# 随机变量

In probability and statistics, a random variable, random quantity, aleatory variable, or stochastic variable is described informally as a variable whose values depend on outcomes of a random phenomenon.[1] The formal mathematical treatment of random variables is a topic in probability theory. In that context, a random variable is understood as a measurable function defined on a probability space that maps from the sample space to the real numbers.[2]

This graph shows how random variable is a function from all possible outcomes to real values. It also shows how random variable is used for defining probability mass functions.

A random variable's possible values might represent the possible outcomes of a yet-to-be-performed experiment, or the possible outcomes of a past experiment whose already-existing value is uncertain (for example, because of imprecise measurements or quantum uncertainty). They may also conceptually represent either the results of an "objectively" random process (such as rolling a die) or the "subjective" randomness that results from incomplete knowledge of a quantity. The meaning of the probabilities assigned to the potential values of a random variable is not part of probability theory itself, but is instead related to philosophical arguments over the interpretation of probability. The mathematics works the same regardless of the particular interpretation in use.

As a function, a random variable is required to be measurable, which allows for probabilities to be assigned to sets of its potential values. It is common that the outcomes depend on some physical variables that are not predictable. For example, when tossing a fair coin, the final outcome of heads or tails depends on the uncertain physical conditions, so the outcome being observed is uncertain. The coin could get caught in a crack in the floor, but such a possibility is excluded from consideration.

The domain of a random variable is called a sample space. It is interpreted as the set of possible outcomes of a random phenomenon. For example, in the case of a coin toss, only two possible outcomes are considered, namely heads or tails.

A random variable has a probability distribution, which specifies the probability of Borel subsets of its range. Random variables can be discrete, that is, taking any of a specified finite or countable list of values (having a countable range), endowed with a probability mass function that is characteristic of the random variable's probability distribution; or continuous, taking any numerical value in an interval or collection of intervals (having an uncountable range), via a probability density function that is characteristic of the random variable's probability distribution; or a mixture of both.

Two random variables with the same probability distribution can still differ in terms of their associations with, or independence from, other random variables. The realizations of a random variable, that is, the results of randomly choosing values according to the variable's probability distribution function, are called random variates.

## Definition定义

A random variable is a measurable function $\displaystyle{ X \colon \Omega \to E }$ from a set of possible outcomes $\displaystyle{ \Omega }$ to a measurable space $\displaystyle{ E }$. The technical axiomatic definition requires $\displaystyle{ \Omega }$ to be a sample space of a probability triple $\displaystyle{ (\Omega, \mathcal{F}, \operatorname{P}) }$ (see the measure-theoretic definition). A random variable is often denoted by capital roman letters such as $\displaystyle{ X }$, $\displaystyle{ Y }$, $\displaystyle{ Z }$, $\displaystyle{ T }$.[3][4]

The probability that $\displaystyle{ X }$ takes on a value in a measurable set $\displaystyle{ S\subseteq E }$ is written as

$\displaystyle{ \operatorname{P}(X \in S) = \operatorname{P}(\{ \omega \in \Omega \mid X(\omega) \in S \}) }$[3]

$\displaystyle{ \operatorname{P}(X \in S) = \operatorname{P}(\{ \omega \in \Omega \mid X(\omega) \in S \}) }$[3]

### Standard case一般示例

In many cases, $\displaystyle{ X }$ is real-valued, i.e. $\displaystyle{ E = \mathbb{R} }$. In some contexts, the term random element (see extensions) is used to denote a random variable not of this form.

When the image (or range) of $\displaystyle{ X }$ is countable, the random variable is called a discrete random variable[5]:399 and its distribution is a discrete probability distribution, i.e. can be described by a probability mass function that assigns a probability to each value in the image of $\displaystyle{ X }$. If the image is uncountably infinite (usually an interval) then $\displaystyle{ X }$ is called a continuous random variable.[6] In the special case that it is absolutely continuous, its distribution can be described by a probability density function, which assigns probabilities to intervals; in particular, each individual point must necessarily have probability zero for an absolutely continuous random variable. Not all continuous random variables are absolutely continuous,[7] a mixture distribution is one such counterexample; such random variables cannot be described by a probability density or a probability mass function.

$\displaystyle{ X }$的像（或范围）是可数的，称随机变量是离散型随机变量[5]:399，称其分布是离散型概率分布，如可以用一个给$\displaystyle{ X }$的像的每个值赋一个概率的概率质量函数来描述。如果像是无限不可数的（通常是一个区间），那么就称$\displaystyle{ X }$连续型随机变量。[6]绝对连续的特殊情况下，它的分布可以用一个为区间赋予概率的概率密度函数来描述；特别地，对于绝对连续型随机变量，每个单独的点的概率都为零。并不是所有的连续型随机变量都是绝对连续的[7]混合分布就是这样的一个反例；这样的随机变量不能用概率密度函数或概率质量函数来描述。

Any random variable can be described by its cumulative distribution function, which describes the probability that the random variable will be less than or equal to a certain value.

The term "random variable" in statistics is traditionally limited to the real-valued case (E=\mathbb{R}). In this case, the structure of the real numbers makes it possible to define quantities such as the expected value and variance of a random variable, its cumulative distribution function, and the moments of its distribution.

