####
**R Data Science:**
R Programming A-Z: R For Data Science With Real Exercises!

This tutorial explains the usage of WHICH function in R and how it works with examples.

In R, the

which(letters=="z")

It returns

1. sapply(ls, is.numeric) returns TRUE TRUE TRUE FALSE. It's TRUE where variable is number else FALSE.

2. which(sapply(ls, is.numeric)) returns 1 2 3. Adding WHICH function returns the position in logical vectors.

3. colnames(ls)[check] returns x1 x2 and x3.

In R, the

**which()**function gives you the**position**of elements of a logical vector that are**TRUE.**It can be a row number or column number or position in a vector.**Examples**

**1. What is the position of alphabet 'z' in a-z letters.**which(letters=="z")

It returns

**26**as 'z' is positioned at 26th place**.****Create a sample data frame**

*The following program creates a data frame consisting four variables. This data frame would be used in the subsequent examples.*

ls = data.frame( x1 = ceiling(runif(10)*10),

x2 = ceiling(runif(10)*10),

x3 = runif(10),

x4= rep(letters[1:5],2))

Sample Data Frame |

**2. Column number of variable "x4" in ls data set**

i=which(names(ls)== "x4")It returns 4 as x4 is placed at 4th column. It works like this -

names(ls) == "x4" returns FALSE FALSE FALSE TRUE. Then WHICH function tells R to calculate the position of TRUE.

**3. Row number in which maximum value of variable "x1" exists**

which(ls$x1 == max(ls$x1))In this case, it gives you the row / observation number wherein the max value of x1 is stored.

**4. Row number in which multiple conditions hold true**which(ls$x1 == 7 & ls$x2 == 4)In this case, we are checking multiple conditions and figuring out the row number wherein conditions met.

**5. Number of cases in which variable x1 is equal to variable x2**

length(which(ls$x1 == ls$x2))

*Let's break out in multiple steps -*- which(ls$x1 == ls$x2) returns the position of all rows where these two variables are equal.
- length() function calculates the length of the returned values of step 1.

**6. Which value is common in both the variables**

ls[which(ls$x1 == ls$x2),"x1"]

**7. Extract names of all the numeric variables**check = which(sapply(ls, is.numeric))

colnames(ls)[check]

*Let's run it step by step -*1. sapply(ls, is.numeric) returns TRUE TRUE TRUE FALSE. It's TRUE where variable is number else FALSE.

2. which(sapply(ls, is.numeric)) returns 1 2 3. Adding WHICH function returns the position in logical vectors.

3. colnames(ls)[check] returns x1 x2 and x3.

Hello,

ReplyDeleteI think that question 5 could be answered in a more concise way.

sum(ls$x1 == ls$x2)

What do you think of my answer?

Best regards

Cédric Guilmin

In R, there are multiple ways to accomplish the same task. I was just explaining the usage of 'WHICH' function. Yes, the question 5 should be answered with sum(ls$x1 == ls$x2) as it is more efficient.

DeleteCheck out the following analysis -

ls = data.frame( x1 = ceiling(runif(10000000)*10),

x2 = ceiling(runif(10000000)*10))

start.time <- Sys.time()

length(which(ls$x1 == ls$x2))

end.time <- Sys.time()

time.taken <- end.time - start.time

time.taken

start.time <- Sys.time()

sum(ls$x1 == ls$x2)

end.time <- Sys.time()

time.taken <- end.time - start.time

time.taken

Thanks!

Delete