Conditions

Conditions and If statements

The use of conditions in R is very important. A logical variable is a variable that can only take two values : TRUE and FALSE. A if statement indicates that the block code between { } will only be executed if the condition between parenthesis is fulfilled (or TRUE, to say it in short).

Take a break & Read

For an introduction in the differents logical conditions and if statements, please look at W3schools If...Else part. Thanks to this reference, you can also test the different notions directly on this website.

Comparison operators

You previously saw this kind of operators but they are really at the base of conditions and If statements. You can see a more detailed and well described documentation by reading again the section 7.2 of the "R pirate Guide" by Nathaniel D. Phillips.

If Statements

Now that you are familiar with comparison operators, we can look forward into the conditional statements. You need to know 2 statements : if and else and 1 R function : ifelse().

The structure of if statement will always be the same :

if (cond) {
    code to execute if `cond` returns `TRUE`
} else {
    code to execute if `cond` returns `FALSE`
}

Warning

Note that you can use if without else but never in the opposite way.

if(cond){
    code to execute if `cond` returns `TRUE`
} 

If you have more than two conditions, you will need to combine if statements. But if you are familiar with python, you may know about the existence of elif, unfortunately it doesn't exist in R. You'll write instead:

my_val <- -5
if (my_val > 0) {
    print("my value is a positive value.")
} else if (my_val < 0) {
    print("my value is a negative value.")
} else {
    print("my value is a null value.")
}
## [1] "my value is a negative value."

Note

R will sequentially test the provided conditions and stop when the condition is met. In this case, the first test my_val > 0 returns FALSE, R will then test whether our value is negative (my_val < 0), which returns TRUE and test will not continue.

These statements are only working when your condition returns a single value (TRUE or FALSE).

Sometimes you need to repeat the same test multiple times, there is a way to do it in a "one line" R command thanks to the ifelse() function:

my_vec <- 1:10
my_test <- ifelse(test = my_vec < 5, #the value in my_vec is inferior to 5
                  yes = my_vec,      #if `test = TRUE` it returns the value
                  no = 5)            #if `test = TRUE` it returns `5`

my_test
## [1] 1 2 3 4 5 5 5 5 5 5

The R function ifelse makes it easier to write code when manipulating values inside a variable. Thanks to this, you can really imagine manipulate more complex variables such as dataframes or matrices.

my_df <- data.frame(gene = c("A1BG", "EPC1", "MTMR7", "SLC20A2", "ZZZ3"),
                    mean_exp = c(1, 4, 0, 10, 3))
my_df
##      gene mean_exp
## 1    A1BG        1
## 2    EPC1        4
## 3   MTMR7        0
## 4 SLC20A2       10
## 5    ZZZ3        3

#create a new column based on another
my_df$is_expressed <- ifelse(my_df$mean_exp == 0,
                             "not_detected",
                             ifelse(my_df$mean_exp < 5,
                             "low detection",
                             "high_detection"))

my_df
##      gene mean_exp   is_expressed
## 1    A1BG        1  low detection
## 2    EPC1        4  low detection
## 3   MTMR7        0   not_detected
## 4 SLC20A2       10 high_detection
## 5    ZZZ3        3  low detection