将列表列表转换为数据框
Posted
技术标签:
【中文标题】将列表列表转换为数据框【英文标题】：Convert list of lists to dataframe 【发布时间】：2017-06-07 20:34:24 【问题描述】：我有一个嵌套列表，名为 mylist，长度为 4。
此列表的每个元素都是一个实验：exp1.1、exp1.2、exp2.1 和 exp2.2。
每个实验都包含对四个植物生长阶段长度（以天为单位）的观察：EM-V6V6-R0R0-R4 和R4-R9。
每个成长阶段都被组织成一个带有year和mean的数据框。
这是完整的数据：
mylist=structure(list(exp1.1 = structure(list(`EM-V6` = structure(list(
year = 2011:2100, mean = c(34, 34, 32, 28, 25, 32, 32, 28,
27, 30, 32, 31, 33, 28, 26, 31, 33, 27, 34, 26, 28, 27, 27,
30, 29, 31, 34, 30, 26, 31, 33, 33, 27, 30, 28, 32, 31, 29,
32, 31, 25, 28, 28, 26, 32, 29, 26, 31, 28, 29, 30, 25, 27,
32, 27, 28, 28, 30, 24, 30, 29, 29, 29, 28, 26, 28, 26, 26,
28, 31, 30, 27, 26, 28, 25, 24, 24, 30, 27, 26, 26, 27, 26,
26, 24, 26, 28, 25, 30, 26)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `V6-R0` = structure(list(year = 2011:2100,
mean = c(30, 33, 33, 32, 29, 30, 32, 31, 32, 30, 33, 30,
32, 33, 33, 32, 29, 31, 32, 28, 31, 29, 36, 29, 30, 30, 33,
31, 33, 30, 34, 32, 29, 31, 28, 30, 30, 29, 34, 31, 32, 31,
30, 28, 32, 29, 29, 32, 29, 28, 29, 29, 32, 31, 27, 32, 29,
31, 29, 29, 30, 29, 29, 29, 28, 28, 30, 30, 30, 32, 29, 29,
30, 29, 29, 29, 28, 28, 29, 30, 29, 29, 29, 30, 28, 30, 30,
29, 29, 29)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R0-R4` = structure(list(year = 2011:2100,
mean = c(31, 32, 32, 33, 32, 32, 33, 31, 34, 32, 33, 33,
32, 31, 33, 31, 32, 32, 32, 30, 32, 31, 34, 30, 31, 32, 34,
33, 34, 32, 36, 33, 32, 32, 31, 30, 32, 32, 32, 32, 32, 32,
31, 30, 30, 31, 32, 32, 30, 30, 32, 31, 31, 32, 30, 32, 29,
32, 31, 30, 32, 30, 30, 31, 32, 30, 31, 30, 31, 32, 31, 31,
30, 30, 30, 31, 30, 30, 31, 30, 31, 30, 30, 30, 31, 32, 30,
31, 30, 30)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R4-R9` = structure(list(year = 2011:2100,
mean = c(27, 29, 28, 28, 27, 30, 29, 27, 30, 26, 30, 28,
29, 28, 29, 27, 29, 28, 25, 26, 26, 25, 27, 27, 27, 28, 30,
28, 29, 27, 29, 28, 29, 28, 26, 26, 28, 28, 30, 28, 27, 25,
26, 25, 25, 26, 26, 27, 25, 25, 26, 25, 27, 28, 24, 27, 25,
28, 26, 24, 27, 26, 27, 25, 26, 26, 24, 26, 25, 26, 24, 25,
25, 26, 26, 25, 25, 25, 25, 25, 26, 25, 25, 25, 25, 26, 26,
26, 25, 24)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame")), .Names = c("EM-V6", "V6-R0", "R0-R4",
"R4-R9")), exp1.2 = structure(list(`EM-V6` = structure(list(year = 2011:2100,
mean = c(34, 34, 32, 28, 25, 32, 32, 28, 27, 30, 32, 31,
33, 28, 26, 31, 33, 27, 34, 26, 28, 27, 27, 30, 29, 31, 34,
30, 26, 31, 33, 33, 27, 30, 28, 32, 31, 29, 32, 31, 25, 28,
28, 26, 32, 29, 26, 31, 28, 29, 30, 25, 27, 32, 27, 28, 28,
30, 24, 30, 29, 29, 29, 28, 26, 28, 26, 26, 28, 31, 30, 27,
26, 28, 25, 24, 24, 30, 27, 26, 26, 27, 26, 26, 24, 26, 28,
25, 30, 26)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `V6-R0` = structure(list(year = 2011:2100,
