Can I identify the same values within a range between 2 columns? - r

Question

I am trying to compare values between two different columns but I need it to accept values within a range of ±3.
I created this 2 tibbles:
example_tp1 <- tibble(Object_centre = c(84, 149, 489, 534, 680.5))
example_tp2 <- tibble(Object_centre = c(84.5, 149.5, 489, 528.5, 542, 680.5))
And I want the program to link the ones that are the same within a ±3 range.
So for example, I want it to identify that 84 and 84.5 are the same, also 149 and 149.5; 489 and 489; 680.5 and 680.5. But I want it to also tell me that 534, 528.5 and 542 do not have a match.
Is there any way to do this?

Answer 1

This could be achieved via the fuzzyjoin package like so:
library(dplyr)
library(fuzzyjoin)
example_tp1 <- tibble(Object_centre = c(84, 149, 489, 534, 680.5))
example_tp2 <- tibble(Object_centre = c(84.5, 149.5, 489, 528.5, 542, 680.5))
match_fun1 <- function(x, y) {
# (x >= y - 3) & (x <= y + 3)
# or following the suggestion by #DarrenTsai
abs(x - y) <= 3
}
fuzzy_full_join(example_tp1, example_tp2,
by = c("Object_centre"),
match_fun = match_fun1)
#> # A tibble: 7 x 2
#> Object_centre.x Object_centre.y
#> <dbl> <dbl>
#> 1 84 84.5
#> 2 149 150.
#> 3 489 489
#> 4 680. 680.
#> 5 534 NA
#> 6 NA 528.
#> 7 NA 542
Created on 2020-08-22 by the reprex package (v0.3.0)

Answer 2

You could look at all combinations of values and see which ones match.
# Data Frame of all combinations
example <- expand.grid(c(84, 149, 489, 534, 680.5), c(84.5, 149.5, 489, 528.5, 542, 680.5))
# Assigns a Match if the values are within a range of 3
example %>%
mutate(match = ifelse(abs(Var1-Var2) <= 3, "Match", "No Match"))
Var1 Var2 match
1 84.0 84.5 Match
2 149.0 84.5 No Match
3 489.0 84.5 No Match
4 534.0 84.5 No Match
5 680.5 84.5 No Match
6 84.0 149.5 No Match
7 149.0 149.5 Match
8 489.0 149.5 No Match
9 ..... ..... ........
10 ..... ..... ........
and so on
You could then filter out only the matches or see which values have no match.

Answer 3

Similar to #Jumble's answer using tidyverse functions :
tidyr::crossing(example_tp1, example_tp2, .name_repair = ~c('col1', 'col2')) %>%
dplyr::filter(abs(col1 - col2) <= 3)
# col1 col2
# <dbl> <dbl>
#1 84 84.5
#2 149 150.
#3 489 489
#4 680. 680.
crossing generates all combinations of example_tp1 and example_tp2 and we keep only those rows where the difference is less than equal to 3.