使用dplyr，可以轻松查看将数据集分组的组。如何使用data.table来获得这些组或定义一种显示它们的打印方法。
示例：
df <- read.table(text = "X Y Goal
32 12 1
13 42 0
55 33 0", header = TRUE)
setDT(df)
df[,lapply(.SD,sum),.(X,Y)][]
输出：
X Y Goal
1: 32 12 1
2: 13 42 0
3: 55 33 0
在dplyr中进行类似操作可以更直观地显示各组，如何在data.table中看到相同的内容？
更新：的确，正如@Gregor在下面指出的那样，summaris(z)e
dyplr中的家庭会自动取消分组。但是，如何获得data.table来打印如下所示的内容。
示例dplyr输出。我正在寻找语句data.table的{{1}}替代项。
# Groups: Species [3]谢谢！
答案 0 :(得分：2)
实际上，如果将目标R作为标题阅读以及OP在以下方面的特定需求，则考虑使用by的基R的tapply（identity的面向对象包装器）。
显示组的打印方法
...可以轻松查看数据集的分组方式
实际上，您可以运行任何操作，例如summary，而by输出仍然具有带分节符的组头！
# ONE GROUP
by(df, random_df$group, identity)
# TWO GROUPS
with(df, by(random_df, list(group1, group2), identity))
# THREE GROUPS
with(df, by(random_df, list(group1, group2, group3), identity))
下面是带有随机种子数据的演示：
set.seed(3222018)
data_tools <- c("sas", "stata", "spss", "python", "r", "julia")
random_df <- data.frame(
group = sample(data_tools, 50, replace=TRUE),
int = sample(1:5, 50, replace=TRUE),
num = rnorm(50),
char = rep(letters[1:3], 20)[1:50],
bool = sample(c(TRUE, FALSE), 50, replace=TRUE),
data = as.Date(sample(1:as.integer(Sys.Date()), 50, replace=TRUE), origin="1970-01-01")
)
by(random_df, random_df$group, identity)