mean = c(30, 33, 33, 32, 29, 30, 32, 31, 32, 30, 33, 30,
32, 33, 33, 32, 29, 31, 32, 28, 31, 29, 36, 29, 30, 30, 33,
31, 33, 30, 34, 32, 29, 31, 28, 30, 30, 29, 34, 31, 32, 31,
30, 28, 32, 29, 29, 32, 29, 28, 29, 29, 32, 31, 27, 32, 29,
31, 29, 29, 30, 29, 29, 29, 28, 28, 30, 30, 30, 32, 29, 29,
30, 29, 29, 29, 28, 28, 29, 30, 29, 29, 29, 30, 28, 30, 30,
29, 29, 29)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R0-R4` = structure(list(year = 2011:2100,
mean = c(31, 32, 32, 33, 32, 32, 33, 31, 34, 32, 33, 33,
32, 31, 33, 31, 32, 32, 32, 30, 32, 31, 34, 30, 31, 32, 34,
33, 34, 32, 36, 33, 32, 32, 31, 30, 32, 32, 32, 32, 32, 32,
31, 30, 30, 31, 32, 32, 30, 30, 32, 31, 31, 32, 30, 32, 29,
32, 31, 30, 32, 30, 30, 31, 32, 30, 31, 30, 31, 32, 31, 31,
30, 30, 30, 31, 30, 30, 31, 30, 31, 30, 30, 30, 31, 32, 30,
31, 30, 30)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R4-R9` = structure(list(year = 2011:2100,
mean = c(27, 29, 28, 28, 27, 30, 29, 27, 30, 26, 30, 28,
29, 28, 29, 27, 29, 28, 25, 26, 26, 25, 27, 27, 27, 28, 30,
28, 29, 27, 29, 28, 29, 28, 26, 26, 28, 28, 30, 28, 27, 25,
26, 25, 25, 26, 26, 27, 25, 25, 26, 25, 27, 28, 24, 27, 25,
28, 26, 24, 27, 26, 27, 25, 26, 26, 24, 26, 25, 26, 24, 25,
25, 26, 26, 25, 25, 25, 25, 25, 26, 25, 25, 25, 25, 26, 26,
26, 25, 24)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame")), .Names = c("EM-V6", "V6-R0", "R0-R4",
"R4-R9")), exp2.1 = structure(list(`EM-V6` = structure(list(year = 2011:2100,
mean = c(34, 34, 32, 28, 25, 32, 32, 28, 27, 30, 32, 31,
33, 28, 26, 31, 33, 27, 34, 26, 28, 27, 27, 30, 29, 31, 34,
30, 26, 31, 33, 33, 27, 30, 28, 32, 31, 29, 32, 31, 25, 28,
28, 26, 32, 29, 26, 31, 28, 29, 30, 25, 27, 32, 27, 28, 28,
30, 24, 30, 29, 29, 29, 28, 26, 28, 26, 26, 28, 31, 30, 27,
26, 28, 25, 24, 24, 30, 27, 26, 26, 27, 26, 26, 24, 26, 28,
25, 30, 26)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `V6-R0` = structure(list(year = 2011:2100,
mean = c(30, 33, 33, 32, 29, 30, 32, 31, 32, 30, 33, 30,
32, 33, 33, 32, 29, 31, 32, 28, 31, 29, 36, 29, 30, 30, 33,
31, 33, 30, 34, 32, 29, 31, 28, 30, 30, 29, 34, 31, 32, 31,
30, 28, 32, 29, 29, 32, 29, 28, 29, 29, 32, 31, 27, 32, 29,
31, 29, 29, 30, 29, 29, 29, 28, 28, 30, 30, 30, 32, 29, 29,
30, 29, 29, 29, 28, 28, 29, 30, 29, 29, 29, 30, 28, 30, 30,
29, 29, 29)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R0-R4` = structure(list(year = 2011:2100,
mean = c(31, 32, 32, 33, 32, 32, 33, 31, 34, 32, 33, 33,
32, 31, 33, 31, 32, 32, 32, 30, 32, 31, 34, 30, 31, 32, 34,
33, 34, 32, 36, 33, 32, 32, 31, 30, 32, 32, 32, 32, 32, 32,
31, 30, 30, 31, 32, 32, 30, 30, 32, 31, 31, 32, 30, 32, 29,
32, 31, 30, 32, 30, 30, 31, 32, 30, 31, 30, 31, 32, 31, 31,
30, 30, 30, 31, 30, 30, 31, 30, 31, 30, 30, 30, 31, 32, 30,
31, 30, 30)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R4-R9` = structure(list(year = 2011:2100,
