在 mutate 中遍历嵌套列表并提取特定列表元素

huangapple
go评论94阅读模式
英文:

map over a nested list in mutate and extract specific list elements

问题

  1. 我有一些看起来像这样的数据：
  3.        price unleveragedData
  4.        <dbl> <list>         
  5.     1 450000 <list [5]>     
  6.     2 400000 <list [5]>     
  7.     3 400000 <list [5]>     
  8.     4 397000 <list [5]>     
  9.     5 750000 <list [5]>     
  10.     6 550000 <list [5]>
  12. 我试图将列表的第 `5` 个元素放入新列中，即：
  14.        price unleveragedData element5
  15.        <dbl> <list>         
  16.     1 450000 <list [5]>         -----
  17.     2 400000 <list [5]>         -----
  18.     3 400000 <list [5]>     
  19.     4 397000 <list [5]>     
  20.     5 750000 <list [5]>     
  21.     6 550000 <list [5]>         -----
  23. 使用以下代码：
  25.     df$unleveragedData %>%
  26.       map(., ~ pluck(., c(5)))
  28. 我可以得到我想要的输出：
  30.     [[1]]
  31.     [1] 260551.4
  33.     [[2]]
  34.     [1] 330786.9
  36.     [[3]]
  37.     [1] 330786.9
  39.     [[4]]
  40.     [1] 287739.3
  42.     [[5]]
  43.     [1] 566416
  45.     [[6]]
  46.     [1] 271879.7
  48. 然而，在 `mutate` 函数内部，我无法让它正常工作。
  50.     df %>%
  51.       mutate(
  52.         element5 = map(unleveragedData, ~ map_dbl(., pluck(., c(5))))
  53.       )
  55. 数据：
  57.     df = structure(list(price = c(450000, 400000, 400000, 397000, 750000, 
  58.     550000), unleveragedData = list(list(-0.0547083151944441, c(-450000, 
  59.     15533.28, 16475.2128, 17473.760928, 18532.32229728, 280205.849576444
  60.     ), "450000-0.08", structure(list(` ` = c("Revenue", "Vacancy", 
  61.     "Gross Revenue", "Operating Expenses", "Net Operating Income"), Year1 = c(16560, 828, 15732, 199, 15533), Year2 = c(17554, 
  62.     878, 16676, 201, 16475), Year3 = c(18607, 930, 17676, 203, 17474
  63.     ), Year4 = c(19723, 986, 18737, 205, 18532), Year5 = c(20907, 
  64.     1045, 19861, 207, 19654), Year6 = c(22161, 1108, 21053, 209, 
  65.     20844), purchasePriceCapRate = c("450000-0.08", "450000-0.08", 
  66.     "450000-0.08", "450000-0.08", "450000-0.08")), row.names = c(NA, 
  67.     -5L), class = "data.frame"), 260551.350870592), list(0.0224165566243759, 
  68.         c(-400000, 19720.512, 20916.35712, 22184.0790912, 23527.991786112, 
  69.         355739.600331834), "400000-0.08", structure(list(` ` = c("Revenue", 
  70.         "Vacancy", "Gross Revenue", "Operating Expenses", "Net Operating Income"
  71.         ), Year1 = c(21024, 1051, 19973, 252, 19721), Year2 = c(22285, 
  72.         1114, 21171, 255, 20916), Year3 = c(23623, 1181, 22441, 257, 
  73.         22184), Year4 = c(25040, 1252, 23788, 260, 23528), Year5 = c(26542, 
  74.         1327, 25215, 263, 24953), Year6 = c(28135, 1407, 26728, 265, 
  75.         26463), purchasePriceCapRate = c("400000-0.08", "400000-0.08", 
  76.         "400000-0.08", "400000-0.08", "400000-0.08")), row.names = c(NA, 
  77.         -5L), class = "data.frame"), 330786.932409621), list(0.0224165566243759, 
  78.         c(-400000, 19720.512, 20916.35712, 22184.0790912, 23527.991786112, 