# random_df$group: julia
# group int num char bool data
# 9 julia 1 0.55209060 c TRUE 2003-01-08
# 12 julia 5 0.25581573 c FALSE 1973-01-20
# 20 julia 2 -0.44872825 b FALSE 1981-12-20
# 30 julia 1 0.09790121 c TRUE 1981-07-11
# 31 julia 1 1.75711081 a FALSE 1984-05-23
# 35 julia 3 -1.78797757 b TRUE 1992-10-23
# 39 julia 5 -1.83548253 c TRUE 1982-04-18
# 44 julia 3 0.52640304 b FALSE 2016-07-28
# 50 julia 2 0.05759068 b FALSE 1987-07-09
# ------------------------------------------------------------
# random_df$group: python
# group int num char bool data
# 8 python 3 -0.75471461 b TRUE 2018-04-08
# 15 python 4 0.03756282 c FALSE 2006-03-29
# 19 python 1 0.43025626 a FALSE 2005-03-30
# 22 python 4 -0.28019319 a FALSE 1997-02-19
# 24 python 2 -0.05318753 c FALSE 1977-08-02
# 45 python 2 -0.11575270 c TRUE 2006-06-23
# 48 python 2 -0.29651827 c TRUE 2015-07-13
# 49 python 3 -0.45816745 a FALSE 1991-09-09
# ------------------------------------------------------------
# random_df$group: r
# group int num char bool data
# 5 r 4 -0.1748250 b FALSE 2005-04-20
# 14 r 2 -0.5868782 b TRUE 2005-12-30
# 25 r 5 -0.5108906 a TRUE 1985-08-26
# 33 r 1 0.3580581 c FALSE 1983-08-16
# 34 r 3 -2.9556022 a TRUE 1998-07-24
# 36 r 5 -2.5008645 c TRUE 1990-12-30
# 43 r 5 -2.1710319 a TRUE 1987-09-29
# ------------------------------------------------------------
# ...
with(random_df, by(random_df, list(group, char), identity))
# : julia
# : a
# group int num char bool data
# 31 julia 1 1.757111 a FALSE 1984-05-23
# ------------------------------------------------------------
# : python
# : a
# group int num char bool data
# 19 python 1 0.4302563 a FALSE 2005-03-30
# 22 python 4 -0.2801932 a FALSE 1997-02-19
# 49 python 3 -0.4581675 a FALSE 1991-09-09
# ------------------------------------------------------------
# : r
# : a
# group int num char bool data
# 25 r 5 -0.5108906 a TRUE 1985-08-26
# 34 r 3 -2.9556022 a TRUE 1998-07-24
# 43 r 5 -2.1710319 a TRUE 1987-09-29
# ------------------------------------------------------------
# : sas
# : a
# group int num char bool data
# 28 sas 2 -0.6086176 a FALSE 2019-01-26
# 40 sas 4 0.5408948 a FALSE 2004-10-24
# 46 sas 1 0.7266734 a TRUE 1996-09-05
# ------------------------------------------------------------
# : spss
# : a
# group int num char bool data
# 1 spss 1 -0.06494476 a TRUE 1999-07-10
# 7 spss 5 1.93460991 a TRUE 1974-04-17
# 16 spss 1 -0.55380012 a FALSE 2010-04-06
# 37 spss 1 0.17309091 a TRUE 1999-01-22
# ------------------------------------------------------------
# : stata
# : a
# group int num char bool data
# 4 stata 1 -1.0498222 a TRUE 1981-05-01
# 10 stata 1 1.9223104 a TRUE 1996-07-23
# 13 stata 1 -0.8381546 a TRUE 1989-01-12
# ------------------------------------------------------------
# ...
with(random_df, by(random_df, list(group, char, int), identity))
# : julia
# : a
# : 1
# group int num char bool data
# 31 julia 1 1.757111 a FALSE 1984-05-23
# ------------------------------------------------------------
# : python
# : a
# : 1
# group int num char bool data
# 19 python 1 0.4302563 a FALSE 2005-03-30
# ------------------------------------------------------------
# : r
# : a
# : 1
# NULL
# ------------------------------------------------------------
# : sas
# : a
# : 1
# group int num char bool data
# 46 sas 1 0.7266734 a TRUE 1996-09-05
# ------------------------------------------------------------
# : spss
# : a
# : 1
# group int num char bool data
# 1 spss 1 -0.06494476 a TRUE 1999-07-10
# 16 spss 1 -0.55380012 a FALSE 2010-04-06
# 37 spss 1 0.17309091 a TRUE 1999-01-22
# ------------------------------------------------------------
# : stata
# : a
# : 1
# group int num char bool data
# 4 stata 1 -1.0498222 a TRUE 1981-05-01
# 10 stata 1 1.9223104 a TRUE 1996-07-23
# 13 stata 1 -0.8381546 a TRUE 1989-01-12
# ------------------------------------------------------------
# ...
为说明起见，您还可以在by内运行{em> any 操作，包括data.table聚合，以仍然返回分组的标题和分节符打印输出：
library(data.table)
...
by(random_df, random_df$group, function(sub)
data.table(sub)[, list(mean_num = mean(num),
median_num = median(num),
min_num = min(num),
max_num = max(num),
total_num = sum(num)), by = char]
)
# random_df$group: julia
# char mean_num median_num min_num max_num total_num
# 1: c -0.2324187 0.1768585 -1.835483 0.5520906 -0.929675
# 2: b -0.4131780 -0.1955688 -1.787978 0.5264030 -1.652712
# 3: a 1.7571108 1.7571108 1.757111 1.7571108 1.757111
# ------------------------------------------------------------------------------
# random_df$group: python
# char mean_num median_num min_num max_num total_num
# 1: b -0.7547146 -0.75471461 -0.7547146 -0.75471461 -0.7547146
# 2: c -0.1069739 -0.08447011 -0.2965183 0.03756282 -0.4278957
# 3: a -0.1027015 -0.28019319 -0.4581675 0.43025626 -0.3081044
# ------------------------------------------------------------------------------
# random_df$group: r
# char mean_num median_num min_num max_num total_num
# 1: b -0.3808516 -0.3808516 -0.5868782 -0.1748250 -0.7617031
# 2: a -1.8791749 -2.1710319 -2.9556022 -0.5108906 -5.6375247
# 3: c -1.0714032 -1.0714032 -2.5008645 0.3580581 -2.1428064
# ------------------------------------------------------------------------------
# random_df$group: sas
# char mean_num median_num min_num max_num total_num
# 1: c 0.08215953 0.163064560 -1.3670420 1.3695510 0.3286381
# 2: b -0.16423225 0.005903355 -1.3995380 0.7308023 -0.6569290
# 3: a 0.21965022 0.540894826 -0.6086176 0.7266734 0.6589507
# ------------------------------------------------------------------------------
# random_df$group: spss
# char mean_num median_num min_num max_num total_num
# 1: a 0.3722390 0.05407307 -0.5538001 1.9346099 1.488956
# 2: b 0.5085635 0.73430008 -0.3003130 0.8659667 2.034254
# 3: c -1.2909541 -1.29095413 -1.2909541 -1.2909541 -1.290954
# ------------------------------------------------------------------------------
# random_df$group: stata
# char mean_num median_num min_num max_num total_num
# 1: a 0.01144453 -0.8381546 -1.0498222 1.9223104 0.0343336
# 2: b -0.65927260 -0.6592726 -1.1870952 -0.1314500 -1.3185452
# 3: c 0.94910811 0.9491081 0.9491081 0.9491081 0.9491081
答案 1 :(得分：1)
# Groups: Species [3]不是由dplyr打印的。它是tibble的打印方式。
您可以检查/模仿它：
> a = iris
> # Set "Groups"
> setattr(a, "vars", "this_is_a_test_group")
> # lie that it is a grouped_df
> setattr(a, "class", c("grouped_df","tbl_df", "tbl", "data.frame"))
