Overview
The beta distribution describes uncertain proportions and probabilities, values between 0 and 1, such as a conversion rate or a project's completion share.
Enter x between 0 and 1, and the shape parameters α and β.
Worked examples
| Inputs | Result |
|---|---|
| x (0-1)0.42Alpha2.3Beta5.2 | Beta PDF at x=0.42: 1.62Approx CDF: 0.76Mean: 0.31 |
Show 2 more examples
| x (0-1)0.5Alpha2Beta2 | Beta PDF at x=0.5: 1.5Approx CDF: 0.5Mean: 0.5 |
| x (0-1)0.2Alpha21Beta81 | Beta PDF at x=0.2: 10.03Approx CDF: 0.46Mean: 0.21 |
Beta PDF at x=0.42 by alpha
With the calculator's starting values, the beta PDF at x=0.42 is 1.62. The table shows how it changes with the alpha.
| Alpha | Beta PDF at x=0.42 |
|---|---|
| 1.5 | 0.94 |
| 2 | 1.37 |
| 2.3starting value | 1.62 |
| 2.5 | 1.77 |
| 3 | 2.08 |
Other inputs: x: 0.42 0-1 · Beta: 5.2. Change them in the calculator above for your own numbers.
How it works
The PDF gives the relative likelihood at x, and the CDF, the chance of a value up to x, is found by numerical integration, so it's an approximation. The mean is α ÷ (α + β). In Bayesian terms, α − 1 and β − 1 can be read as prior successes and failures.
The formula
| α, β | Shape parameters, both above 0 |
Practical tips
- With α = 2.3 and β = 5.2, the mean is about 0.31 and P(X ≤ 0.42) is about 0.76.
- α = β = 1 is the uniform distribution, flat between 0 and 1.
- After 20 successes in 100 trials with a flat prior, the posterior is Beta(21, 81).
Frequently asked questions
Why can the PDF be above 1?
It's a density, not a probability. Only its area must equal 1.
Where is the beta distribution used?
A/B testing, project estimates (PERT), and modeling rates in Bayesian statistics.
How accurate is the CDF?
It uses Simpson's rule with 200 steps, which is usually accurate to two decimals but less so when α or β is below 1.