However, the definition above is valid for any measurable space E of values. Thus one can consider random elements of other sets E, such as random boolean values, categorical values, complex numbers, vectors, matrices, sequences, trees, sets, shapes, manifolds, and functions. One may then specifically refer to a random variable of type E, or an E-valued random variable.

### Extensions

This more general concept of a random element is particularly useful in disciplines such as graph theory, machine learning, natural language processing, and other fields in discrete mathematics and computer science, where one is often interested in modeling the random variation of non-numerical data structures. In some cases, it is nonetheless convenient to represent each element of E, using one or more real numbers. In this case, a random element may optionally be represented as a vector of real-valued random variables (all defined on the same underlying probability space \Omega, which allows the different random variables to covary). For example:

The term "random variable" in statistics is traditionally limited to the real-valued case ($\displaystyle{ E=\mathbb{R} }$). In this case, the structure of the real numbers makes it possible to define quantities such as the expected value and variance of a random variable, its cumulative distribution function, and the moments of its distribution.

However, the definition above is valid for any measurable space $\displaystyle{ E }$ of values. Thus one can consider random elements of other sets $\displaystyle{ E }$, such as random boolean values, categorical values, complex numbers, vectors, matrices, sequences, trees, sets, shapes, manifolds, and functions. One may then specifically refer to a random variable of type $\displaystyle{ E }$, or an $\displaystyle{ E }$-valued random variable.

This more general concept of a random element is particularly useful in disciplines such as graph theory, machine learning, natural language processing, and other fields in discrete mathematics and computer science, where one is often interested in modeling the random variation of non-numerical data structures. In some cases, it is nonetheless convenient to represent each element of $\displaystyle{ E }$, using one or more real numbers. In this case, a random element may optionally be represented as a vector of real-valued random variables (all defined on the same underlying probability space $\displaystyle{ \Omega }$, which allows the different random variables to covary). For example:

• A random word may be represented as a random integer that serves as an index into the vocabulary of possible words. Alternatively, it can be represented as a random indicator vector, whose length equals the size of the vocabulary, where the only values of positive probability are $\displaystyle{ (1 \ 0 \ 0 \ 0 \ \cdots) }$, $\displaystyle{ (0 \ 1 \ 0 \ 0 \ \cdots) }$, $\displaystyle{ (0 \ 0 \ 1 \ 0 \ \cdots) }$ and the position of the 1 indicates the word.
• A random sentence of given length $\displaystyle{ N }$ may be represented as a vector of $\displaystyle{ N }$ random words.

If a random variable X\colon \Omega \to \mathbb{R} defined on the probability space (\Omega, \mathcal{F}, \operatorname{P}) is given, we can ask questions like "How likely is it that the value of X is equal to 2?". This is the same as the probability of the event \{ \omega : X(\omega) = 2 \}\,\! which is often written as P(X = 2)\,\! or p_X(2) for short.

• A random graph on $\displaystyle{ N }$ given vertices may be represented as a $\displaystyle{ N \times N }$ matrix of random variables, whose values specify the adjacency matrix of the random graph.
• A random function $\displaystyle{ F }$ may be represented as a collection of random variables $\displaystyle{ F(x) }$, giving the function's values at the various points $\displaystyle{ x }$ in the function's domain. The $\displaystyle{ F(x) }$ are ordinary real-valued random variables provided that the function is real-valued. For example, a stochastic process is a random function of time, a random vector is a random function of some index set such as $\displaystyle{ 1,2,\ldots, n }$, and random field is a random function on any set (typically time, space, or a discrete set).

Recording all these probabilities of output ranges of a real-valued random variable X yields the probability distribution of X. The probability distribution "forgets" about the particular probability space used to define X and only records the probabilities of various values of X. Such a probability distribution can always be captured by its cumulative distribution function

## Distribution functions

F_X(x) = \operatorname{P}(X \le x)

F _ X (x) = 操作者名{ P }(X le x)

If a random variable $\displaystyle{ X\colon \Omega \to \mathbb{R} }$ defined on the probability space $\displaystyle{ (\Omega, \mathcal{F}, \operatorname{P}) }$ is given, we can ask questions like "How likely is it that the value of $\displaystyle{ X }$ is equal to 2?". This is the same as the probability of the event $\displaystyle{ \{ \omega : X(\omega) = 2 \}\,\! }$ which is often written as $\displaystyle{ P(X = 2)\,\! }$ or $\displaystyle{ p_X(2) }$ for short.

and sometimes also using a probability density function, p_X. In measure-theoretic terms, we use the random variable X to "push-forward" the measure P on \Omega to a measure p_X on \mathbb{R}.

The underlying probability space \Omega is a technical device used to guarantee the existence of random variables, sometimes to construct them, and to define notions such as correlation and dependence or independence based on a joint distribution of two or more random variables on the same probability space. In practice, one often disposes of the space \Omega altogether and just puts a measure on \mathbb{R} that assigns measure 1 to the whole real line, i.e., one works with probability distributions instead of random variables. See the article on quantile functions for fuller development.

