How to remove duplicate columns after dplyr join?

I like to do things in as few steps as possible. I think this would reduce the numbers of steps:

combine<-df1%>%
  left_join(df2, by="id", suffix=c("",".y")%>%
  select(-ends_with(".y"))

The minus sign in the select command means that you want to select everything EXCEPT those variables. If you want to delete ALL of the duplicated columns (no column a at all), you could do this:

combine<-df1%>%
  left_join(df2, by="id", suffix=c(".x",".y")%>%
  select(-ends_with(".x"),-ends_with(".y"))

First we perform the join by id

combined <- df1 %>% left_join(df2, by="id")

Then we rename those with .x and drop those with .y

combined <- combined %>% 
  rename_at(
    vars(ends_with(".x")),
    ~str_replace(., "\\..$","")
  ) %>% 
  select_at(
    vars(-ends_with(".y"))
  )

The more generic approach would be to drop the columns before left join otherwise your combined dataset can be very large initially:

df1<- data.frame(id= seq(1:0), a=rnorm(1:10,0.2),b=rpois(10,0.2))
df2<- data.frame(id= seq(1:0), a=rnorm(1:10,0.2),c=rnorm(10,0.2))

varList<- names(df2)[!(names(df2) %in% names(df1))] # get non common names
varList<- c(varList,"id") # appending key parameter

combined <- df1 %>% left_join((df2 %>% select(varList)), by="id")

combined dataset will not have any .x or .y

By "there are too many of columns like a" do you mean you want to find all the columns which are common to both sources? In that case, why not literally use the intersection (the default behaviour)?

## two data.frames, only id = 3, a = 4 matches
(df1 <- data.frame(id = 1:3, a = 2:4, b = 3:5))
#>   id a b
#> 1  1 2 3
#> 2  2 3 4
#> 3  3 4 5
(df2 <- data.frame(id = 3:2, a = 4:5, c = 1:2))
#>   id a c
#> 1  3 4 1
#> 2  2 5 2

## this produces a.x and a.y                  
dplyr::left_join(df1, df2, by = "id")
#>   id a.x b a.y  c
#> 1  1   2 3  NA NA
#> 2  2   3 4   5  2
#> 3  3   4 5   4  1

## which columns are common?
intersect(names(df1), names(df2))
#> [1] "id" "a"

## this produces id, a, b, c
dplyr::left_join(df1, df2, by = intersect(names(df1), names(df2)))
#>   id a b  c
#> 1  1 2 3 NA
#> 2  2 3 4 NA
#> 3  3 4 5  1

## this is, however, the default behaviour for left_join
## i.e. use all columns which are present in both
dplyr::left_join(df1, df2)
#> Joining, by = c("id", "a")
#>   id a b  c
#> 1  1 2 3 NA
#> 2  2 3 4 NA
#> 3  3 4 5  1

^{Created on 2020-05-06 by the reprex package (v0.3.0)}

I think this is the simplest way to achieve what you're trying to do

df <- left_join(df1, df2, by = "id", suffix = c("", ".annoying_duplicate_column")) %>%
  select(-ends_with(".annoying_duplicate_column"))

(Combining the answer by @Ernest Han with the very helpful comment by @David T above)