Cumulative distribution function

Properties

Every cumulative distribution function ${\displaystyle F_X}$ is non-decreasing and right-continuous, which makes it a càdlàg function. Furthermore, $${\displaystyle \underset{x\to -\mathrm{\infty }}{lim}{F}_X(x)=0,\underset{x\to +\mathrm{\infty }}{lim}{F}_X(x)=1.}$$

Every function with these three properties is a CDF, i.e., for every such function, a random variable can be defined such that the function is the cumulative distribution function of that random variable.

If ${\displaystyle X}$ is a purely discrete random variable, then it attains values ${\displaystyle x_1,x_2,\dots }$ with probability ${\displaystyle p_i=p(x_i)}$, and the CDF of ${\displaystyle X}$ will be discontinuous at the points ${\displaystyle x_i}$: $${\displaystyle F_X(x)=\mathrm{P}(X\le x)=\sum _{x_i\le x}\mathrm{P}(X=x_i)=\sum _{x_i\le x}p(x_i).}$$

If the CDF ${\displaystyle F_X}$ of a real valued random variable ${\displaystyle X}$ is continuous, then ${\displaystyle X}$ is a continuous random variable; if furthermore ${\displaystyle F_X}$ is absolutely continuous, then there exists a Lebesgue-integrable function ${\displaystyle f_X(x)}$ such that $${\displaystyle F_X(b)-F_X(a)=\mathrm{P}(a<X\le b)={\int }_a^b{f}_X(x)dx}$$ for all real numbers ${\displaystyle a}$ and ${\displaystyle b}$. The function ${\displaystyle f_X}$ is equal to the derivative of ${\displaystyle F_X}$ almost everywhere, and it is called the probability density function of the distribution of ${\displaystyle X}$.

If ${\displaystyle X}$ has finite L1-norm, that is, the expectation of ${\displaystyle |X|}$ is finite, then the expectation is given by the Riemann–Stieltjes integral $${\displaystyle \mathrm{E}(X)={\int }_{-\mathrm{\infty }}^{\mathrm{\infty }}td{F}_X(t)}$$

and for any ${\displaystyle x\ge 0}$, $${\displaystyle x(1-F_X(x))\le {\int }_x^{\mathrm{\infty }}td{F}_X(t)}$$ as well as $${\displaystyle x{F}_X(-x)\le {\int }_{-\mathrm{\infty }}^{-x}(-t)d{F}_X(t)}$$ as shown in the diagram (consider the areas of the two red rectangles and their extensions to the right or left up to the graph of ${\displaystyle F_X}$). In particular, we have $${\displaystyle \underset{x\to -\mathrm{\infty }}{lim}x{F}_X(x)=0,\underset{x\to +\mathrm{\infty }}{lim}x(1-F_X(x))=0.}$$ In addition, the (finite) expected value of the real-valued random variable ${\displaystyle X}$ can be defined on the graph of its cumulative distribution function as illustrated by the drawing in the definition of expected value for arbitrary real-valued random variables.

Examples

As an example, suppose ${\displaystyle X}$ is uniformly distributed on the unit interval ${\displaystyle [0,1]}$. Then the CDF of ${\displaystyle X}$ is given by $${\displaystyle F_X(x)=\{\begin{array}{ll}0& :x<0\\ x& :0\le x\le 1\\ 1& :x>1\end{array}}$$

Suppose instead that ${\displaystyle X}$ takes only the discrete values 0 and 1, with equal probability. Then the CDF of ${\displaystyle X}$ is given by $${\displaystyle F_X(x)=\{\begin{array}{ll}0& :x<0\\ 1/2& :0\le x<1\\ 1& :x\ge 1\end{array}}$$

Suppose ${\displaystyle X}$ is exponential distributed. Then the CDF of ${\displaystyle X}$ is given by $${\displaystyle F_X(x;\lambda )=\{\begin{array}{ll}1-e^{-\lambda x}& x\ge 0,\\ 0& x<0.\end{array}}$$ Here λ > 0 is the parameter of the distribution, often called the rate parameter.

Suppose ${\displaystyle X}$ is normal distributed. Then the CDF of ${\displaystyle X}$ is given by $${\displaystyle F(x;\mu ,\sigma )=\frac{1}{\sigma \sqrt{2\pi }}{\int }_{-\mathrm{\infty }}^x\exp \left(-\frac{(t-\mu {)}^2}{2{\sigma }^2}\right)dt.}$$ Here the parameter ${\displaystyle \mu }$ is the mean or expectation of the distribution; and ${\displaystyle \sigma }$ is its standard deviation.

