从数据框列表中提取列，并将其存储在另一个数据框列表中

Question

我有一个名为-cj1-的列表，其中包含多个数据框

dput(head(cj1[1:2]))
list(structure(list(individual = c("a12TTT.pdf", "a15.pdf", "a17.pdf", 
"a18.pdf", "a21.pdf", "a2TTT.pdf", "a5.pdf", "B11.pdf", "B12.pdf", 
"B13.pdf", "B22.pdf", "B24.pdf", "B4.pdf", "B7.pdf", "B8.pdf", 
"cw10-1.pdf", "cw13-1.pdf", "cw15-1TTT.pdf", "cw17-1.pdf", "cw18.pdf", 
"cw3.pdf", "cw4.pdf", "cw7_1TTT.pdf"), id = 1:23, Ntot = c(13, 
9, 16, 15, 9, 13, 10, 10, 11, 10, 14, 10, 11, 12, 11, 10, 15, 
12, 14, 11, 9, 10, 11), N1 = c(5, 5, 10, 11, 7, 9, 5, 5, 6, 8, 
8, 8, 9, 8, 7, 1, 0, 6, 3, 4, 2, 4, 2), ND = c(0, 0, 0, 0, 0, 
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), N0 = c(8, 
4, 6, 4, 2, 4, 5, 5, 5, 2, 6, 2, 2, 4, 4, 9, 15, 6, 11, 7, 7, 
6, 9), score = c(5.06923076923077, 4.96666666666667, 9.925, 10.86, 
6.83333333333333, 8.88461538461539, 5, 5, 5.97272727272727, 7.82, 
7.95714285714286, 7.82, 8.80909090909091, 7.9, 6.91818181818182, 
1.24, 0.3, 6, 3.17142857142857, 4.08181818181818, 2.16666666666667, 
4.06, 2.19090909090909), propscore = c(0.389940828402367, 0.551851851851852, 
0.6203125, 0.724, 0.759259259259259, 0.683431952662722, 0.5, 
0.5, 0.54297520661157, 0.782, 0.568367346938776, 0.782, 0.800826446280992, 
0.658333333333333, 0.628925619834711, 0.124, 0.02, 0.5, 0.226530612244898, 
0.371074380165289, 0.240740740740741, 0.406, 0.199173553719008
), theta = c(-0.571211122446447, 0.418736780198501, 0.464533662219296, 
0.760432013134893, 1.43961032059382, 0.935963883364303, 0.0742361005467161, 
0.416783201347136, 0.232586422933618, 1.65345248955369, 0.178947462869717, 
1.3980442736112, 1.5300599487058, 0.340087410746963, 0.616985944469495, 
-1.73246102772711, -4.06186172096556, -0.347700710331151, -1.21009964741398, 
0.239145600406579, -1.88836418690337, -0.276451472526056, -0.611455626388059
), se.theta = c(0.689550115014498, 0.689441554709003, 0.595659709892116, 
0.609506508256404, 0.917792293663691, 0.652011367164736, 0.720534163064516, 
0.695969555549033, 0.661019531367007, 0.87050969318314, 0.605775647419845, 
0.797443937820774, 0.768436114096332, 0.695748274310803, 0.709380679025605, 
1.00089414765463, 1.8701468050665, 0.68959824350285, 0.733014089189809, 
0.656392513303483, 0.952935324276941, 0.71608982789968, 0.771906532861938
), outfit = c(1.24922700170817, 1.46067763769417, 0.915183304626819, 
0.753992664091072, 0.37410361433915, 0.727316037037668, 0.616907868814702, 
1.01528298230254, 1.01594232662062, 0.616808170683195, 0.646097057961938, 
0.622993494551005, 0.807441271101246, 0.788526018181888, 1.2157399735092, 
0.341189086206191, 0.021052091633073, 0.543024513106335, 1.04183076617928, 
1.1772656963046, 0.736106160865241, 0.756316095787985, 0.58320701094964
), infit = c(1.4078580948461, 1.42854494963967, 1.09762978932861, 
0.893957122448352, 0.64936943769433, 0.899191443180872, 0.724956556509282, 
1.14975990693782, 1.08074439712469, 0.978248081241133, 0.755557633771936, 
0.823903684368671, 0.911855771375284, 0.954272320131035, 0.926253596526142, 
0.634052701587448, 0.0504659822408584, 0.712539957033173, 0.966034039620798, 
1.1901663169553, 0.81371119642719, 0.817417869881874, 0.737574872116582
)), row.names = c(NA, -23L), class = "data.frame"), structure(list(
    parlabel = c("Ties", "Home"), par = c("delta", "eta"), est = c(-43.5016417611571, 
    0.337872999554289), se = c(366043197.615422, 0.215169736220537
    )), row.names = c(NA, -2L), class = "data.frame"))

以下是数据框的外观：

head(cj1[[1]],2)
  individual id Ntot N1 ND N0    score propscore      theta  se.theta   outfit
1 a12TTT.pdf  1   13  5  0  8 5.069231 0.3899408 -0.5712111 0.6895501 1.249227
2    a15.pdf  2    9  5  0  4 4.966667 0.5518519  0.4187368 0.6894416 1.460678
     infit
1 1.407858
2 1.428545

我想创建一个名为-results1-的单独列表，其中包含数据框，这些数据框包括列1和9，分别命名为individual和theta。

我尝试过：

results1<-sapply(cj1, "[",c("individual","theta") )

在[.data.frame(X[[i]], ...)