  79.         355739.600331834), "400000-0.08", structure(list(` ` = c("Revenue", 
  80.         "Vacancy", "Gross Revenue", "Operating Expenses", "Net Operating Income"
  81.         ), Year1 = c(21024, 1051, 19973, 252, 19721), Year2 = c(22285, 
  82.         1114, 21171, 255, 20916), Year3 = c(23623, 1181, 22441, 257, 
  83.         22184), Year4 = c(25040, 1252, 23788, 260, 23528), Year5 = c(26542, 
  84.         1327, 25215, 263, 24953), Year6 = c(28135, 1407, 26728, 265, 
  85.         26463), purchasePriceCapRate = c("400000-0.08", "400000-0.08", 
  86.         "400000-0.08", "400000-0.08", "400000-0.08")), row.names = c(NA, 
  87.         -5L), class = "data.frame"), 330786.932409621), list(-0.00700507916565851, 
  88.         c(-397000, 17154.144, 18194.36544, 19297.1098944, 20466.129841344, 
  89.         309444.720836595), "397000-0.08", structure(list(` ` = c("Revenue", 
  90.         "Vacancy", "Gross Revenue", "Operating Expenses", "Net Operating Income"
  91.         ), Year1 = c(18288, 914, 17374, 219, 17154), Year2 = c(19385, 
  92.         969, 18416, 222, 18194), Year3 = c(20548, 1027, 195
  94. <details>
  95. <summary>英文:</summary>
  97. I have some data which looks like:
  99.        price unleveragedData
  100.        &lt;dbl&gt; &lt;list&gt;         
  101.     1 450000 &lt;list [5]&gt;     
  102.     2 400000 &lt;list [5]&gt;     
  103.     3 400000 &lt;list [5]&gt;     
  104.     4 397000 &lt;list [5]&gt;     
  105.     5 750000 &lt;list [5]&gt;     
  106.     6 550000 &lt;list [5]&gt;
  108. I am trying to put into a new column the element `5` of the lists - i.e.
  110.        price unleveragedData element5
  111.        &lt;dbl&gt; &lt;list&gt;         
  112.     1 450000 &lt;list [5]&gt;         -----
  113.     2 400000 &lt;list [5]&gt;         -----
  114.     3 400000 &lt;list [5]&gt;     
  115.     4 397000 &lt;list [5]&gt;     
  116.     5 750000 &lt;list [5]&gt;     
  117.     6 550000 &lt;list [5]&gt;         -----
  119. Using the following:
  121.     df$unleveragedData %&gt;% 
  122.       map(., ~ pluck(., c(5)))
  124. I can get the output I want:
  126.     [[1]]
  127.     [1] 260551.4
  128.     
  129.     [[2]]
  130.     [1] 330786.9
  131.     
  132.     [[3]]
  133.     [1] 330786.9
  134.     
  135.     [[4]]
  136.     [1] 287739.3
  137.     
  138.     [[5]]
  139.     [1] 566416
  140.     
  141.     [[6]]
  142.     [1] 271879.7
  144. However, inside the `mutate` function I can&#39;t get it to work.
  146.     df %&gt;% 
  147.       mutate(
  148.         element5 = map(unleveragedData, ~ map_dbl(., pluck(., c(5))))
  149.       )
  151. Data:
  153.     df = structure(list(price = c(450000, 400000, 400000, 397000, 750000, 
  154.     550000), unleveragedData = list(list(-0.0547083151944441, c(-450000, 
  155.     15533.28, 16475.2128, 17473.760928, 18532.32229728, 280205.849576444
  156.     ), &quot;450000-0.08&quot;, structure(list(` ` = c(&quot;Revenue&quot;, &quot;Vacancy&quot;, 