mean = c(27, 29, 28, 28, 27, 30, 29, 27, 30, 26, 30, 28,
29, 28, 29, 27, 29, 28, 25, 26, 26, 25, 27, 27, 27, 28, 30,
28, 29, 27, 29, 28, 29, 28, 26, 26, 28, 28, 30, 28, 27, 25,
26, 25, 25, 26, 26, 27, 25, 25, 26, 25, 27, 28, 24, 27, 25,
28, 26, 24, 27, 26, 27, 25, 26, 26, 24, 26, 25, 26, 24, 25,
25, 26, 26, 25, 25, 25, 25, 25, 26, 25, 25, 25, 25, 26, 26,
26, 25, 24)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame")), .Names = c("EM-V6", "V6-R0", "R0-R4",
"R4-R9")), exp2.2 = structure(list(`EM-V6` = structure(list(year = 2011:2100,
mean = c(34, 34, 32, 28, 25, 32, 32, 28, 27, 30, 32, 31,
33, 28, 26, 31, 33, 27, 34, 26, 28, 27, 27, 30, 29, 31, 34,
30, 26, 31, 33, 33, 27, 30, 28, 32, 31, 29, 32, 31, 25, 28,
28, 26, 32, 29, 26, 31, 28, 29, 30, 25, 27, 32, 27, 28, 28,
30, 24, 30, 29, 29, 29, 28, 26, 28, 26, 26, 28, 31, 30, 27,
26, 28, 25, 24, 24, 30, 27, 26, 26, 27, 26, 26, 24, 26, 28,
25, 30, 26)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `V6-R0` = structure(list(year = 2011:2100,
mean = c(30, 33, 33, 32, 29, 30, 32, 31, 32, 30, 33, 30,
32, 33, 33, 32, 29, 31, 32, 28, 31, 29, 36, 29, 30, 30, 33,
31, 33, 30, 34, 32, 29, 31, 28, 30, 30, 29, 34, 31, 32, 31,
30, 28, 32, 29, 29, 32, 29, 28, 29, 29, 32, 31, 27, 32, 29,
31, 29, 29, 30, 29, 29, 29, 28, 28, 30, 30, 30, 32, 29, 29,
30, 29, 29, 29, 28, 28, 29, 30, 29, 29, 29, 30, 28, 30, 30,
29, 29, 29)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R0-R4` = structure(list(year = 2011:2100,
mean = c(31, 32, 32, 33, 32, 32, 33, 31, 34, 32, 33, 33,
32, 31, 33, 31, 32, 32, 32, 30, 32, 31, 34, 30, 31, 32, 34,
33, 34, 32, 36, 33, 32, 32, 31, 30, 32, 32, 32, 32, 32, 32,
31, 30, 30, 31, 32, 32, 30, 30, 32, 31, 31, 32, 30, 32, 29,
32, 31, 30, 32, 30, 30, 31, 32, 30, 31, 30, 31, 32, 31, 31,
30, 30, 30, 31, 30, 30, 31, 30, 31, 30, 30, 30, 31, 32, 30,
31, 30, 30)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame"), `R4-R9` = structure(list(year = 2011:2100,
mean = c(27, 29, 28, 28, 27, 30, 29, 27, 30, 26, 30, 28,
29, 28, 29, 27, 29, 28, 25, 26, 26, 25, 27, 27, 27, 28, 30,
28, 29, 27, 29, 28, 29, 28, 26, 26, 28, 28, 30, 28, 27, 25,
26, 25, 25, 26, 26, 27, 25, 25, 26, 25, 27, 28, 24, 27, 25,
28, 26, 24, 27, 26, 27, 25, 26, 26, 24, 26, 25, 26, 24, 25,
25, 26, 26, 25, 25, 25, 25, 25, 26, 25, 25, 25, 25, 26, 26,
26, 25, 24)), .Names = c("year", "mean"), row.names = c(NA,
-90L), class = "data.frame")), .Names = c("EM-V6", "V6-R0", "R0-R4",
"R4-R9"))), .Names = c("exp1.1", "exp1.2", "exp2.1", "exp2.2"
))
我需要做的是将此嵌套列表“取消列出”到如下所示的数据框中：
YEAR EXP EM-V6 V6-R0 R0-R4 R4-R9
2011 exp1.1 34 30 31 27
2011 exp1.2 34 30 31 27
2011 exp2.1 34 30 31 27
2011 exp1.1 34 30 31 27
这意味着：
- first year, first experiment, and growth stages.
- first year, second experiment and growth stages.
