对于列中首次出现的值，请记录其他列中的最小值

Question

我有一个数据帧'all'，它的子集看起来像这个

ONSETS_S    DURATIONS_S CUMUlATIVE_DURATION Phase Active  Block   TRIAL
0           2.926       2.926               1     NA      NA       NA
2.926       1.925       4.851               1     NA      NA       NA
4.851       4.851       9.702               1     Passive Baseline 1
9.702       1.54        11.242              1     Passive Baseline 1
11.242      4.851       16.093              1     Passive Baseline 1
107.183     1.925       109.109             1     NA      NA       NA
109.108     4.851       113.96              1     Active  Baseline 1
114.135     1.848       115.808             1     Active  Baseline 1
115.807     4.851       120.659             1     Active  Baseline 1
312.232     1.925       314.16              1     NA      NA       NA
314.157     4.851       319.011             1     Passive Baseline 1
319.008     2.387       321.398             1     Passive Baseline 1
321.395     4.851       326.249             1     Passive Baseline 1
415.797     1.925       417.725             1     NA      NA       NA
417.722     4.851       422.576             1     Active  GD       2
422.762     2.31        424.886             1     Active  GD       2
424.883     4.851       429.737             1     Active  GD       2

在整个数据帧中，总共有6个阶段，块中有5种类型的条目，每个阶段每个块最多有2次试验。当阶段、活动和块条件值固定时，我想为每个试验在ONSETS_S中创建一个最小值矩阵。例如，当Phase == 1 & Active == "Passive" & Block == "Baseline"时，输出应该是P_1_Passive_Onsets <- [4.851 314.157]

Answer 1

使用dplyr的简单解决方案。

目前还不完全清楚你想要什么，但从你的描述中，听起来你想在TRIAL、Phase、Active和Block级别之间进行分组，并找到每个级别的最小/最大值。

library(dplyr)
dat %>% 
group_by(TRIAL, Phase, Active, Block) %>% 
summarise(min_val = min(ONSETS_S), max_val = max(ONSETS_S))

# A tibble: 4 × 6
# Groups:   TRIAL, Phase, Active [4]
TRIAL Phase Active  Block    min_val max_val
<int> <int> <chr>   <chr>      <dbl>   <dbl>
1     1     1 Active  Baseline  109.      116.
2     1     1 Passive Baseline    4.85    321.
3     2     1 Active  GD        418.      425.
4    NA     1 NA      NA          0       416.

或使用data.table

library(data.table)
setDT(dat)
dat[, list(min_val = min(ONSETS_S), max_val = max(ONSETS_S)), by=c("TRIAL", "Phase", "Active", "Block")]
TRIAL Phase  Active    Block min_val max_val
1:    NA     1    <NA>     <NA>   0.000 415.797
2:     1     1 Passive Baseline   4.851 321.395
3:     1     1  Active Baseline 109.108 115.807
4:     2     1  Active       GD 417.722 424.883