Recording all these probabilities of output ranges of a real-valued random variable $\displaystyle{ X }$ yields the probability distribution of $\displaystyle{ X }$. The probability distribution "forgets" about the particular probability space used to define $\displaystyle{ X }$ and only records the probabilities of various values of $\displaystyle{ X }$. Such a probability distribution can always be captured by its cumulative distribution function

$\displaystyle{ F_X(x) = \operatorname{P}(X \le x) }$

and sometimes also using a probability density function, $\displaystyle{ p_X }$. In measure-theoretic terms, we use the random variable $\displaystyle{ X }$ to "push-forward" the measure $\displaystyle{ P }$ on $\displaystyle{ \Omega }$ to a measure $\displaystyle{ p_X }$ on $\displaystyle{ \mathbb{R} }$.

The underlying probability space $\displaystyle{ \Omega }$ is a technical device used to guarantee the existence of random variables, sometimes to construct them, and to define notions such as correlation and dependence or independence based on a joint distribution of two or more random variables on the same probability space. In practice, one often disposes of the space $\displaystyle{ \Omega }$ altogether and just puts a measure on $\displaystyle{ \mathbb{R} }$ that assigns measure 1 to the whole real line, i.e., one works with probability distributions instead of random variables. See the article on quantile functions for fuller development.

In an experiment a person may be chosen at random, and one random variable may be the person's height. Mathematically, the random variable is interpreted as a function which maps the person to the person's height. Associated with the random variable is a probability distribution that allows the computation of the probability that the height is in any subset of possible values, such as the probability that the height is between 180 and 190 cm, or the probability that the height is either less than 150 or more than 200 cm.

## Examples

Another random variable may be the person's number of children; this is a discrete random variable with non-negative integer values. It allows the computation of probabilities for individual integer values – the probability mass function (PMF) – or for sets of values, including infinite sets. For example, the event of interest may be "an even number of children". For both finite and infinite event sets, their probabilities can be found by adding up the PMFs of the elements; that is, the probability of an even number of children is the infinite sum \operatorname{PMF}(0) + \operatorname{PMF}(2) + \operatorname{PMF}(4) + \cdots.

### Discrete random variable

In examples such as these, the sample space is often suppressed, since it is mathematically hard to describe, and the possible values of the random variables are then treated as a sample space. But when two random variables are measured on the same sample space of outcomes, such as the height and number of children being computed on the same random persons, it is easier to track their relationship if it is acknowledged that both height and number of children come from the same random person, for example so that questions of whether such random variables are correlated or not can be posed.

In an experiment a person may be chosen at random, and one random variable may be the person's height. Mathematically, the random variable is interpreted as a function which maps the person to the person's height. Associated with the random variable is a probability distribution that allows the computation of the probability that the height is in any subset of possible values, such as the probability that the height is between 180 and 190 cm, or the probability that the height is either less than 150 or more than 200 cm.

If $\displaystyle{ \{a_n\}, \{b_n\} }$ are countable sets of real numbers, $\displaystyle{ b_n \gt 0 }$ and \sum_n b_n=1, then F=\sum_n b_n \delta_{a_n} is a discrete distribution function. Here \delta_t(x) = 0 for x < t, \delta_t(x) = 1 for x \ge t. Taking for instance an enumeration of all rational numbers as \{a_n\}, one gets a discrete distribution function that is not a step function or piecewise constant. There are no "gaps", which would correspond to numbers which have a finite probability of occurring. Instead, continuous random variables almost never take an exact prescribed value c (formally, $\displaystyle{ \forall c \in \mathbb{R}:\; \Pr(X = c) = 0 }$) but there is a positive probability that its value will lie in particular intervals which can be arbitrarily small. Continuous random variables usually admit probability density functions (PDF), which characterize their CDF and probability measures;

such distributions are also called absolutely continuous; but some continuous distributions are singular, or mixes of an absolutely continuous part and a singular part.

Another random variable may be the person's number of children; this is a discrete random variable with non-negative integer values. It allows the computation of probabilities for individual integer values – the probability mass function (PMF) – or for sets of values, including infinite sets. For example, the event of interest may be "an even number of children". For both finite and infinite event sets, their probabilities can be found by adding up the PMFs of the elements; that is, the probability of an even number of children is the infinite sum $\displaystyle{ \operatorname{PMF}(0) + \operatorname{PMF}(2) + \operatorname{PMF}(4) + \cdots }$.

An example of a continuous random variable would be one based on a spinner that can choose a horizontal direction. Then the values taken by the random variable are directions. We could represent these directions by North, West, East, South, Southeast, etc. However, it is commonly more convenient to map the sample space to a random variable which takes values which are real numbers. This can be done, for example, by mapping a direction to a bearing in degrees clockwise from North. The random variable then takes values which are real numbers from the interval [0, 360), with all parts of the range being "equally likely". In this case, X = the angle spun. Any real number has probability zero of being selected, but a positive probability can be assigned to any range of values. For example, the probability of choosing a number in [0, 180] is . Instead of speaking of a probability mass function, we say that the probability density of X is 1/360. The probability of a subset of [0, 360) can be calculated by multiplying the measure of the set by 1/360. In general, the probability of a set for a given continuous random variable can be calculated by integrating the density over the given set.