A table of the CDF of the standard normal distribution (${\displaystyle \mu =0}$ and ${\displaystyle \sigma =1}$) is often used in statistical applications, where it is named the standard normal table, the unit normal table, or the Z table.

Suppose ${\displaystyle X}$ is binomial distributed. Then the CDF of ${\displaystyle X}$ is given by $${\displaystyle F(x;n,p)=Pr(X\le x)=\sum _{i=0}^{min(n,\lfloor x\rfloor )}(\genfrac{}{}{0ex}{}{n}{i})p^i(1-p{)}^{n-i}.}$$ Here ${\displaystyle p}$ is the probability of success and the function denotes the discrete probability distribution of the number of successes in a sequence of ${\displaystyle n}$ independent experiments, and ${\displaystyle \lfloor x\rfloor }$ is the "floor" under ${\displaystyle x}$, i.e. the greatest integer less than or equal to ${\displaystyle x}$.

Suppose ${\displaystyle X}$ is the number of dots for an idealized roll of a die. Then its cumulative distribution function ${\displaystyle F:\mathbb{R}\to \mathbb{R}}$ is given by the adjacent drawing.

As mentioned above (at the end of section Properties), ${\displaystyle F}$ can be used to determine the expected value of ${\displaystyle X,}$ which leads to $${\displaystyle \begin{array}{cc}\mathrm{E}(X)& ={\int }_0^{\mathrm{\infty }}(1-F(x))dx-{\int }_{-\mathrm{\infty }}^{0}F(x)dx\\ & =\frac{1}{6}+\frac{2}{6}+\frac{3}{6}+\frac{4}{6}+\frac{5}{6}+\frac{6}{6}=3.5\end{array}}$$ (area of the yellow-highlighted surface; this sum also corresponds to the specific formula for random variables with finitely many outcomes).

Derived functions

Complementary cumulative distribution function (tail distribution)

Sometimes, it is useful to study the opposite question and ask how often the random variable is above a particular level. This is called the complementary cumulative distribution function (ccdf) or simply the tail distribution or exceedance, and is defined as $${\displaystyle {\overline{F}}_X(x)=\mathrm{P}(X>x)=1-F_X(x).}$$

This has applications in statistical hypothesis testing, for example, because the one-sided p-value is the probability of observing a test statistic at least as extreme as the one observed. Thus, provided that the test statistic, T, has a continuous distribution, the one-sided p-value is simply given by the ccdf: for an observed value ${\displaystyle t}$ of the test statistic $${\displaystyle p=\mathrm{P}(T\ge t)=\mathrm{P}(T>t)=1-F_T(t).}$$

In survival analysis, ${\displaystyle {\overline{F}}_X(x)}$ is called the survival function and denoted ${\displaystyle S(x)}$, while the term reliability function is common in engineering.

Properties

• For a non-negative continuous random variable having an expectation, Markov's inequality states that $${\displaystyle {\overline{F}}_X(x)\le \frac{\mathrm{E}(X)}{x}.}$$
• As ${\displaystyle x\to \mathrm{\infty },{\overline{F}}_X(x)\to 0}$, and in fact ${\displaystyle {\overline{F}}_X(x)=o(1/x)}$ provided that ${\displaystyle \mathrm{E}(X)}$ is finite.
  Proof:
  Assuming ${\displaystyle X}$ has a density function ${\displaystyle f_X}$, for any ${\displaystyle c>0}$ $${\displaystyle \mathrm{E}(X)={\int }_0^{\mathrm{\infty }}x{f}_X(x)dx\ge {\int }_0^{c}x{f}_X(x)dx+c{\int }_c^{\mathrm{\infty }}{f}_X(x)dx}$$ Then, on recognizing $${\displaystyle {\overline{F}}_X(c)={\int }_c^{\mathrm{\infty }}{f}_X(x)dx}$$ and rearranging terms, $${\displaystyle 0\le c{\overline{F}}_X(c)\le \mathrm{E}(X)-{\int }_0^{c}x{f}_X(x)dx\to 0\text{ as }c\to \mathrm{\infty }}$$ as claimed.
• For a random variable having an expectation, $${\displaystyle \mathrm{E}(X)={\int }_0^{\mathrm{\infty }}{\overline{F}}_X(x)dx-{\int }_{-\mathrm{\infty }}^0{F}_X(x)dx}$$ and for a non-negative random variable the second term is 0.
  If the random variable can only take non-negative integer values, this is equivalent to $${\displaystyle \mathrm{E}(X)=\sum _{n=0}^{\mathrm{\infty }}{\overline{F}}_X(n).}$$

Folded cumulative distribution

While the plot of a cumulative distribution ${\displaystyle F}$ often has an S-like shape, an alternative illustration is the folded cumulative distribution or mountain plot, which folds the top half of the graph over, that is

${\displaystyle F_{\text{fold}}(x)=F(x)1_{\{F(x)\le 0.5\}}+(1-F(x))1_{\{F(x)>0.5\}}}$

where ${\displaystyle 1_{\{A\}}}$ denotes the indicator function and the second summand is the survivor function, thus using two scales, one for the upslope and another for the downslope. This form of illustration emphasises the median, dispersion (specifically, the mean absolute deviation from the median) and skewness of the distribution or of the empirical results.

