In data.table in R, how can you reference columns that have `<=` in their names without making R think its an operation?

Suppose we have the following data.table object in R, and we wish to subset the data by those where x<=0.5 == 1:

x=rnorm(1000)
df <- data.table(x=x, `x<=0.5`=1*(x<=0.5))
df[`x<=0.5` == 1]

This returns an error saying:

Error in .parse_on(substitute(on), isnull_inames): Found more than one operator in one 'on' statement: x<=0.5==x<=0.5. Please specify a single operator.

How can I actually reference it without this error? I assume it is because it thinks the less than or equal sign is not a name but an operation.

We may use get on a string to return the value

df[get("x<=0.5") == 1]

-output

               x x<=0.5
  1:  0.3903123      1
  2:  0.1426477      1
  3: -0.3169492      1
  4: -1.3024570      1
  5: -1.5403663      1
 ---                  
691: -0.4453730      1
692: -2.3406638      1
693: -0.3276438      1
694:  0.4063935      1
695: -0.5860276      1

Or use it in .SDcols

df[df[, .SD[[1]] == 1, .SDcols = "x<=0.5"]]
</details>