Cumulative Distribution Function (CDF)

When analyzing data, we often need to know the probability that a random variable falls within a certain range, or below a specific value. While Probability Density Functions (PDFs) and Probability Mass Functions (PMFs) tell us the likelihood at a single point or interval, they don't directly give us cumulative information. This is where the Cumulative Distribution Function (CDF) becomes indispensable, providing a complete picture of the probability distribution up to any given point.

What is the Cumulative Distribution Function (CDF)?

The CDF, denoted as $F_X(x)$ , quantifies the probability that a random variable $X$ will take a value less than or equal to a specific value $x$ . It essentially accumulates the probabilities from the lowest possible value up to $x$ . This function provides a comprehensive view of the distribution's shape, showing how probabilities accumulate across the variable's range.

📐 General CDF Definition

For any random variable $X$ , the Cumulative Distribution Function (CDF) is defined as:

F_X(x) = P(X \le x)

where $P(X \le x)$ is the probability that $X$ takes a value less than or equal to $x$ .

CDF for Continuous Random Variables

For a continuous random variable $X$ , the CDF is derived by integrating its Probability Density Function (PDF), $f_X(t)$ , from negative infinity up to the point $x$ . This integral represents the area under the PDF curve, which is the accumulated probability. The CDF for continuous variables is always a continuous function itself.

📐 Continuous CDF Formula

If $X$ is a continuous random variable with PDF $f_X(t)$ , its CDF is:

F_X(x) = \int_{-\infty}^{x} f_X(t) dt

For example, for a Uniform distribution $U(a,b)$ , the PDF is $f_X(t) = \frac{1}{b-a}$ for $a \le t \le b$ , and 0 otherwise.

pythonCalculating CDF for a Continuous Uniform Distribution

CDF for Discrete Random Variables

For a discrete random variable $X$ , the CDF is calculated by summing its Probability Mass Function (PMF), $f_X(t)$ , for all values $t$ less than or equal to $x$ . This summation accumulates the probabilities of individual outcomes. The CDF for discrete variables is a step function, meaning it increases in discrete jumps at each possible value of $X$ .

📐 Discrete CDF Formula

If $X$ is a discrete random variable with PMF $f_X(t)$ , its CDF is:

F_X(x) = \sum_{t \le x} P(X=t) = \sum_{t \le x} f_X(t)

For example, for a fair six-sided die, the PMF is $f_X(t) = \frac{1}{6}$ for $t \in \{1, 2, 3, 4, 5, 6\}$ .

pythonCalculating CDF for a Discrete Die Roll

Check Your Understanding

What is the maximum possible value a CDF can take?

Key Properties of CDFs

CDFs possess several fundamental properties that make them predictable and useful. First, they are monotonically non-decreasing, meaning $F_X(x_1) \le F_X(x_2)$ if $x_1 < x_2$ . This makes sense, as accumulating more probability can only increase or keep the cumulative sum the same. Second, the CDF always starts at 0 and ends at 1, reflecting that the probability of a variable being less than negative infinity is 0, and less than positive infinity is 1.

✨ CDF Boundary Conditions

Every valid CDF $F_X(x)$ must satisfy these conditions:

$0 \le F_X(x) \le 1$ for all $x$ .
$F_X(-\infty) = \lim_{x \to -\infty} F_X(x) = 0$ .
$F_X(\infty) = \lim_{x \to \infty} F_X(x) = 1$ .
$F_X(x)$ is non-decreasing: if $x_1 < x_2$ , then $F_X(x_1) \le F_X(x_2)$ .

Using CDF to Calculate Probabilities

The primary utility of the CDF is to easily calculate probabilities for various ranges. To find the probability that $X$ is less than or equal to a value $x$ , we directly use $F_X(x)$ . If we need the probability that $X$ is greater than $x$ , we use the complement rule: $1 - F_X(x)$ . For a range $a < X le b$ , we subtract the CDF values: $F_X(b) - F_X(a)$ . This makes CDFs incredibly versatile for practical probability calculations.