Inverse distribution function (quantile function)

If the CDF F is strictly increasing and continuous then ${\displaystyle F^{-1}(p),p\in [0,1],}$ is the unique real number ${\displaystyle x}$ such that ${\displaystyle F(x)=p}$. This defines the inverse distribution function or quantile function.

Some distributions do not have a unique inverse (for example if ${\displaystyle f_X(x)=0}$ for all ${\displaystyle a<x<b}$, causing ${\displaystyle F_X}$ to be constant). In this case, one may use the generalized inverse distribution function, which is defined as

${\displaystyle F^{-1}(p)=inf\{x\in \mathbb{R}:F(x)\ge p\},\mathrm{\forall }p\in [0,1].}$

• Example 1: The median is ${\displaystyle F^{-1}(0.5)}$.
• Example 2: Put ${\displaystyle \tau =F^{-1}(0.95)}$. Then we call ${\displaystyle \tau }$ the 95th percentile.

Some useful properties of the inverse cdf (which are also preserved in the definition of the generalized inverse distribution function) are:

1. ${\displaystyle F^{-1}}$ is nondecreasing
2. ${\displaystyle F^{-1}(F(x))\le x}$
3. ${\displaystyle F(F^{-1}(p))\ge p}$
4. ${\displaystyle F^{-1}(p)\le x}$ if and only if ${\displaystyle p\le F(x)}$
5. If ${\displaystyle Y}$ has a ${\displaystyle U[0,1]}$ distribution then ${\displaystyle F^{-1}(Y)}$ is distributed as ${\displaystyle F}$. This is used in random number generation using the inverse transform sampling-method.
6. If ${\displaystyle \{X_{\alpha }\}}$ is a collection of independent ${\displaystyle F}$-distributed random variables defined on the same sample space, then there exist random variables ${\displaystyle Y_{\alpha }}$ such that ${\displaystyle Y_{\alpha }}$ is distributed as ${\displaystyle U[0,1]}$ and ${\displaystyle F^{-1}(Y_{\alpha })=X_{\alpha }}$ with probability 1 for all ${\displaystyle \alpha }$.

The inverse of the cdf can be used to translate results obtained for the uniform distribution to other distributions.

Empirical distribution function

The empirical distribution function is an estimate of the cumulative distribution function that generated the points in the sample. It converges with probability 1 to that underlying distribution. A number of results exist to quantify the rate of convergence of the empirical distribution function to the underlying cumulative distribution function.

Multivariate case

Definition for two random variables

When dealing simultaneously with more than one random variable the joint cumulative distribution function can also be defined. For example, for a pair of random variables ${\displaystyle X,Y}$, the joint CDF ${\displaystyle F_{XY}}$ is given by

${\displaystyle F_{X,Y}(x,y)=\mathrm{P}(X\le x,Y\le y)}$ ()

where the right-hand side represents the probability that the random variable ${\displaystyle X}$ takes on a value less than or equal to ${\displaystyle x}$ and that ${\displaystyle Y}$ takes on a value less than or equal to ${\displaystyle y}$.

Example of joint cumulative distribution function:

For two continuous variables X and Y: $${\displaystyle Pr(a<X<b\text{ and }c<Y<d)={\int }_a^b{\int }_c^{d}f(x,y)dydx;}$$

For two discrete random variables, it is beneficial to generate a table of probabilities and address the cumulative probability for each potential range of X and Y, and here is the example:

given the joint probability mass function in tabular form, determine the joint cumulative distribution function.

