Exponential and Gamma Densities

Definition

The exponential distribution with rate parameter $\lambda > 0$ has probability density function

$$f(x) = \lambda e^{-\lambda x}, \quad x \ge 0$$

and $f(x) = 0$ for $x < 0$.

The gamma distribution with shape parameter $\alpha > 0$ and rate parameter $\lambda > 0$ has probability density function

$$f(x) = \frac{\lambda^\alpha}{\Gamma(\alpha)} x^{\alpha-1} e^{-\lambda x}, \quad x \ge 0$$

and $f(x)=0$ for $x<0$, where $\Gamma(\alpha)$ is the gamma function.

These densities describe random variables that take only nonnegative values and are commonly used to represent waiting times and durations.

---

Main Content

1. Exponential Density

Form and meaning of the density

The exponential density is

$$f(x)=\lambda e^{-\lambda x}, \quad x \ge 0$$

where $\lambda$ is called the rate parameter. A larger $\lambda$ means events happen more quickly, so the expected waiting time is shorter. The curve starts at its maximum value $\lambda$ when $x=0$, then decreases continuously toward 0 as $x$ increases.

Example: If $\lambda=2$, then

$$f(x)=2e^{-2x}, \quad x\ge 0$$

This models a situation where short waiting times are more likely than long ones.

Properties and interpretation

The exponential distribution has several key properties:

• It is supported only on $x \ge 0$.
• It is right-skewed.

• Its mean is

  $$E[X] = \frac{1}{\lambda}$$

• Its variance is

  $$\operatorname{Var}(X)=\frac{1}{\lambda^2}$$

• Its cumulative distribution function is

  $$F(x)=1-e^{-\lambda x}, \quad x\ge 0$$

A famous feature is the memoryless property:

$$P(X>s+t \mid X>s)=P(X>t)$$

This means that the future waiting time does not depend on how long you have already waited. This property is unique to the exponential distribution among continuous distributions.

---

2. Gamma Density

Form and meaning of the density

The gamma density is

$$f(x)=\frac{\lambda^\alpha}{\Gamma(\alpha)}x^{\alpha-1}e^{-\lambda x}, \quad x\ge 0$$

where:

• $\alpha$ is the shape parameter
• $\lambda$ is the rate parameter
• $\Gamma(\alpha)$ is the gamma function, which generalizes factorials

If $\alpha$ is an integer $n$, then

$$\Gamma(n)=(n-1)!$$

Example: For $\alpha=3$,

$$f(x)=\frac{\lambda^3}{2}x^2e^{-\lambda x}, \quad x\ge 0$$

This density often models the time until the third event occurs in a Poisson process.

Properties and interpretation

The gamma distribution is very flexible because its shape changes with $\alpha$:

• If $\alpha<1$, the density is heavily concentrated near 0 and decreases rapidly.
• If $\alpha=1$, it becomes the exponential distribution.
• If $\alpha>1$, it rises from 0, reaches a peak, and then falls.

Important formulas:

• Mean:

  $$E[X]=\frac{\alpha}{\lambda}$$

• Variance:

  $$\operatorname{Var}(X)=\frac{\alpha}{\lambda^2}$$

The gamma distribution is widely used because it can represent waiting times for multiple independent events, accumulated damage, rainfall totals, and many other positive-valued quantities.

---

3. Relationship Between Exponential and Gamma Densities

Exponential as a special case of gamma

The exponential distribution is a special case of the gamma distribution when the shape parameter is 1:

$$\text{Gamma}(\alpha=1,\lambda) = \text{Exponential}(\lambda)$$

Substituting $\alpha=1$ into the gamma density gives

$$f(x)=\frac{\lambda^1}{\Gamma(1)}x^{0}e^{-\lambda x} =\lambda e^{-\lambda x}$$

since $\Gamma(1)=1$.

This shows that the exponential distribution is not separate from the gamma family; it is the simplest gamma distribution.

Interpretation through waiting times

In a Poisson process, events occur randomly over time at a constant average rate. Then:

• The time until the first event is exponential.
• The time until the $n$-th event is gamma with shape $n$.

For example, if calls arrive at a call center according to a Poisson process, the waiting time until the first call is exponential, while the waiting time until the fifth call is gamma with $\alpha=5$.

This relationship is important because it connects density functions with real stochastic processes and explains why these distributions appear together in practice.

---

Working / Process

1. Identify the positive random quantity

Determine whether the variable represents a waiting time, lifetime, delay, or accumulated amount. If the values are always nonnegative and the process is random over time, exponential or gamma models may be appropriate.

2. Choose the correct density model

• Use the exponential density if events occur independently at a constant rate and the waiting time to a single event is of interest.
• Use the gamma density if the waiting time to the $n$-th event is needed or if the data have a more flexible shape than exponential.

3. Apply the formula and compute quantities

Use the density function to find probabilities, means, variances, or percentiles. For example:

• Probability of waiting less than $t$ units:

  $$P(X\le t)=F(t)$$

• For exponential:

  $$F(t)=1-e^{-\lambda t}$$

• For gamma, use the cumulative distribution function or tables/software because the integral may not have a simple elementary form for non-integer $\alpha$.

Example workflow:

• Suppose $\lambda=0.5$ for an exponential model.

• To find the probability that waiting time is at most 3 units:

  $$P(X\le 3)=1-e^{-0.5(3)}=1-e^{-1.5}$$

• This gives the probability directly from the density’s cumulative form.

---

Advantages / Applications

Modeling waiting times and lifetimes

Exponential and gamma densities are ideal for representing the time until events occur. This includes equipment failure times, arrival times in queues, and time to recovery or relapse in medical studies.

Flexibility in real-world data modeling

The exponential distribution is simple and mathematically elegant, while the gamma distribution provides more flexibility by adjusting its shape. This makes gamma useful when data are skewed but not well described by the exponential model.

Use in probability, statistics, and engineering

These distributions appear in:

• reliability engineering
• queueing systems
• survival analysis
• hydrology and meteorology
• Bayesian statistics
• signal processing

They are especially important because many theoretical results, simulations, and estimation methods are built around them.

---

Summary

• Exponential density models the waiting time until one event and has a constant rate $\lambda$.
• Gamma density generalizes the exponential distribution and models waiting time until multiple events.
• Exponential is the special case of gamma when the shape parameter equals 1.
• Important terms to remember: rate parameter, shape parameter, gamma function, memoryless property, waiting time, Poisson process.