In examples such as these, the sample space is often suppressed, since it is mathematically hard to describe, and the possible values of the random variables are then treated as a sample space. But when two random variables are measured on the same sample space of outcomes, such as the height and number of children being computed on the same random persons, it is easier to track their relationship if it is acknowledged that both height and number of children come from the same random person, for example so that questions of whether such random variables are correlated or not can be posed.

More formally, given any interval $\displaystyle{ I = [a, b] = \{x \in \mathbb{R} : a \le x \le b \} }$, a random variable X_I \sim \operatorname{U}(I) = \operatorname{U}[a, b] is called a "continuous uniform random variable" (CURV) if the probability that it takes a value in a subinterval depends only on the length of the subinterval. This implies that the probability of X_I falling in any subinterval [c, d] \sube [a, b] is proportional to the length of the subinterval, that is, if , one has

If $\displaystyle{ \{a_n\}, \{b_n\} }$ are countable sets of real numbers, $\displaystyle{ b_n \gt 0 }$ and $\displaystyle{ \sum_n b_n=1 }$, then $\displaystyle{ F=\sum_n b_n \delta_{a_n} }$ is a discrete distribution function. Here $\displaystyle{ \delta_t(x) = 0 }$ for $\displaystyle{ x \lt t }$, $\displaystyle{ \delta_t(x) = 1 }$ for $\displaystyle{ x \ge t }$. Taking for instance an enumeration of all rational numbers as $\displaystyle{ \{a_n\} }$, one gets a discrete distribution function that is not a step function or piecewise constant.[5]

$\displaystyle{ “数学显示屏” ====Coin toss==== 抛硬币 \Pr\left( X_I \in [c,d]\right) Pr 左(x _ i 在[ c，d ]右) = \frac{d - c}{b - a}\Pr\left( X_I \in I\right)= \frac{d - c}{b - a} }$

= frac { d-c }{ b-a } Pr left (x _ i in i right) = frac { d-c }{ b-a } </math >

The possible outcomes for one coin toss can be described by the sample space $\displaystyle{ \Omega = \{\text{heads}, \text{tails}\} }$. We can introduce a real-valued random variable $\displaystyle{ Y }$ that models a \$1 payoff for a successful bet on heads as follows:

$\displaystyle{ where the last equality results from the unitarity axiom of probability. The probability density function of a CURV X \sim \operatorname {U}[a, b] is given by the indicator function of its interval of support normalized by the interval's length: \lt math display="block"\gt f_X(x) = \begin{cases} 最后一个等式来自于统一性的概率公理。一个 CURV x sim 操作器名称{ u }[ a，b ]的概率密度函数是由它的支持区间的指示函数被区间的长度归一化得到的: \lt math display = " block" \gt f _ x (x) = begin { cases } Y(\omega) = \displaystyle{1 \over b-a}, & a \le x \le b \\ 显示风格{1 over b-a } ，& a le x le b \begin{cases} 0, & \text{otherwise}. 0，& text { otherwise }. 1, & \text{if } \omega = \text{heads}, \\[6pt] \end{cases} }$Of particular interest is the uniform distribution on the unit interval [0, 1]. Samples of any desired probability distribution \operatorname{D} can be generated by calculating the quantile function of \operatorname{D} on a randomly-generated number distributed uniformly on the unit interval. This exploits properties of cumulative distribution functions, which are a unifying framework for all random variables.

0, & \text{if } \omega = \text{tails}.

\end{cases}

[/itex]

A mixed random variable is a random variable whose cumulative distribution function is neither piecewise-constant (a discrete random variable) nor everywhere-continuous. This definition enables us to measure any subset B\in \mathcal{E} in the target space by looking at its preimage, which by assumption is measurable.

If the coin is a fair coin, Y has a probability mass function $\displaystyle{ f_Y }$ given by:

In more intuitive terms, a member of \Omega is a possible outcome, a member of \mathcal{F} is a measurable subset of possible outcomes, the function P gives the probability of each such measurable subset, E represents the set of values that the random variable can take (such as the set of real numbers), and a member of \mathcal{E} is a "well-behaved" (measurable) subset of E (those for which the probability may be determined). The random variable is then a function from any outcome to a quantity, such that the outcomes leading to any useful subset of quantities for the random variable have a well-defined probability.

$\displaystyle{ f_Y(y) = When E is a topological space, then the most common choice for the σ-algebra \mathcal{E} is the Borel σ-algebra \mathcal{B}(E), which is the σ-algebra generated by the collection of all open sets in E. In such case the (E, \mathcal{E})-valued random variable is called an E-valued random variable. Moreover, when the space E is the real line \mathbb{R}, then such a real-valued random variable is called simply a random variable. 当E是一个拓扑空间时，那么最常见的σ代数/mathcal{E}的选择是博尔σ代数/mathcal{B}(E)，它是由E中所有开放集的集合生成的σ代数，在这种情况下，(E，\mathcal{E})值的随机变量称为E值的随机变量。此外，当空间E为实线\mathbb{R}时，则这样的实值随机变量称为简单随机变量。 \begin{cases} \tfrac 12,& \text{if }y=1,\\[6pt] \tfrac 12,& \text{if }y=0, \end{cases} In this case the observation space is the set of real numbers. Recall, (\Omega, \mathcal{F}, P) is the probability space. For a real observation space, the function X\colon \Omega \rightarrow \mathbb{R} is a real-valued random variable if 在这种情况下，观测空间是实数的集合。回想一下，(Omega，mathcal {F} ，P)是概率空间。对于一个真实的观察空间，函数 Xcolon Omega right tarrow mathbb {R}是一个真值随机变量，如果 }$

\{ \omega : X(\omega) \le r \} \in \mathcal{F} \qquad \forall r \in \mathbb{R}.

