The Difference Between rnorm() and runif() in R

Spread the love

In R, we have various functions to create random variables that follow specific probability distributions. Two such functions are rnorm() and runif(). rnorm() is used to generate random numbers that follow a normal distribution, while runif() is used to generate random numbers that follow a uniform distribution. While they are fundamentally similar in that they both generate random numbers, they exhibit major differences in their underlying distributions, parameters, and usage.

Basic Definitions

Before we delve deeper into their differences, let’s first establish what rnorm() and runif() are in the context of R.


rnorm() is a function in R that generates a set of random numbers according to a normal distribution (also known as Gaussian distribution). The syntax for rnorm() is as follows:

rnorm(n, mean = 0, sd = 1)

Here n is the number of observations (random numbers) to generate, mean is the mean of the normal distribution, and sd is the standard deviation.


runif() is a function in R that generates a set of random numbers according to a uniform distribution. The syntax for runif() is as follows:

runif(n, min = 0, max = 1)

Here n is the number of observations (random numbers) to generate, min is the minimum limit of the distribution, and max is the maximum limit.

Differences Between rnorm() and runif()

1. Underlying Distributions

The fundamental difference between rnorm() and runif() is the underlying probability distribution they use to generate the random numbers.

rnorm() generates numbers from a normal distribution, which is bell-shaped and symmetric, meaning it has a peak at the mean, and the data around the mean is dense. In other words, most of the observations in a normal distribution are close to the mean, and the frequency of the observations decreases as we move away from the mean on either side.

On the other hand, runif() generates numbers from a uniform distribution. In a uniform distribution, all numbers within the specified range have an equal probability of being drawn. The distribution is a flat line because no value is more likely to occur than others within the specified minimum and maximum bounds.

2. Parameters

The parameters required by rnorm() and runif() are different due to their distribution characteristics.

rnorm() requires the mean (mean) and standard deviation (sd) of the distribution. The mean is the central value of the distribution, around which the data is symmetrically distributed. The standard deviation is a measure of the dispersion or how spread out the values are around the mean.

Conversely, runif() requires the minimum (min) and maximum (max) values of the distribution. The generated random numbers will fall within this specified range, and each will have an equal probability of being chosen.

3. Usage

The rnorm() and runif() functions are used in different scenarios that align with their distribution characteristics.

rnorm() is often used when the data is expected to have a lot of values close to a central value with symmetry on both sides, like in the case of heights of people, test scores, etc. It’s also a go-to choice for many statistical tests and procedures as they often assume that the data is normally distributed.

runif(), on the other hand, is useful when there is no apparent skew or preference for particular values in the data within the specified range, such as the roll of a fair die, or drawing a card from a well-shuffled deck.

Practical Examples

rnorm() Example:

# Generate 5 random numbers from a normal distribution with mean 10 and standard deviation 2
random_numbers <- rnorm(5, mean = 10, sd = 2)

runif() Example:

# Generate 5 random numbers from a uniform distribution between 0 and 100
random_numbers <- runif(5, min = 0, max = 100)

These examples help illustrate how rnorm() and runif() work and how their resulting outputs are fundamentally different.


In summary, rnorm() and runif() are integral parts of the random number generation capabilities of R. While they both generate random numbers, they do so based on different probability distributions – normal and uniform, respectively. Their choice depends on the problem at hand, the data’s distribution characteristics, and the specific requirements of statistical analysis.

Posted in RTagged

Leave a Reply