>
> a
# A tibble: 150 x 5
# Groups: this_is_a_test_group [?]
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
<dbl> <dbl> <dbl> <dbl> <fct>
1 5.1 3.5 1.4 0.2 setosa
2 4.9 3 1.4 0.2 setosa
3 4.7 3.2 1.3 0.2 setosa
4 4.6 3.1 1.5 0.2 setosa
5 5 3.6 1.4 0.2 setosa
6 5.4 3.9 1.7 0.4 setosa
7 4.6 3.4 1.4 0.3 setosa
8 5 3.4 1.5 0.2 setosa
9 4.4 2.9 1.4 0.2 setosa
10 4.9 3.1 1.5 0.1 setosa
# ... with 140 more rows
组Species [3]（[3]）的数量由indices属性控制。
这是每个组的行索引的嵌套列表。
> setattr(a, "indices", list(1:5, 5:10, 10:15))
> a
# A tibble: 150 x 5
# Groups: this_is_a_test_group [3]
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
<dbl> <dbl> <dbl> <dbl> <fct>
1 5.1 3.5 1.4 0.2 setosa
2 4.9 3 1.4 0.2 setosa
3 4.7 3.2 1.3 0.2 setosa
4 4.6 3.1 1.5 0.2 setosa
5 5 3.6 1.4 0.2 setosa
6 5.4 3.9 1.7 0.4 setosa
7 4.6 3.4 1.4 0.3 setosa
8 5 3.4 1.5 0.2 setosa
9 4.4 2.9 1.4 0.2 setosa
10 4.9 3.1 1.5 0.1 setosa
# ... with 140 more rows
这实际上并不能真正回答您的问题。但是在评论中展示它会更困难。
要获得这样的输出（在实际data.table之前使用groups[n]，则必须编写一个新的打印方法，该方法是对data.table打印的包装，但是要打印那些“ vars”属性在调用data.table打印之前。
并创建一个新类，例如“ custom.groupped.data.table”，将使用该打印方法。
或附加这些属性，然后将其转换为小对象。
编辑：
好吧，显然其中一些功能已在data.table打印方法中实现（但默认情况下未启用）。
具体是：print.keys = TRUE
> b = data.table(iris)
> setkey(b, "Species")
> print(b, print.keys = TRUE)
Key: <Species>
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1: 5.1 3.5 1.4 0.2 setosa
2: 4.9 3.0 1.4 0.2 setosa
3: 4.7 3.2 1.3 0.2 setosa
4: 4.6 3.1 1.5 0.2 setosa
5: 5.0 3.6 1.4 0.2 setosa
---
146: 6.7 3.0 5.2 2.3 virginica
147: 6.3 2.5 5.0 1.9 virginica
148: 6.5 3.0 5.2 2.0 virginica
149: 6.2 3.4 5.4 2.3 virginica
150: 5.9 3.0 5.1 1.8 virginica
> print(b, print.keys = TRUE, class = TRUE)
Key: <Species>
Sepal.Length Sepal.Width Petal.Length Petal.Width Species
<num> <num> <num> <num> <fctr>
1: 5.1 3.5 1.4 0.2 setosa
2: 4.9 3.0 1.4 0.2 setosa
3: 4.7 3.2 1.3 0.2 setosa
4: 4.6 3.1 1.5 0.2 setosa
5: 5.0 3.6 1.4 0.2 setosa
---
146: 6.7 3.0 5.2 2.3 virginica
147: 6.3 2.5 5.0 1.9 virginica
148: 6.5 3.0 5.2 2.0 virginica
149: 6.2 3.4 5.4 2.3 virginica
150: 5.9 3.0 5.1 1.8 virginica