  157.     &quot;Gross Revenue&quot;, &quot;Operating Expenses&quot;, &quot;Net Operating Income&quot;
  158.     ), Year1 = c(16560, 828, 15732, 199, 15533), Year2 = c(17554, 
  159.     878, 16676, 201, 16475), Year3 = c(18607, 930, 17676, 203, 17474
  160.     ), Year4 = c(19723, 986, 18737, 205, 18532), Year5 = c(20907, 
  161.     1045, 19861, 207, 19654), Year6 = c(22161, 1108, 21053, 209, 
  162.     20844), purchasePriceCapRate = c(&quot;450000-0.08&quot;, &quot;450000-0.08&quot;, 
  163.     &quot;450000-0.08&quot;, &quot;450000-0.08&quot;, &quot;450000-0.08&quot;)), row.names = c(NA, 
  164.     -5L), class = &quot;data.frame&quot;), 260551.350870592), list(0.0224165566243759, 
  165.         c(-400000, 19720.512, 20916.35712, 22184.0790912, 23527.991786112, 
  166.         355739.600331834), &quot;400000-0.08&quot;, structure(list(` ` = c(&quot;Revenue&quot;, 
  167.         &quot;Vacancy&quot;, &quot;Gross Revenue&quot;, &quot;Operating Expenses&quot;, &quot;Net Operating Income&quot;
  168.         ), Year1 = c(21024, 1051, 19973, 252, 19721), Year2 = c(22285, 
  169.         1114, 21171, 255, 20916), Year3 = c(23623, 1181, 22441, 257, 
  170.         22184), Year4 = c(25040, 1252, 23788, 260, 23528), Year5 = c(26542, 
  171.         1327, 25215, 263, 24953), Year6 = c(28135, 1407, 26728, 265, 
  172.         26463), purchasePriceCapRate = c(&quot;400000-0.08&quot;, &quot;400000-0.08&quot;, 
  173.         &quot;400000-0.08&quot;, &quot;400000-0.08&quot;, &quot;400000-0.08&quot;)), row.names = c(NA, 
  174.         -5L), class = &quot;data.frame&quot;), 330786.932409621), list(0.0224165566243759, 
  175.         c(-400000, 19720.512, 20916.35712, 22184.0790912, 23527.991786112, 
  176.         355739.600331834), &quot;400000-0.08&quot;, structure(list(` ` = c(&quot;Revenue&quot;, 
  177.         &quot;Vacancy&quot;, &quot;Gross Revenue&quot;, &quot;Operating Expenses&quot;, &quot;Net Operating Income&quot;
  178.         ), Year1 = c(21024, 1051, 19973, 252, 19721), Year2 = c(22285, 
  179.         1114, 21171, 255, 20916), Year3 = c(23623, 1181, 22441, 257, 
  180.         22184), Year4 = c(25040, 1252, 23788, 260, 23528), Year5 = c(26542, 
  181.         1327, 25215, 263, 24953), Year6 = c(28135, 1407, 26728, 265, 
  182.         26463), purchasePriceCapRate = c(&quot;400000-0.08&quot;, &quot;400000-0.08&quot;, 
  183.         &quot;400000-0.08&quot;, &quot;400000-0.08&quot;, &quot;400000-0.08&quot;)), row.names = c(NA, 
  184.         -5L), class = &quot;data.frame&quot;), 330786.932409621), list(-0.00700507916565851, 
  185.         c(-397000, 17154.144, 18194.36544, 19297.1098944, 20466.129841344, 
  186.         309444.720836595), &quot;397000-0.08&quot;, structure(list(` ` = c(&quot;Revenue&quot;, 
  187.         &quot;Vacancy&quot;, &quot;Gross Revenue&quot;, &quot;Operating Expenses&quot;, &quot;Net Operating Income&quot;
  188.         ), Year1 = c(18288, 914, 17374, 219, 17154), Year2 = c(19385, 
  189.         969, 18416, 222, 18194), Year3 = c(20548, 1027, 19521, 224, 
