R：如何使用索引列聚合数据框？

Question

我有一个如下所示的数据框：

head(test_df, n =15)
# print the first 15rows of the dataframe

               value frequency                        index
1  -2.90267705917358         1                            1
2  -2.90254878997803         1                            1
3  -2.90252590179443         1                            1
4  -2.90219354629517         1                            1
5  -2.90201354026794         1                            1
6   -2.9016375541687         1                            1
7  -2.90107154846191         1                            1
8  -2.90089440345764         1                            1
9  -2.89996957778931         1                            1
10 -2.89970088005066         1                            1
11 -2.89928865432739         1                            2
12 -2.89920520782471         1                            2
13 -2.89907360076904         1                            2
14 -2.89888191223145         1                            2
15  -2.8988630771637         1                            2

数据框有 3 列和 61819 行。要聚合数据框，我想获取具有相同 'index' 的所有行的 'value' 和 'frequency' 列的平均值。

我已经找到了一些有用的链接，请参阅：

https://www.r-bloggers.com/2018/07/how-to-aggregate-data-in-r/

但是，我还没有解决问题。

test_df_ag <- stats::aggregate(test_df[1:2], by = test_df[3], FUN = 'mean')
# aggregate the dataframe based on the 'index' column (build the mean)

   index value frequency
1      1    NA         1
2      2    NA         1
3      3    NA         1
4      4    NA         1
5      5    NA         1
6      6    NA         1
7      7    NA         1
8      8    NA         1
9      9    NA         1
10    10    NA         1
11    11    NA         1
12    12    NA         1
13    13    NA         1
14    14    NA         1
15    15    NA         1

因为我只得到列 'value' 的 NA 值，我想知道它是否 可能只是数据类型问题？！ 但是当我尝试转换时我失败的数据类型...

base::typeof(test_df$value)
# query the data type of the 'value' column
[1] "integer"

Answer 1

试试 tidyverse。 test_summary <- test_df %>% group_by(index) %>% summarise(n=n(), mean_value=mean(value, na.rm=T),mean_frequency=mean(frequency, na.rm=T)).

哦，当然，您应该确保您已经检查了数据的质量并了解数据集中任何 NA 的假设和原因。

Answer 2

1. 这是一个基本的 R 解决方案。

aggregate(cbind(value, frequency) ~ index, data = test_df, FUN = mean)
#  index     value frequency
#1     1 -2.901523         1
#2     2 -2.899062         1

2. 和一个简单的 dplyr 解决方案。

library(dplyr)

test_df %>%
  group_by(index) %>%
  summarize(across(1:2, mean))
## A tibble: 2 x 3
#  index value frequency
#* <int> <dbl>     <dbl>
#1     1 -2.90         1
#2     2 -2.90         1

数据

test_df <-
structure(list(value = c(-2.90267705917358, -2.90254878997803, 
-2.90252590179443, -2.90219354629517, -2.90201354026794, -2.9016375541687, 
-2.90107154846191, -2.90089440345764, -2.89996957778931, -2.89970088005066, 
-2.89928865432739, -2.89920520782471, -2.89907360076904, -2.89888191223145, 
-2.8988630771637), frequency = c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), index = c(1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L)), class = "data.frame", row.names = c("1", 
"2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", 
"14", "15"))

Answer 3

使用data.table

library(data.table)
setDT(test_df)[, lapply(.SD, mean), by = index, .SDcols = 1:2]

R：如何使用索引列聚合数据框？

R: How to aggregate a dataframe using an index column?

statistics

aggregate

r

mean

dataframe

数据