What is the exponential distribution used for?

The exponential distribution models time between events in a Poisson process: customer arrival times, component lifetimes, service durations, radioactive decay intervals, and earthquake occurrences. It's used whenever you're measuring waiting time until the next random event.

How do you calculate exponential probability?

The exponential PDF is f(x) = λe^(-λx) for x ≥ 0. The CDF is F(x) = 1 - e^(-λx). For P(X > x), use e^(-λx) directly. Range probabilities P(a ≤ X ≤ b) equal F(b) - F(a). The calculator handles these exponential computations automatically.

What's the relationship between exponential and Poisson?

If events occur according to a Poisson process with rate λ, the time between successive events follows an exponential distribution with the same rate λ. Poisson counts events in a fixed time; exponential measures time until an event. They're two perspectives on the same process.

What are the mean and variance of exponential distribution?

Both the mean and standard deviation equal 1/λ, where λ is the rate parameter. The variance equals 1/λ². For example, if events occur at rate λ = 2 per hour, the average waiting time is 0.5 hours with standard deviation also 0.5 hours.

Exponential Distribution Explorer

Modify Parameters and See Results

Time between events in a Poisson process

Parameters

Rate Parameter (λ): 1.0

Rate Parameter (λ)

1.00

Average number of events per unit time

Scale Parameter (1/λ)

1.00

Average time between events (mean)

Expected Value (E[X])

1.00

Mean waiting time until the next event

Statistics

Expected Value

1.0000

Variance

1.0000

Std Deviation

1.0000

Mode

0.0000

Probability Calculator

Enter lower bound a:

Enter upper bound b:

Key Properties

Real-World Applications

Time until the next customer arrives at a store
Lifetime of electronic components (time until failure)
Time between successive radioactive decays
Duration of phone calls in a call center
Time until the next earthquake in a region
Waiting time for the next bus arrival

Adjusting the Rate Parameter

Reading the PDF Curve

Understanding the CDF Display

The Explorer at Its Opening Parameter

Computing Point Probabilities

Calculating Cumulative Probabilities

Range Probability Calculations

What is the Exponential Distribution?

The Memoryless Property

Distribution Statistics

Related Distributions and Tools

Adjusting the Rate Parameter

Use the λ (lambda) slider to set the event rate - how frequently events occur per unit time. Values range from 0.1 (rare events, long waits) to 5 (frequent events, short waits).

As λ increases, the PDF curve becomes steeper and more concentrated near zero, reflecting shorter average waiting times. Lower λ creates a more gradual decline, indicating longer typical waits between events.

The mean waiting time equals 1/λ. At λ = 2 events per hour, average waiting time is 0.5 hours. At λ = 0.5, expect 2-hour average waits. The relationship is perfectly reciprocal.

Reading the PDF Curve

The PDF shows the characteristic exponentially decreasing curve, starting at maximum height λ at x = 0 and declining toward zero as x increases. The curve never touches zero but approaches it asymptotically.

The steepness of the decline is controlled by λ. Higher rates create steeper drops, while lower rates produce more gradual decreases. The curve always has its mode at x = 0 - the shortest possible waiting time.

About 63% of the probability mass lies within one mean (1/λ) from zero. The practical display range extends to about 4/λ, capturing over 98% of all probability.

Understanding the CDF Display

The CDF shows P(X ≤ x) = 1 - e^(-λx), rising from 0 and approaching 1 asymptotically. Unlike the normal's S-curve, the exponential CDF rises quickly initially then gradually levels off.

The curve's initial steepness indicates high probability of short waiting times. At x = 1/λ (the mean), the CDF equals approximately 0.632 - about 63.2% of events occur within one average waiting time.

The CDF never quite reaches 1, reflecting the theoretical possibility of arbitrarily long waiting times, though probabilities become negligible for large x.

The Explorer at Its Opening Parameter

\lambda = 1

the density starts at its maximum,

f(0) = \lambda = 1

, and decays from there. There is no interior peak: the most probable region is always immediately after zero.