中出现错误：未定义的列被选中

library(dplyr)
> results1 <- lapply(cj1, function(x) x%>% select(individual,theta))

错误:

无法对不存在的列进行子集操作。
x 列 individual 不存在。
运行rlang::last_error()查看发生错误的位置。

我可以从一个数据框中减去这些列:

cj[[1]][c(1,9)]

我无法将此应用于整个列表。

- Vasile

Answer 1

您可以使用以下解决方案。我们使用.x引用您列表的每个单独元素。这里的.x可以是您要选择的每个数据框，其中我们只想选择c("individual","theta")这两列。但是，由于只有一个数据框包含这样的列名，因此我使用keep函数来实际保留仅包含所需列名的元素的数据框。请记住，在称为purrr-style公式的这种编码形式中，我们需要在.x之前使用~。因此，您可以使用map函数（相当于基本R中的lapply）并使用此语法将任何函数应用于每个单独元素（此处为数据框）。

library(purrr)

cj1 %>%
  map_if(~ all(c("individual","theta") %in% names(.x)), 
         ~ .x %>% select(individual, theta)) %>%
  keep(~ all(c("individual","theta") %in% names(.x)))

[[1]]
      individual      theta
1     a12TTT.pdf -0.5712111
2        a15.pdf  0.4187368
3        a17.pdf  0.4645337
4        a18.pdf  0.7604320
5        a21.pdf  1.4396103
6      a2TTT.pdf  0.9359639
7         a5.pdf  0.0742361
8        B11.pdf  0.4167832
9        B12.pdf  0.2325864
10       B13.pdf  1.6534525
11       B22.pdf  0.1789475
12       B24.pdf  1.3980443
13        B4.pdf  1.5300599
14        B7.pdf  0.3400874
15        B8.pdf  0.6169859
16    cw10-1.pdf -1.7324610
17    cw13-1.pdf -4.0618617
18 cw15-1TTT.pdf -0.3477007
19    cw17-1.pdf -1.2100996
20      cw18.pdf  0.2391456
21       cw3.pdf -1.8883642
22       cw4.pdf -0.2764515
23  cw7_1TTT.pdf -0.6114556

或者我们可以节省一行代码，使其更加简洁：

cj1 %>%
  keep(~ all(c("individual","theta") %in% names(.x))) %>%
  map(~ .x %>% select(individual, theta))

[[1]]
      individual      theta
1     a12TTT.pdf -0.5712111
2        a15.pdf  0.4187368
3        a17.pdf  0.4645337
4        a18.pdf  0.7604320
5        a21.pdf  1.4396103
6      a2TTT.pdf  0.9359639
7         a5.pdf  0.0742361
8        B11.pdf  0.4167832
9        B12.pdf  0.2325864
10       B13.pdf  1.6534525
11       B22.pdf  0.1789475
12       B24.pdf  1.3980443
13        B4.pdf  1.5300599
14        B7.pdf  0.3400874
15        B8.pdf  0.6169859
16    cw10-1.pdf -1.7324610
17    cw13-1.pdf -4.0618617
18 cw15-1TTT.pdf -0.3477007
19    cw17-1.pdf -1.2100996
20      cw18.pdf  0.2391456
21       cw3.pdf -1.8883642
22       cw4.pdf -0.2764515
23  cw7_1TTT.pdf -0.6114556

这里是另一种使用基础 R 的解决方案，语法略有不同。请注意，\(x) 相当于 function(x)，这是自R 4.1.0以来可用的新功能。

cj1 |>
  lapply(\(x) { 
    if(all(c("individual","theta") %in% names(x))) {
      `[`(x, c("individual","theta"))
    }
  }
) -> cj2

cj2 <- cj2[-which(sapply(cj2, is.null))] |> as.data.frame()

- Anoushiravan R

---

安纳什拉万，我需要用 cj1 替换所有的 .x 吗？ - Vasile

不，你要用cj1替换df并使用第二个解决方案。我也更新了我的解决方案。 - Anoushiravan R

谢谢，我使用了第一个解决方案（还没有看到你的评论），它起作用了。 - Vasile

第二个更为简洁。如果有需要我进一步解释的内容，请告诉我，我也添加了一些额外的细节。 - Anoushiravan R

Answer 2

在基础R中，您可以使用lapply尝试此解决方案 -

cols <- c("individual","theta")
lapply(cj1, function(x) if(all(cols %in% names(x))) x[cols])

#[[1]]
#      individual   theta
#1     a12TTT.pdf -0.5712
#2        a15.pdf  0.4187
#3        a17.pdf  0.4645
#4        a18.pdf  0.7604
#5        a21.pdf  1.4396
#6      a2TTT.pdf  0.9360
#7         a5.pdf  0.0742
#8        B11.pdf  0.4168
#9        B12.pdf  0.2326
#10       B13.pdf  1.6535
#11       B22.pdf  0.1789
#12       B24.pdf  1.3980
#13        B4.pdf  1.5301
#14        B7.pdf  0.3401
#15        B8.pdf  0.6170
#16    cw10-1.pdf -1.7325
#17    cw13-1.pdf -4.0619
#18 cw15-1TTT.pdf -0.3477
#19    cw17-1.pdf -1.2101
#20      cw18.pdf  0.2391
#21       cw3.pdf -1.8884
#22       cw4.pdf -0.2765
#23  cw7_1TTT.pdf -0.6115

#[[2]]
#NULL

如果您想要删除NULL列表，您可以添加Filter -

Filter(length, lapply(cj1, function(x) if(all(cols %in% names(x))) x[cols]))

- Ronak Shah