( vald_end = data %>% map_dbl( ., function(x) max(x$time_cum)),
+ h1 = vald_end - 15*60,
+ index_1 = data %>% map_dbl( ., function(x) which.max(x$time_cum > h1)),
+ index_2 = index_1 - 1,
+ vald_start = data %>% map_dbl( ., ~ nth( .$time_cum,
+ n = 490,
+ order_by = .$time_cum))) ->
+ x_r
There were 20 warnings (use warnings() to see them)
> x_r
# A tibble: 21 x 9
meta_event meta_aktivitet meta_subject_id data vald_end h1 index_1 index_2 vald_start
<chr> <chr> <chr> <list> <dbl> <dbl> <dbl> <dbl> <dbl>
1 001 001 100001 <tibble [5~ 4015 3115 491 490 2470
2 001 001 100002 <tibble [5~ 3770 2870 459 458 3575
3 001 001 100003 <tibble [3~ 3840 2940 346 345 NA
4 001 001 100004 <tibble [5~ 3630 2730 456 455 3455
5 001 001 100005 <tibble [5~ 3945 3045 493 492 2460
6 001 001 100006 <tibble [5~ 3860 2960 422 421 3660
7 001 001 100007 <tibble [4~ 3480 2580 356 355 NA
8 002 001 100001 <tibble [5~ 4570 3670 498 497 2450
9 002 001 100002 <tibble [5~ 3755 2855 424 423 3640
10 002 001 100003 <tibble [4~ 3560 2660 361 360 NA
# ... with 11 more rows
来自丹麦的问候。
Dan Olesen