{ omega: X (omega) le r }{F} qquad for all r {F}。

#### Dice roll

This definition is a special case of the above because the set \{(-\infty, r]: r \in \R\} generates the Borel σ-algebra on the set of real numbers, and it suffices to check measurability on any generating set. Here we can prove measurability on this generating set by using the fact that \{ \omega : X(\omega) \le r \} = X^{-1}((-\infty, r]).

If the sample space is the set of possible numbers rolled on two dice, and the random variable of interest is the sum S of the numbers on the two dice, then S is a discrete random variable whose distribution is described by the probability mass function plotted as the height of picture columns here.

A random variable can also be used to describe the process of rolling dice and the possible outcomes. The most obvious representation for the two-dice case is to take the set of pairs of numbers n1 and n2 from {1, 2, 3, 4, 5, 6} (representing the numbers on the two dice) as the sample space. The total number rolled (the sum of the numbers in each pair) is then a random variable X given by the function that maps the pair to the sum:

The probability distribution of a random variable is often characterised by a small number of parameters, which also have a practical interpretation. For example, it is often enough to know what its "average value" is. This is captured by the mathematical concept of expected value of a random variable, denoted \operatorname{E}[X], and also called the first moment. In general, \operatorname{E}[f(X)] is not equal to f(\operatorname{E}[X]). Once the "average value" is known, one could then ask how far from this average value the values of X typically are, a question that is answered by the variance and standard deviation of a random variable. \operatorname{E}[X] can be viewed intuitively as an average obtained from an infinite population, the members of which are particular evaluations of X.

$\displaystyle{ X((n_1, n_2)) = n_1 + n_2 }$

Mathematically, this is known as the (generalised) problem of moments: for a given class of random variables X, find a collection \{f_i\} of functions such that the expectation values \operatorname{E}[f_i(X)] fully characterise the distribution of the random variable X.

and (if the dice are fair) has a probability mass function ƒX given by:

Moments can only be defined for real-valued functions of random variables (or complex-valued, etc.). If the random variable is itself real-valued, then moments of the variable itself can be taken, which are equivalent to moments of the identity function f(X)=X of the random variable. However, even for non-real-valued random variables, moments can be taken of real-valued functions of those variables. For example, for a categorical random variable X that can take on the nominal values "red", "blue" or "green", the real-valued function [X = \text{green}] can be constructed; this uses the Iverson bracket, and has the value 1 if X has the value "green", 0 otherwise. Then, the expected value and other moments of this function can be determined.

$\displaystyle{ f_X(S) = \frac{\min(S-1, 13-S)}{36}, \text{ for } S \in \{2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12\} }$

### Continuous random variable

A new random variable Y can be defined by applying a real Borel measurable function g\colon \mathbb{R} \rightarrow \mathbb{R} to the outcomes of a real-valued random variable X. That is, Y=g(X). The cumulative distribution function of Y is then

Formally, a continuous random variable is a random variable whose cumulative distribution function is continuous everywhere.[8] There are no "gaps", which would correspond to numbers which have a finite probability of occurring. Instead, continuous random variables almost never take an exact prescribed value c (formally, $\displaystyle{ \forall c \in \mathbb{R}:\; \Pr(X = c) = 0 }$) but there is a positive probability that its value will lie in particular intervals which can be arbitrarily small. Continuous random variables usually admit probability density functions (PDF), which characterize their CDF and probability measures;

F_Y(y) = \operatorname{P}(g(X) \le y).

F _ Y(y) ={P}(g (X) le y)。

such distributions are also called absolutely continuous; but some continuous distributions are singular, or mixes of an absolutely continuous part and a singular part.

If function g is invertible (i.e., h = g^{-1} exists, where h is g's inverse function) and is either increasing or decreasing, then the previous relation can be extended to obtain

An example of a continuous random variable would be one based on a spinner that can choose a horizontal direction. Then the values taken by the random variable are directions. We could represent these directions by North, West, East, South, Southeast, etc. However, it is commonly more convenient to map the sample space to a random variable which takes values which are real numbers. This can be done, for example, by mapping a direction to a bearing in degrees clockwise from North. The random variable then takes values which are real numbers from the interval [0, 360), with all parts of the range being "equally likely". In this case, X = the angle spun. Any real number has probability zero of being selected, but a positive probability can be assigned to any range of values. For example, the probability of choosing a number in [0, 180] is 模板:Frac. Instead of speaking of a probability mass function, we say that the probability density of X is 1/360. The probability of a subset of [0, 360) can be calculated by multiplying the measure of the set by 1/360. In general, the probability of a set for a given continuous random variable can be calculated by integrating the density over the given set.

