How to use the switch statement in R functions?

RSwitch Statement

R Problem Overview


I would like to use for my function in R the statement switch() to trigger different computation according to the value of the function's argument.

For instance, in Matlab you can do that by writing

switch(AA)        
case '1'   
...   
case '2'   
...   
case '3'  
...  
end

I found this post - switch() statement usage - that explain how to use switch, but not really helpful to me as I want to perform more sophisticated computation (matrix operations) and not a simple mean.

R Solutions


Solution 1 - R

Well, switch probably wasn't really meant to work like this, but you can:

AA = 'foo'
switch(AA, 
foo={
  # case 'foo' here...
  print('foo')
},
bar={
  # case 'bar' here...
  print('bar')    
},
{
   print('default')
}
)

...each case is an expression - usually just a simple thing, but here I use a curly-block so that you can stuff whatever code you want in there...

Solution 2 - R

those various ways of switch ...

# by index
switch(1, "one", "two")
## [1] "one"


# by index with complex expressions
switch(2, {"one"}, {"two"})
## [1] "two"

 
# by index with complex named expression
switch(1, foo={"one"}, bar={"two"})
## [1] "one"

 
# by name with complex named expression
switch("bar", foo={"one"}, bar={"two"})
## [1] "two"

Solution 3 - R

I hope this example helps. You ca use the curly braces to make sure you've got everything enclosed in the switcher changer guy (sorry don't know the technical term but the term that precedes the = sign that changes what happens). I think of switch as a more controlled bunch of if () {} else {} statements.

Each time the switch function is the same but the command we supply changes.

do.this <- "T1"

switch(do.this,
    T1={X <- t(mtcars)
        colSums(mtcars)%*%X
    },
    T2={X <- colMeans(mtcars)
        outer(X, X)
    },
    stop("Enter something that switches me!")
)
#########################################################
do.this <- "T2"

switch(do.this,
    T1={X <- t(mtcars)
        colSums(mtcars)%*%X
    },
    T2={X <- colMeans(mtcars)
        outer(X, X)
    },
    stop("Enter something that switches me!")
)
########################################################
do.this <- "T3"

switch(do.this,
    T1={X <- t(mtcars)
        colSums(mtcars)%*%X
    },
    T2={X <- colMeans(mtcars)
        outer(X, X)
    },
    stop("Enter something that switches me!")
)

Here it is inside a function:

FUN <- function(df, do.this){
    switch(do.this,
        T1={X <- t(df)
            P <- colSums(df)%*%X
        },
        T2={X <- colMeans(df)
            P <- outer(X, X)
        },
        stop("Enter something that switches me!")
    )
    return(P)
}

FUN(mtcars, "T1")
FUN(mtcars, "T2")
FUN(mtcars, "T3")

Solution 4 - R

This is a more general answer to the missing "Select cond1, stmt1, ... else stmtelse" connstruction in R. It's a bit gassy, but it works an resembles the switch statement present in C

while (TRUE) {
  if (is.na(val)) {
    val <- "NULL"
    break
  }
  if (inherits(val, "POSIXct") || inherits(val, "POSIXt")) {
    val <- paste0("#",  format(val, "%Y-%m-%d %H:%M:%S"), "#")
    break
  }
  if (inherits(val, "Date")) {
    val <- paste0("#",  format(val, "%Y-%m-%d"), "#")
    break
  }
  if (is.numeric(val)) break
  val <- paste0("'", gsub("'", "''", val), "'")
  break
}

Attributions

All content for this solution is sourced from the original question on Stackoverflow.

The content on this page is licensed under the Attribution-ShareAlike 4.0 International (CC BY-SA 4.0) license.

Content TypeOriginal AuthorOriginal Content on Stackoverflow
QuestionSimonView Question on Stackoverflow
Solution 1 - RTommyView Answer on Stackoverflow
Solution 2 - RpetermeissnerView Answer on Stackoverflow
Solution 3 - RTyler RinkerView Answer on Stackoverflow
Solution 4 - RRolf LView Answer on Stackoverflow