📐 CDF Probability Rules

Using the CDF, we can calculate various probabilities:

$P(X \le x) = F_X(x)$
$P(X > x) = 1 - F_X(x)$
$P(a < X \le b) = F_X(b) - F_X(a)$
For continuous variables, $P(X < x) = P(X \le x) = F_X(x)$ and $P(X=x) = 0$ .
For discrete variables, $P(X < x) = F_X(x^-)$ (where $x^-$ is the largest value less than $x$ ) and $P(X=x) = F_X(x) - F_X(x^-)$ .

pythonCalculating Probabilities with Normal CDF

Try It Yourself

Using the standard normal distribution (mean=0, std_dev=1), calculate the probability that a random variable

X

falls between -2 and 2 (i.e.,

P(-2 < X \le 2)

python

Check Your Understanding

F_X(5) = 0.8

, what does

P(X > 5)

equal?

Visualizing the CDF

A CDF plot typically forms an S-shaped curve for continuous distributions, or a series of steps for discrete ones. The curve starts at 0 on the left (negative infinity) and smoothly rises to 1 on the right (positive infinity). The steepness of the curve indicates regions where the probability accumulates rapidly, corresponding to higher probability densities (or masses). A flatter section means less probability is accumulating in that range.

Cumulative Distribution Function (CDF) of a Standard Normal Distribution

This chart displays the CDF of a standard normal distribution (mean=0, standard deviation=1). The y-axis represents the cumulative probability

F_X(x)

, while the x-axis represents the value

x

. The curve starts near 0, rises smoothly, and approaches 1, forming a characteristic S-shape.

Loading chart...

Key Insight: The S-shape of the CDF visually represents how probability accumulates across the range of a continuous random variable, with steeper sections indicating higher probability density.

Relationship with PDF/PMF

The CDF, PDF, and PMF are intrinsically linked. For continuous variables, the PDF is the derivative of the CDF, meaning $f_X(x) = \frac{d}{dx} F_X(x)$ . Conversely, the CDF is the integral of the PDF. For discrete variables, the PMF at a point $x$ can be found by taking the difference between the CDF at $x$ and the CDF just before $x$ (i.e., $f_X(x) = F_X(x) - F_X(x^-)$ ). Understanding these relationships allows us to move between different representations of a probability distribution.

CDF vs. PDF/PMF

Feature	Cumulative Distribution Function (CDF)	Probability Density/Mass Function (PDF/PMF)
What it measures	Cumulative probability $P(X \le x)$	Probability at a point (discrete) or density at a point (continuous)
Range of values	[0, 1]	Non-negative (can be > 1 for PDF, but integral is 1)
Function type (continuous)	Continuous, non-decreasing	Continuous, non-negative
Function type (discrete)	Step function, non-decreasing	Discrete values at specific points
Derivation from other	Integral of PDF; sum of PMF	Derivative of CDF (continuous); difference of CDF (discrete)
Use case	Calculating $P(X \le x)$ , $P(X > x)$ , $P(a < X \le b)$	Understanding likelihood at specific points/intervals, shape of distribution

A comparison highlighting the distinct roles and properties of CDFs and PDF/PMFs.

Key Takeaways

The Cumulative Distribution Function (CDF), $F_X(x) = P(X \le x)$ , gives the probability that a random variable $X$ takes a value less than or equal to $x$ .
For continuous variables, the CDF is the integral of the PDF: $F_X(x) = \int_{-\infty}^{x} f_X(t) dt$ .
For discrete variables, the CDF is the sum of the PMF: $F_X(x) = \sum_{t \le x} f_X(t)$ .
CDFs are always non-decreasing and range from 0 to 1, with $F_X(-\infty)=0$ and $F_X(\infty)=1$ .
Use CDFs to calculate probabilities for ranges: $P(a < X \le b) = F_X(b) - F_X(a)$ and $P(X > x) = 1 - F_X(x)$ .
The visual representation of a continuous CDF is typically an S-shaped curve, while a discrete CDF is a step function.

← All lessons in Data Distributions