  190.         19297), Year4 = c(21781, 1089, 20692, 226, 20466), Year5 = c(23088, 
  191.         1154, 21934, 228, 21705), Year6 = c(24473, 1224, 23250, 231, 
  192.         23019), purchasePriceCapRate = c(&quot;397000-0.08&quot;, &quot;397000-0.08&quot;, 
  193.         &quot;397000-0.08&quot;, &quot;397000-0.08&quot;, &quot;397000-0.08&quot;)), row.names = c(NA, 
  194.         -5L), class = &quot;data.frame&quot;), 287739.317917958), list(0.00205549716813258, 
  195.         c(-750000, 33768, 35815.68, 37986.4368, 40287.657168, 609143.15125314
  196.         ), &quot;750000-0.08&quot;, structure(list(` ` = c(&quot;Revenue&quot;, &quot;Vacancy&quot;, 
  197.         &quot;Gross Revenue&quot;, &quot;Operating Expenses&quot;, &quot;Net Operating Income&quot;
  198.         ), Year1 = c(36000, 1800, 34200, 432, 33768), Year2 = c(38160, 
  199.         1908, 36252, 436, 35816), Year3 = c(40450, 2022, 38427, 441, 
  200.         37986), Year4 = c(42877, 2144, 40733, 445, 40288), Year5 = c(45449, 
  201.         2272, 43177, 450, 42727), Year6 = c(48176, 2409, 45767, 454, 
  202.         45313), purchasePriceCapRate = c(&quot;750000-0.08&quot;, &quot;750000-0.08&quot;, 
  203.         &quot;750000-0.08&quot;, &quot;750000-0.08&quot;, &quot;750000-0.08&quot;)), row.names = c(NA, 
  204.         -5L), class = &quot;data.frame&quot;), 566415.98015346), list(-0.0866171399087425, 
  205.         c(-550000, 16208.64, 17191.5264, 18233.489664, 19338.07544064, 
  206.         292388.712601507), &quot;550000-0.08&quot;, structure(list(` ` = c(&quot;Revenue&quot;, 
  207.         &quot;Vacancy&quot;, &quot;Gross Revenue&quot;, &quot;Operating Expenses&quot;, &quot;Net Operating Income&quot;
  208.         ), Year1 = c(17280, 864, 16416, 207, 16209), Year2 = c(18317, 
  209.         916, 17401, 209, 17192), Year3 = c(19416, 971, 18445, 212, 
  210.         18233), Year4 = c(20581, 1029, 19552, 214, 19338), Year5 = c(21816, 
  211.         1091, 20725, 216, 20509), Year6 = c(23125, 1156, 21968, 218, 
  212.         21750), purchasePriceCapRate = c(&quot;550000-0.08&quot;, &quot;550000-0.08&quot;, 
  213.         &quot;550000-0.08&quot;, &quot;550000-0.08&quot;, &quot;550000-0.08&quot;)), row.names = c(NA, 
  214.         -5L), class = &quot;data.frame&quot;), 271879.670473661))), class = c(&quot;rowwise_df&quot;, 
  215.     &quot;tbl_df&quot;, &quot;tbl&quot;, &quot;data.frame&quot;), row.names = c(NA, -6L), groups = structure(list(
  216.         .rows = structure(list(1L, 2L, 3L, 4L, 5L, 6L), ptype = integer(0), class = c(&quot;vctrs_list_of&quot;, 
  217.         &quot;vctrs_vctr&quot;, &quot;list&quot;))), row.names = c(NA, -6L), class = c(&quot;tbl_df&quot;, 
  218.     &quot;tbl&quot;, &quot;data.frame&quot;)))
  220. </details>
  223. # 答案1
  224. **得分**: 4
  226. ```markdown
  227. 存在一个`rowwise`分组，如果要使用`map`，应该将其解除分组(`ungroup`)
  229. ```R
  230. library(dplyr)
  231. library(purrr)
  232. df %>% 
  233.   ungroup() %>%
  234.   mutate(
  235.     element5 = map_dbl(unleveragedData, ~ nth(.x, 5))
  236.   )

