Probability density and cumulative distribution functions
Studying this course? Keep your progress in one place.
Key facts
- A continuous random variable takes any value in an interval; probabilities are areas, and .
- A pdf has and total area ; .
- The cdf is ; .
- and .
- For a piecewise pdf, each later cdf piece must include the earlier pieces' probability.
Continuous random variables
Probabilities are areas, so any single value has probability zero.
A continuous random variable can take any value in an interval, such as a time, a mass or a length, so it cannot be listed value by value. Probabilities are found as areas, not heights.
For a continuous variable for every single value . So , and are all equal.
| Discrete | Continuous | |
|---|---|---|
| Values | Values can be listed | Any value in an interval |
| Single value | can be positive | |
| Probabilities | Probabilities are summed | Probabilities are areas |
Discrete
- Values:
- Values can be listed
- Single value:
- can be positive
- Probabilities:
- Probabilities are summed
Continuous
- Values:
- Any value in an interval
- Single value:
- Probabilities:
- Probabilities are areas
is continuous. Which statement is true?
The probability density function
It describes how probability is spread out; probabilities are areas under it, and the total area is 1.
The probability density function must satisfy and . Then
Here is restricted to simple polynomials, often piecewise: a different expression on each interval, and 0 outside. To find an unknown constant , set the total area equal to 1.
Worked example
for , and 0 otherwise, is a pdf. Find .
- 1
The total area is 1: .
- 2
.
for , and 0 otherwise. What is ?
The cumulative distribution function
The cdf gives the probability of being at or below .
The cumulative distribution function is . It rises from 0 to 1 and never decreases. Then and .
For a piecewise pdf, the cdf for a later piece must include the total probability of the earlier pieces.
for . What is ?
The link between f and F
The pdf is the derivative of the cdf.
Differentiate to get and integrate to get . For example, if for , then on the same interval, and outside it.
- f(x)
- F(x)
Worked example
for . Find and .
- 1
.
- 2
.
- 3
and .
for . What is on this interval?
Try an exam question
The continuous random variable has pdf for , and otherwise. Find and find .
[4 marks]
- [1]Total area is 1: .
- [1].
- [1] for .
- [1].
That's the notes covered.
Carry on to the next subtopic.
Try a question
Find the value of .
1 mark, marked against the mark scheme
Find the value of .
1 mark, marked against the mark scheme