条件选择两个数据帧R

Question

假设A和B是两个dataframes，最快的方法是什么：

查找A>0andB<0？即CCD_ 7大于零，条件是CCD_。

两个dataframes具有相等的尺寸，非常大(>70GB(，并且对A中的column1和B中的coulmn1等进行运算

A=c(-31.621, -40.149, -21.519, -33.773, -34.023, -38.345, -39.238, 
-30.221, -22.278, 27.055)
B=c(-29.321, -26.398, -25.663, -26.423, -23.721, -17.418, -15.979, 
-21.927, -16.398, -21.084)

Answer 1

这就是您想要的吗？

选项1-加入然后处理条件

library(tibble)
library(dplyr)
A <- c(-31.621, -40.149, -21.519, -33.773, -34.023, -38.345, -39.238, 
-30.221, -22.278, 27.055) %>% 
enframe(name = "ID")
B <- c(-29.321, -26.398, -25.663, -26.423, -23.721, -17.418, -15.979, 
-21.927, -16.398, -21.084) %>% 
enframe(name = "ID")
A %>% 
left_join(., B, by = "ID", suffix = c("_A", "_B")) %>% 
mutate(condition = case_when(value_A > 0 & value_B < 0 ~ T,
TRUE ~ F))
#> # A tibble: 10 x 4
#>       ID value_A value_B condition
#>    <int>   <dbl>   <dbl> <lgl>    
#>  1     1   -31.6   -29.3 FALSE    
#>  2     2   -40.1   -26.4 FALSE    
#>  3     3   -21.5   -25.7 FALSE    
#>  4     4   -33.8   -26.4 FALSE    
#>  5     5   -34.0   -23.7 FALSE    
#>  6     6   -38.3   -17.4 FALSE    
#>  7     7   -39.2   -16.0 FALSE    
#>  8     8   -30.2   -21.9 FALSE    
#>  9     9   -22.3   -16.4 FALSE    
#> 10    10    27.1   -21.1 TRUE

选项2-处理条件，然后加入

library(tibble)
library(dplyr)
A <- c(-31.621, -40.149, -21.519, -33.773, -34.023, -38.345, -39.238, 
-30.221, -22.278, 27.055) %>% 
enframe(name = "ID") %>% 
filter(value > 0)
B <- c(-29.321, -26.398, -25.663, -26.423, -23.721, -17.418, -15.979, 
-21.927, -16.398, -21.084) %>% 
enframe(name = "ID") %>% 
filter(value < 0)
A %>% 
left_join(., B, by = "ID", suffix = c("_A", "_B"))
#> # A tibble: 1 x 3
#>      ID value_A value_B
#>   <int>   <dbl>   <dbl>
#> 1    10    27.1   -21.1

选项3-处理条件并查找与条件匹配的索引

library(tibble)
library(dplyr)
A <- c(-31.621, -40.149, -21.519, -33.773, -34.023, -38.345, -39.238, 
-30.221, -22.278, 27.055) %>% 
enframe(name = "ID") %>% 
filter(value > 0)
B <- c(-29.321, -26.398, -25.663, -26.423, -23.721, -17.418, -15.979, 
-21.927, -16.398, -21.084) %>% 
enframe(name = "ID") %>% 
filter(value < 0)
intersect(A$ID, B$ID)
#> [1] 10

^{由reprex包(v0.3.0(于2020-02-25创建}