$\displaystyle{ F_Y(y) = \operatorname{P}(g(X) \le y) = \lt math \gt F _ y (y) = p }(g (x) le y) = More formally, given any [[Interval (mathematics)|interval]] \lt math display="inline"\gt I = [a, b] = \{x \in \mathbb{R} : a \le x \le b \} }$, a random variable $\displaystyle{ X_I \sim \operatorname{U}(I) = \operatorname{U}[a, b] }$ is called a "continuous uniform random variable" (CURV) if the probability that it takes a value in a subinterval depends only on the length of the subinterval. This implies that the probability of $\displaystyle{ X_I }$ falling in any subinterval $\displaystyle{ [c, d] \sube [a, b] }$ is proportional to the length of the subinterval, that is, if acdb, one has

\begin{cases}

{ cases }

\operatorname{P}(X \le h(y)) = F_X(h(y)),

{ p }(x le h (y)) = f _ x (h (y)) ,

$\displaystyle{ & \text{if } h = g^{-1} \text{ increasing} ,\\ 文本{ if } h = g ^ {-1} text { incremental } , \Pr\left( X_I \in [c,d]\right) \\ \\ = \frac{d - c}{b - a}\Pr\left( X_I \in I\right)= \frac{d - c}{b - a} }$

\operatorname{P}(X \ge h(y)) = 1 - F_X(h(y)),

{ p }(xgeh (y)) = 1-f _ x (h (y)) ,

& \text{if } h = g^{-1} \text{ decreasing} .


& text { if } h = g ^ {-1} text { reducing }.

where the last equality results from the unitarity axiom of probability. The probability density function of a CURV $\displaystyle{ X \sim \operatorname {U}[a, b] }$ is given by the indicator function of its interval of support normalized by the interval's length: $\displaystyle{ f_X(x) = \begin{cases} \end{cases} }$

{ cases } </math >

\displaystyle{1 \over b-a}, & a \le x \le b \\

 0, & \text{otherwise}.


With the same hypotheses of invertibility of g, assuming also differentiability, the relation between the probability density functions can be found by differentiating both sides of the above expression with respect to y, in order to obtain

\end{cases}[/itex]Of particular interest is the uniform distribution on the unit interval $\displaystyle{ [0, 1] }$. Samples of any desired probability distribution $\displaystyle{ \operatorname{D} }$ can be generated by calculating the quantile function of $\displaystyle{ \operatorname{D} }$ on a randomly-generated number distributed uniformly on the unit interval. This exploits properties of cumulative distribution functions, which are a unifying framework for all random variables.

f_Y(y) = f_X\bigl(h(y)\bigr) \left| \frac{d h(y)}{d y} \right|.

F _ y (y) = f _ x bigl (h (y) bigr)左 | frac { d h (y)}{ d y }右 | 。

### Mixed type

If there is no invertibility of g but each y admits at most a countable number of roots (i.e., a finite, or countably infinite, number of x_i such that y = g(x_i)) then the previous relation between the probability density functions can be generalized with

A mixed random variable is a random variable whose cumulative distribution function is neither piecewise-constant (a discrete random variable) nor everywhere-continuous.[8] It can be realized as the sum of a discrete random variable and a continuous random variable; in which case the 模板:Abbr will be the weighted average of the CDFs of the component variables.[8]

f_Y(y) = \sum_{i} f_X(g_{i}^{-1}(y)) \left| \frac{d g_{i}^{-1}(y)}{d y} \right|

f _ Y (y) = sum _ { i } f _ x (g _ { i } ^ {-1}(y))) | frac { g _ { i } ^ {-1}(y)}{ d y } |

An example of a random variable of mixed type would be based on an experiment where a coin is flipped and the spinner is spun only if the result of the coin toss is heads. If the result is tails, X = −1; otherwise X = the value of the spinner as in the preceding example. There is a probability of 模板:Frac that this random variable will have the value −1. Other ranges of values would have half the probabilities of the last example.

where x_i = g_i^{-1}(y), according to the inverse function theorem. The formulas for densities do not demand g to be increasing.

Most generally, every probability distribution on the real line is a mixture of discrete part, singular part, and an absolutely continuous part; see 模板:Section link. The discrete part is concentrated on a countable set, but this set may be dense (like the set of all rational numbers).

In the measure-theoretic, axiomatic approach to probability, if a random variable X on \Omega and a Borel measurable function g\colon \mathbb{R} \rightarrow \mathbb{R}, then Y = g(X) is also a random variable on \Omega, since the composition of measurable functions is also measurable. (However, this is not necessarily true if g is Lebesgue measurable.) The same procedure that allowed one to go from a probability space (\Omega, P) to (\mathbb{R}, dF_{X}) can be used to obtain the distribution of Y.

## Measure-theoretic definition

The most formal, axiomatic definition of a random variable involves measure theory. Continuous random variables are defined in terms of sets of numbers, along with functions that map such sets to probabilities. Because of various difficulties (e.g. the Banach–Tarski paradox) that arise if such sets are insufficiently constrained, it is necessary to introduce what is termed a sigma-algebra to constrain the possible sets over which probabilities can be defined. Normally, a particular such sigma-algebra is used, the Borel σ-algebra, which allows for probabilities to be defined over any sets that can be derived either directly from continuous intervals of numbers or by a finite or countably infinite number of unions and/or intersections of such intervals.[2]

Let X be a real-valued, continuous random variable and let Y = X^2.