-输出

  1. # 一个 tibble: 6 × 3
  2.    price unleveragedData element5
  3.    <dbl> <list>             <dbl>
  4. 1 450000 <list [5]>       260551.
  5. 2 400000 <list [5]>       330787.
  6. 3 400000 <list [5]>       330787.
  7. 4 397000 <list [5]>       287739.
  8. 5 750000 <list [5]>       566416.
  9. 6 550000 <list [5]>       271880.

另外，由于是rowwise，我们也可以直接提取

  1. df %>%
  2.    mutate(element5 = nth(unleveragedData, 5)) %>%
  3.    ungroup()

-输出

  1. # 一个 tibble: 6 × 3
  2.    price unleveragedData element5
  3.    <dbl> <list>             <dbl>
  4. 1 450000 <list [5]>       260551.
  5. 2 400000 <list [5]>       330787.
  6. 3 400000 <list [5]>       330787.
  7. 4 397000 <list [5]>       287739.
  8. 5 750000 <list [5]>       566416.
  9. 6 550000 <list [5]>       271880.

或者使用 pluck

  1. df %>% 
  2.    mutate(element5 = pluck(unleveragedData, 5)) %>%
  3.    ungroup()

-输出

  1. # 一个 tibble: 6 × 3
  2.    price unleveragedData element5
  3.    <dbl> <list>             <dbl>
  4. 1 450000 <list [5]>       260551.
  5. 2 400000 <list [5]>       330787.
  6. 3 400000 <list [5]>       330787.
  7. 4 397000 <list [5]>       287739.
  8. 5 750000 <list [5]>       566416.
  9. 6 550000 <list [5]>       271880.

在rowwise上使用map也是可能的，只需要将其包装在pick中

  1. df %>%
  2.    mutate(element5 = map_dbl(pick(unleveragedData), pluck, 5))

-输出

  1. # 一个 tibble: 6 × 3
  2. # Rowwise: 
  3.    price unleveragedData element5
  4.    <dbl> <list>             <dbl>
  5. 1 450000 <list [5]>       260551.
  6. 2 400000 <list [5]>       330787.
  7. 3 400000 <list [5]>       330787.
  8. 4 397000 <list [5]>       287739.
  9. 5 750000 <list [5]>       566416.
  10. 6 550000 <list [5]>       271880.
  2. <details>
  3. <summary>英文:</summary>
  4. There is a `rowwise` grouping, which should be `ungroup`ed if we want to use `map`

library(dplyr)
library(purrr)
df %>%
ungroup %>%
mutate(
element5 = map_dbl(unleveragedData, ~ nth(.x, 5))
)

  1. -output

A tibble: 6 × 3

price unleveragedData element5
<dbl> <list> <dbl>
1 450000 <list [5]> 260551.
2 400000 <list [5]> 330787.
3 400000 <list [5]> 330787.
4 397000 <list [5]> 287739.
5 750000 <list [5]> 566416.
6 550000 <list [5]> 271880.

  2. ---
  3. Also, as it is `rowwise`, we can directly extract as well

df %>%
mutate(element5 = nth(unleveragedData, 5)) %>%
ungroup

  1. -output

A tibble: 6 × 3

price unleveragedData element5
<dbl> <list> <dbl>
1 450000 <list [5]> 260551.
2 400000 <list [5]> 330787.
3 400000 <list [5]> 330787.
4 397000 <list [5]> 287739.
5 750000 <list [5]> 566416.
6 550000 <list [5]> 271880.

  2. Or with `pluck`

df %>%
mutate(element5 = pluck(unleveragedData, 5)) %>%
ungroup

  1. -output

A tibble: 6 × 3

price unleveragedData element5
<dbl> <list> <dbl>
1 450000 <list [5]> 260551.
2 400000 <list [5]> 330787.
3 400000 <list [5]> 330787.
4 397000 <list [5]> 287739.
5 750000 <list [5]> 566416.
6 550000 <list [5]> 271880.

  2. ---
  3. It is possible to do this in `map` on a `rowwise`, if we wrap it in `pick`

df %>%
mutate(element5 = map_dbl(pick(unleveragedData), pluck, 5))

  1. -output

A tibble: 6 × 3

Rowwise:

price unleveragedData element5
<dbl> <list> <dbl>
1 450000 <list [5]> 260551.
2 400000 <list [5]> 330787.
3 400000 <list [5]> 330787.
4 397000 <list [5]> 287739.
5 750000 <list [5]> 566416.
6 550000 <list [5]> 271880.

  2. </details>

huangapple
  • 本文由 发表于 2023年2月27日 09:12:02
  • 转载请务必保留本文链接：https://go.coder-hub.com/75576055.html
  • dplyr
  • purrr
  • r
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定