![](/img/trans.png)
[英]Compare two pairs of columns from one dataframe to detect mismatches and show the value from another column in the same row
[英]Compare one column in a dataframe to two columns of another dataframe
我有兩個數據幀,我需要知道第一個 dataframe 的值是否介於第二個 dataframe 中的兩個值(最小值和最大值)之間。
我之前對其他兩個數據框做了類似的事情,我使用了嵌套loop
和between {dplyr}
。 但是,另一個數據集只有三個變量,我可以使用 8 個if
語句。 這就是我卡住的地方,dataframe1 有 62 個變量和 477 個觀察值,而 dataframe2 有 124 個變量和 50 個觀察值(最小值和最大值)。 下面我有兩個數據框的示例以及我正在尋找的結果。
所以我正在尋找一個解決方案,我不必寫大約一千個if else
語句。 我希望有人可以提供幫助,或者如果這可能的話。
數據看起來如何的示例,我仍然可以更改數據框,但這就是我所處的位置。
Df1
id type data1 data2 data3
1 1 ab 0 0 0
2 2 cd 0 0 0
3 3 dd 0 10 5
4 4 ed 0 0 0
5 5 kd 0 0 15
6 6 xd 0 5 0
7 7 ab 0 0 0
8 8 cd 0 0 0
9 9 dd 0 10 10
10 10 ed 0 0 0
11 11 kd 0 0 12
12 12 xd 0 12 0
13 13 ab 0 0 0
14 14 cd 0 0 0
15 15 dd 0 5 15
16 16 ed 0 0 0
17 17 kd 0 0 15
18 18 xd 0 7 0
19 19 ab 0 0 0
20 20 cd 0 0 0
21 21 dd 0 18 10
22 22 ed 0 0 0
23 23 kd 0 0 5
我通常將“類型”相互匹配,然后匹配數據是否在下邊界和上邊界之間。
Df2
type data1 data1max data2 data2max data3 data3max
1 ab NA NA NA NA NA NA
2 dd NA NA 5 20 10 100
3 xd NA NA 1 30 NA NA
4 ed NA NA NA NA NA NA
5 cd NA NA NA NA NA NA
6 kd NA NA NA NA 5 20
並在觀察到的數據與合格數據匹配時產生計數。
Df3
id type qualifyingfields
1 1 ab 0
2 2 cd 0
3 3 dd 1
4 4 ed 0
5 5 kd 1
6 6 xd 1
7 7 ab 0
8 8 cd 0
9 9 dd 2
10 10 ed 0
11 11 kd 1
12 12 xd 1
13 13 ab 0
14 14 cd 0
15 15 dd 2
16 16 ed 0
17 17 kd 1
18 18 xd 1
19 19 ab 0
20 20 cd 0
21 21 dd 1
22 22 ed 0
23 23 kd 1
library(dplyr)
library(tidyr)
df1 %>%
right_join(., df2, by = "type", suffix = c("val", "min")) %>%
group_by(type, id) %>%
pivot_longer(-c(id, type), names_to = "data", values_to = "value") %>%
separate(col = data, into = c("data", "var"), sep = "(?<=\\d)") %>%
pivot_wider(names_from = var, values_from = value) %>%
group_by(id, type, data) %>%
mutate(qualifyingfields = sum(between(val, min, max), na.rm = T)) %>%
group_by(id, type) %>%
summarise(qualifyingfields = sum(qualifyingfields))
#> # A tibble: 23 x 3
#> # Groups: type, id [23]
#> id type qualifyingfields
#> <int> <chr> <int>
#> 1 1 ab 0
#> 2 2 cd 0
#> 3 3 dd 1
#> 4 4 ed 0
#> 5 5 kd 1
#> 6 6 xd 1
#> 7 7 ab 0
#> 8 8 cd 0
#> 9 9 dd 2
#> 10 10 ed 0
#> # ... with 13 more rows
數據:
df1 <- read.table(text=" id type data1 data2 data3
1 1 ab 0 0 0
2 2 cd 0 0 0
3 3 dd 0 10 5
4 4 ed 0 0 0
5 5 kd 0 0 15
6 6 xd 0 5 0
7 7 ab 0 0 0
8 8 cd 0 0 0
9 9 dd 0 10 10
10 10 ed 0 0 0
11 11 kd 0 0 12
12 12 xd 0 12 0
13 13 ab 0 0 0
14 14 cd 0 0 0
15 15 dd 0 5 15
16 16 ed 0 0 0
17 17 kd 0 0 15
18 18 xd 0 7 0
19 19 ab 0 0 0
20 20 cd 0 0 0
21 21 dd 0 18 10
22 22 ed 0 0 0
23 23 kd 0 0 5",
header=T, stringsAsFactors=F)
df2 <- read.table(text=" type data1 data1max data2 data2max data3 data3max
1 ab NA NA NA NA NA NA
2 dd NA NA 5 20 10 100
3 xd NA NA 1 30 NA NA
4 ed NA NA NA NA NA NA
5 cd NA NA NA NA NA NA
6 kd NA NA NA NA 5 20",
header=T, stringsAsFactors=F, na.strings = "NA")
這是一個更通用的解決方案,適用於數據,無論有多少data[n]
列
library('dplyr')
library('tidyr')
# Make dataframes tidy
Df1_tidy <- Df1 %>%
gather(key='data_name', value='value', -(id:type))
Df2_tidy <- Df2 %>%
gather(key='data_name', value='value', -type) %>%
mutate(limit=ifelse(grepl('max', data_name), 'Max', 'Min'),
data_name=gsub('max', '', data_name)) %>%
spread(limit, value)
# Count qualifying fields
Df3 <- full_join(Df1_tidy, Df2_tidy) %>%
group_by(id, type) %>%
summarise(qualifyingfields = sum(value >= Min & value <= Max, na.rm=T)) %>%
ungroup()
Df3
# # A tibble: 23 x 3
# id type qualifyingfields
# <int> <chr> <int>
# 1 1 ab 0
# 2 2 cd 0
# 3 3 dd 1
# 4 4 ed 0
# 5 5 kd 1
# 6 6 xd 1
# 7 7 ab 0
# 8 8 cd 0
# 9 9 dd 2
# 10 10 ed 0
# # ... with 13 more rows
獲取數據(從@M--響應復制):
df1 <- read.table(text=" id type data1 data2 data3
1 1 ab 0 0 0
2 2 cd 0 0 0
3 3 dd 0 10 5
4 4 ed 0 0 0
5 5 kd 0 0 15
6 6 xd 0 5 0
7 7 ab 0 0 0
8 8 cd 0 0 0
9 9 dd 0 10 10
10 10 ed 0 0 0
11 11 kd 0 0 12
12 12 xd 0 12 0
13 13 ab 0 0 0
14 14 cd 0 0 0
15 15 dd 0 5 15
16 16 ed 0 0 0
17 17 kd 0 0 15
18 18 xd 0 7 0
19 19 ab 0 0 0
20 20 cd 0 0 0
21 21 dd 0 18 10
22 22 ed 0 0 0
23 23 kd 0 0 5",
header=T, stringsAsFactors=F)
df2 <- read.table(text=" type data1 data1max data2 data2max data3 data3max
1 ab NA NA NA NA NA NA
2 dd NA NA 5 20 10 100
3 xd NA NA 1 30 NA NA
4 ed NA NA NA NA NA NA
5 cd NA NA NA NA NA NA
6 kd NA NA NA NA 5 20",
header=T, stringsAsFactors=F, na.strings = "NA")
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.