The measure-theoretic definition is as follows.

F_Y(y) = \operatorname{P}(X^2 \le y).

F _ Y (y) ={ p }(X ^ 2 le y)。

Let $\displaystyle{ (\Omega, \mathcal{F}, P) }$ be a probability space and $\displaystyle{ (E, \mathcal{E}) }$ a measurable space. Then an $\displaystyle{ (E, \mathcal{E}) }$-valued random variable is a measurable function $\displaystyle{ X\colon \Omega \to E }$, which means that, for every subset $\displaystyle{ B\in\mathcal{E} }$, its preimage $\displaystyle{ X^{-1}(B)\in \mathcal{F} }$ where $\displaystyle{ X^{-1}(B) = \{\omega : X(\omega)\in B\} }$.[9] This definition enables us to measure any subset $\displaystyle{ B\in \mathcal{E} }$ in the target space by looking at its preimage, which by assumption is measurable.

If y < 0, then P(X^2 \leq y) = 0, so

In more intuitive terms, a member of $\displaystyle{ \Omega }$ is a possible outcome, a member of $\displaystyle{ \mathcal{F} }$ is a measurable subset of possible outcomes, the function $\displaystyle{ P }$ gives the probability of each such measurable subset, $\displaystyle{ E }$ represents the set of values that the random variable can take (such as the set of real numbers), and a member of $\displaystyle{ \mathcal{E} }$ is a "well-behaved" (measurable) subset of $\displaystyle{ E }$ (those for which the probability may be determined). The random variable is then a function from any outcome to a quantity, such that the outcomes leading to any useful subset of quantities for the random variable have a well-defined probability.

When $\displaystyle{ E }$ is a topological space, then the most common choice for the σ-algebra $\displaystyle{ \mathcal{E} }$ is the Borel σ-algebra $\displaystyle{ \mathcal{B}(E) }$, which is the σ-algebra generated by the collection of all open sets in $\displaystyle{ E }$. In such case the $\displaystyle{ (E, \mathcal{E}) }$-valued random variable is called an $\displaystyle{ E }$-valued random variable. Moreover, when the space $\displaystyle{ E }$ is the real line $\displaystyle{ \mathbb{R} }$, then such a real-valued random variable is called simply a random variable.

If y \geq 0, then

### Real-valued random variables

$\displaystyle{ \operatorname{P}(X^2 \le y) = \operatorname{P}(|X| \le \sqrt{y}) (x ^ 2 le y) = operatorname { p }(| x | le sqrt { y }) In this case the observation space is the set of real numbers. Recall, \lt math\gt (\Omega, \mathcal{F}, P) }$ is the probability space. For a real observation space, the function $\displaystyle{ X\colon \Omega \rightarrow \mathbb{R} }$ is a real-valued random variable if

= \operatorname{P}(-\sqrt{y} \le  X \le \sqrt{y}),[/itex]


= { p }(- sqrt { y } le x le sqrt { y }) ，</math >

$\displaystyle{ \{ \omega : X(\omega) \le r \} \in \mathcal{F} \qquad \forall r \in \mathbb{R}. }$

so

This definition is a special case of the above because the set $\displaystyle{ \{(-\infty, r]: r \in \R\} }$ generates the Borel σ-algebra on the set of real numbers, and it suffices to check measurability on any generating set. Here we can prove measurability on this generating set by using the fact that $\displaystyle{ \{ \omega : X(\omega) \le r \} = X^{-1}((-\infty, r]) }$.

F _ y (y) = f _ x (sqrt { y })-f _ x (- sqrt { y }) qquad hbox { if } quad y ge 0.

## Moments

The probability distribution of a random variable is often characterised by a small number of parameters, which also have a practical interpretation. For example, it is often enough to know what its "average value" is. This is captured by the mathematical concept of expected value of a random variable, denoted $\displaystyle{ \operatorname{E}[X] }$, and also called the first moment. In general, $\displaystyle{ \operatorname{E}[f(X)] }$ is not equal to $\displaystyle{ f(\operatorname{E}[X]) }$. Once the "average value" is known, one could then ask how far from this average value the values of $\displaystyle{ X }$ typically are, a question that is answered by the variance and standard deviation of a random variable. $\displaystyle{ \operatorname{E}[X] }$ can be viewed intuitively as an average obtained from an infinite population, the members of which are particular evaluations of $\displaystyle{ X }$.

Suppose X is a random variable with a cumulative distribution

Mathematically, this is known as the (generalised) problem of moments: for a given class of random variables $\displaystyle{ X }$, find a collection $\displaystyle{ \{f_i\} }$ of functions such that the expectation values $\displaystyle{ \operatorname{E}[f_i(X)] }$ fully characterise the distribution of the random variable $\displaystyle{ X }$.