The red line marks

E[X] = 1/\lambda = 1

, and the plotted window runs from

0

to about

6.91

Exponential, lambda = 1: PDF above, CDF below

Density is largest at the very edge of the support, f(0) = 1, with no interior peak. The window ends at -ln(0.001) = 6.908, so the CDF below stops at exactly 0.999 by construction.

That window edge is chosen, not arbitrary. The tool uses

x_{\max} = -\frac{\ln(0.001)}{\lambda} = 6.908

, the point where the surviving tail is exactly one thousandth — so the CDF tab ends at

0.999

by construction rather than by accident.

The mean sits at

1

but the median is

\frac{\ln 2}{\lambda} = 0.693

, noticeably to its left. That gap is the signature of right skew, and it means "average waiting time" and "typical waiting time" are different questions with different answers.

Mean and standard deviation are both

1/\lambda

, so the coefficient of variation is exactly

1

for every

\lambda

. Changing the rate rescales the whole picture without altering its shape — which is the memoryless property seen from the outside.

Computing Point Probabilities

For continuous distributions, individual point probabilities are always zero. Instead, use the PDF value at x, which gives the probability density: f(x) = λe^(-λx).

The PDF height indicates relative likelihood - higher values mean that region is more probable. The PDF at x = 0 equals λ, the maximum density, declining exponentially from there.

To find actual probabilities, you need intervals. Use the range calculator to find P(a ≤ X ≤ b), which computes the area under the PDF curve between a and b.

Calculating Cumulative Probabilities

Use P(X ≤ x) to find the probability of waiting x time units or less. The formula 1 - e^(-λx) provides exact values without numerical integration.

P(X > x) gives the survival probability - the chance of waiting more than x units. This equals e^(-λx), showing exponential decay. The memoryless property appears here: if you've waited time s, the probability of waiting additional time t is still e^(-λt).

P(X ≥ x) is identical to P(X > x) for continuous distributions since P(X = x) = 0. This differs from discrete distributions where the distinction matters.

Range Probability Calculations

The range calculator computes P(a ≤ X ≤ b) = F(b) - F(a) = e^(-λa) - e^(-λb), giving the probability that waiting time falls between a and b.

Four boundary options exist, but for continuous distributions all four give identical results since individual points have zero probability:
• [a, b], (a, b), [a, b), (a, b] all equal e^(-λa) - e^(-λb)

Example: For λ = 1, find P(0.5 ≤ X ≤ 2). This gives e^(-0.5) - e^(-2) ≈ 0.606 - 0.135 = 0.471, about 47% probability the wait is between 0.5 and 2 time units.

What is the Exponential Distribution?

The exponential distribution models waiting times between events in a Poisson process. If events occur randomly at constant rate λ, the time until the next event follows an exponential distribution.

The distribution requires constant event rate and independent occurrences. Each waiting period is independent of previous history - the memoryless property makes it unique among continuous distributions.

Applications include customer service times, equipment lifetimes, time between arrivals, radioactive decay intervals, and earthquake occurrences. For theoretical foundations, see exponential distribution theory page.

The Memoryless Property

The exponential is the only continuous distribution with the memoryless property: P(X > s+t | X > s) = P(X > t). If you've waited time s without an event, the probability of waiting additional time t is the same as starting fresh.

This means past waiting provides no information about future waiting. The distribution "forgets" how long you've already waited - each moment is statistically identical to any other.

The memoryless property arises from the constant hazard rate λ. Unlike aging processes where failure becomes more likely over time, exponential models truly random, time-independent events.

Distribution Statistics

The mean and standard deviation both equal 1/λ. With λ = 3 events per hour, average waiting time is 1/3 hour (20 minutes) with the same standard deviation.

The variance equals 1/λ², so variance equals mean squared. This high variance relative to the mean reflects the distribution's long right tail - occasional very long waits are possible.

The median equals (ln 2)/λ ≈ 0.693/λ, always less than the mean 1/λ. This reflects right skewness - the mode is at 0, median is lower than mean, creating positive skew.