- first year, third experiment and growth stages
- first year, fourth experiment and growth stages
- second year, first experiment and growth stages
等等。
如何执行该数据转换？
【问题讨论】：
@akrun。你确定这个问题是骗人的吗？另一个 Q How to reshape data from long to wide format? 不处理嵌套列表。 我同意，其他问题解决了完全不同的请求。请考虑将此问题取消标记为重复的@akrun。 @UweBlock 它已经被删除了。 【参考方案1】：我们可以使用tidyverse，代码更简洁易读
library(dplyr)
library(tidyr)
library(purrr)
res1 <- mylist %>%
#bind the inner datasets and create an id column
map(bind_rows, .id = "id") %>%
#bind the outer datasets and create an EXP column
bind_rows(.id = "EXP") %>%
#reshape to wide format
spread(id, mean)
head(res1, 4)
# EXP year EM-V6 R0-R4 R4-R9 V6-R0
#1 exp1.1 2011 34 31 27 30
#2 exp1.1 2012 34 32 29 33
#3 exp1.1 2013 32 32 28 33
#4 exp1.1 2014 28 33 28 32
或者我们可以通过使用lapply 循环遍历mylist 来解决此问题，然后通过cbind 使用内部names 元素的names 创建一个新列'name'使用Map，然后rbind list 元素和do.call(rbind，现在再做一个Map 以基于'mylist' 的names、rbind、list 元素和reshape 创建一个新列base R 将其转换为“宽”
res <- do.call(rbind, Map(cbind, lapply(mylist, function(x)
do.call(rbind, Map(cbind, x, name = names(x)))), EXP= names(mylist)))
res2 <- reshape(res, idvar = c("year", "EXP"),
timevar = "name", direction = "wide")
row.names(res2) <- NULL
head(res2, 4)
# year EXP mean.EM-V6 mean.V6-R0 mean.R0-R4 mean.R4-R9
#1 2011 exp1.1 34 30 31 27
#2 2012 exp1.1 34 33 32 29
#3 2013 exp1.1 32 33 32 28
#4 2014 exp1.1 28 32 33 28
注意：未使用外部包 (100% base R)
或使用 reshape2 中的 dcast 转换为“宽”格式
library(reshape2)
res2 <- dcast(res, year + EXP~name, value.var = "mean")
head(res2, 4)
# year EXP EM-V6 V6-R0 R0-R4 R4-R9
#1 2011 exp1.1 34 30 31 27
#2 2011 exp1.2 34 30 31 27
#3 2011 exp2.1 34 30 31 27
#4 2011 exp2.2 34 30 31 27
【讨论】：
谢谢，它有效！我很高兴我问了这个问题，因为答案远远超出了我的 R 技能。 @thiagoveloso 我还更新了一个不使用外部包的解决方案。【参考方案2】：使用data.table-package 中的rbindlist 两次的替代方法：
library(data.table)
# bind the dataframes in the 'listed lists' together and include the year with the 'id'-parameter
# the resulting 'data.table's are returned as a list
step1 <- lapply(mylist, rbindlist, id = 'stages')
# bind the resulting list together and include the experiment id
step2 <- rbindlist(step1, id = 'experiment')
# reshape to wide format
dcast(step2, year + experiment ~ stages, value.var = 'mean')
或者一口气：
dcast(rbindlist(lapply(mylist, rbindlist, id = 'stages'), id = 'experiment'),
year + experiment ~ stages, value.var = 'mean')
给出：
year experiment EM-V6 R0-R4 R4-R9 V6-R0
1: 2011 exp1.1 34 31 27 30
2: 2011 exp1.2 34 31 27 30
3: 2011 exp2.1 34 31 27 30
4: 2011 exp2.2 34 31 27 30
5: 2012 exp1.1 34 32 29 33
---
356: 2099 exp2.2 30 30 25 29
357: 2100 exp1.1 26 30 24 29
358: 2100 exp1.2 26 30 24 29
359: 2100 exp2.1 26 30 24 29
360: 2100 exp2.2 26 30 24 29
【讨论】：
它也可以，非常感谢。我会接受你的回答，因为它使用了我已经在我的代码中加载的包（无需加载额外的包）。【参考方案3】：备用tidyverse：
library(tidyverse)
map_df(mylist, ~bind_rows(., .id="id"), .id="EXP") %>%
spread(id, mean)
【讨论】：
它还显示了不同且更简洁的 purrr 语法 一个上升有点意味着至少有一个人认为它已经足够不同了。以上是关于将列表列表转换为数据框的主要内容，如果未能解决你的问题，请参考以下文章