Solution: using the given table of probabilities for each potential range of X and Y, the joint cumulative distribution function may be constructed in tabular form:

Definition for more than two random variables

For ${\displaystyle N}$ random variables ${\displaystyle X_1,\dots ,X_N}$, the joint CDF ${\displaystyle F_{X_1,\dots ,X_N}}$ is given by

${\displaystyle F_{X_1,\dots ,X_N}(x_1,\dots ,x_N)=\mathrm{P}(X_1\le x_1,\dots ,X_N\le x_N)}$ ()

Interpreting the ${\displaystyle N}$ random variables as a random vector ${\displaystyle \mathbf{X}=(X_1,\dots ,X_N{)}^T}$ yields a shorter notation: $${\displaystyle F_{\mathbf{X}}(\mathbf{x})=\mathrm{P}(X_1\le x_1,\dots ,X_N\le x_N)}$$

Properties

Every multivariate CDF is:

1. Monotonically non-decreasing for each of its variables,
2. Right-continuous in each of its variables,
3. ${\displaystyle 0\le F_{X_1\dots X_n}(x_1,\dots ,x_n)\le 1,}$
4. ${\displaystyle \underset{x_1,\dots ,x_n\to +\mathrm{\infty }}{lim}{F}_{X_1\dots X_n}(x_1,\dots ,x_n)=1}$ and ${\displaystyle \underset{x_i\to -\mathrm{\infty }}{lim}{F}_{X_1\dots X_n}(x_1,\dots ,x_n)=0,}$ for all i.

Not every function satisfying the above four properties is a multivariate CDF, unlike in the single dimension case. For example, let ${\displaystyle F(x,y)=0}$ for ${\displaystyle x<0}$ or ${\displaystyle x+y<1}$ or ${\displaystyle y<0}$ and let ${\displaystyle F(x,y)=1}$ otherwise. It is easy to see that the above conditions are met, and yet ${\displaystyle F}$ is not a CDF since if it was, then $\mathrm{P}\left(\frac{1}{3}<X\le 1,\frac{1}{3}<Y\le 1\right)=-1$ as explained below.

The probability that a point belongs to a hyperrectangle is analogous to the 1-dimensional case: $${\displaystyle F_{X_1,X_2}(a,c)+F_{X_1,X_2}(b,d)-F_{X_1,X_2}(a,d)-F_{X_1,X_2}(b,c)=\mathrm{P}(a<X_1\le b,c<X_2\le d)=\int \cdots }$$

Complex case

Complex random variable

The generalization of the cumulative distribution function from real to complex random variables is not obvious because expressions of the form ${\displaystyle P(Z\le 1+2i)}$ make no sense. However expressions of the form ${\displaystyle P(\mathrm{\Re }(Z)\le 1,\mathrm{\Im }(Z)\le 3)}$ make sense. Therefore, we define the cumulative distribution of a complex random variables via the joint distribution of their real and imaginary parts: $${\displaystyle F_Z(z)=F_{\mathrm{\Re }(Z),\mathrm{\Im }(Z)}(\mathrm{\Re }(z),\mathrm{\Im }(z))=P(\mathrm{\Re }(Z)\le \mathrm{\Re }(z),\mathrm{\Im }(Z)\le \mathrm{\Im }(z)).}$$

Complex random vector

Generalization of Eq.4 yields $${\displaystyle \begin{array}{cc}{F}_{\mathbf{Z}}(\mathbf{z})& =F_{\mathrm{\Re }(Z_1),\mathrm{\Im }(Z_1),\dots ,\mathrm{\Re }(Z_n),\mathrm{\Im }(Z_n)}(\mathrm{\Re }(z_1),\mathrm{\Im }(z_1),\dots ,\mathrm{\Re }(z_n),\mathrm{\Im }(z_n))\\ & =\mathrm{P}(\mathrm{\Re }(Z_1)\le \mathrm{\Re }(z_1),\mathrm{\Im }(Z_1)\le \mathrm{\Im }(z_1),\dots ,\mathrm{\Re }(Z_n)\le \mathrm{\Re }(z_n),\mathrm{\Im }(Z_n)\le \mathrm{\Im }(z_n))\end{array}}$$ as definition for the CDS of a complex random vector ${\displaystyle \mathbf{Z}=(Z_1,\dots ,Z_N{)}^T}$.

Use in statistical analysis

The concept of the cumulative distribution function makes an explicit appearance in statistical analysis in two (similar) ways. Cumulative frequency analysis is the analysis of the frequency of occurrence of values of a phenomenon less than a reference value. The empirical distribution function is a formal direct estimate of the cumulative distribution function for which simple statistical properties can be derived and which can form the basis of various statistical hypothesis tests. Such tests can assess whether there is evidence against a sample of data having arisen from a given distribution, or evidence against two samples of data having arisen from the same (unknown) population distribution.

Kolmogorov–Smirnov and Kuiper's tests

The Kolmogorov–Smirnov test is based on cumulative distribution functions and can be used to test to see whether two empirical distributions are different or whether an empirical distribution is different from an ideal distribution. The closely related Kuiper's test is useful if the domain of the distribution is cyclic as in day of the week. For instance Kuiper's test might be used to see if the number of tornadoes varies during the year or if sales of a product vary by day of the week or day of the month.