F_{X}(x) = P(X \leq x) = \frac{1}{(1 + e^{-x})^{\theta}}


F _ { x }(x) = p (x leq x) = frac {1}{(1 + e ^ {-x }) ^ { theta }}

Moments can only be defined for real-valued functions of random variables (or complex-valued, etc.). If the random variable is itself real-valued, then moments of the variable itself can be taken, which are equivalent to moments of the identity function $\displaystyle{ f(X)=X }$ of the random variable. However, even for non-real-valued random variables, moments can be taken of real-valued functions of those variables. For example, for a categorical random variable X that can take on the nominal values "red", "blue" or "green", the real-valued function $\displaystyle{ [X = \text{green}] }$ can be constructed; this uses the Iverson bracket, and has the value 1 if $\displaystyle{ X }$ has the value "green", 0 otherwise. Then, the expected value and other moments of this function can be determined.

where \theta > 0 is a fixed parameter. Consider the random variable Y = \mathrm{log}(1 + e^{-X}). Then,

## Functions of random variables

F_{Y}(y) = P(Y \leq y) = P(\mathrm{log}(1 + e^{-X}) \leq y) = P(X \geq -\mathrm{log}(e^{y} - 1)).\,


f _ { y }(y) = p (y leq y) = p (mathrm { log }(1 + e ^ {-x }) leq y) = p (x geq-mathrm { log }(e ^ { y }-1)) ,

A new random variable Y can be defined by applying a real Borel measurable function $\displaystyle{ g\colon \mathbb{R} \rightarrow \mathbb{R} }$ to the outcomes of a real-valued random variable $\displaystyle{ X }$. That is, $\displaystyle{ Y=g(X) }$. The cumulative distribution function of $\displaystyle{ Y }$ is then

The last expression can be calculated in terms of the cumulative distribution of X, so

$\displaystyle{ F_Y(y) = \operatorname{P}(g(X) \le y). }$

$\displaystyle{ If function \lt math\gt g }$ is invertible (i.e., $\displaystyle{ h = g^{-1} }$ exists, where $\displaystyle{ h }$ is $\displaystyle{ g }$'s inverse function) and is either increasing or decreasing, then the previous relation can be extended to obtain

\begin{align} { align }

F_Y(y) & = 1 - F_X(-\log(e^y - 1)) \\[5pt]

F _ y (y) & = 1-f _ x (- log (e ^ y-1))[5 pt ]

\displaystyle{ F_Y(y) = \operatorname{P}(g(X) \le y) = & = 1 - \frac{1}{(1 + e^{\log(e^y - 1)})^\theta} \\[5pt] 1-frac {1}{(1 + e ^ { log (e ^ y-1)}) ^ theta }[5 pt ] \begin{cases} & = 1 - \frac{1}{(1 + e^y - 1)^\theta} \\[5pt] 1-frac {1}{(1 + e ^ y-1) ^ theta }[5 pt ] \operatorname{P}(X \le h(y)) = F_X(h(y)), & = 1 - e^{-y \theta}. & = 1-e ^ {-y theta }. & \text{if } h = g^{-1} \text{ increasing} ,\\ \end{align} { align } \\ }

\operatorname{P}(X \ge h(y)) = 1 - F_X(h(y)),

& \text{if } h = g^{-1} \text{ decreasing} .


which is the cumulative distribution function (CDF) of an exponential distribution.

\end{cases}[/itex]

With the same hypotheses of invertibility of $\displaystyle{ g }$, assuming also differentiability, the relation between the probability density functions can be found by differentiating both sides of the above expression with respect to $\displaystyle{ y }$, in order to obtain[8]

Suppose X is a random variable with a standard normal distribution, whose density is

$\displaystyle{ f_Y(y) = f_X\bigl(h(y)\bigr) \left| \frac{d h(y)}{d y} \right|. }$

f_X(x) = \frac{1}{\sqrt{2\pi}}e^{-x^2/2}.


F _ x (x) = frac {1}{ sqrt {2 pi } e ^ {-x ^ 2/2}.

If there is no invertibility of $\displaystyle{ g }$ but each $\displaystyle{ y }$ admits at most a countable number of roots (i.e., a finite, or countably infinite, number of $\displaystyle{ x_i }$ such that $\displaystyle{ y = g(x_i) }$) then the previous relation between the probability density functions can be generalized with

Consider the random variable Y = X^2. We can find the density using the above formula for a change of variables:

$\displaystyle{ f_Y(y) = \sum_{i} f_X(g_{i}^{-1}(y)) \left| \frac{d g_{i}^{-1}(y)}{d y} \right| }$

f_Y(y) = \sum_{i} f_X(g_{i}^{-1}(y)) \left| \frac{d g_{i}^{-1}(y)}{d y} \right|.

F _ y (y) = sum _ { i } f _ x (g _ { i } ^ {-1}(y)))左 | frac { g _ { i } ^ {-1}(y)}{ d y }右 | 。

where $\displaystyle{ x_i = g_i^{-1}(y) }$, according to the inverse function theorem. The formulas for densities do not demand $\displaystyle{ g }$ to be increasing.

In this case the change is not monotonic, because every value of Y has two corresponding values of X (one positive and negative). However, because of symmetry, both halves will transform identically, i.e.,

In the measure-theoretic, axiomatic approach to probability, if a random variable $\displaystyle{ X }$ on $\displaystyle{ \Omega }$ and a Borel measurable function $\displaystyle{ g\colon \mathbb{R} \rightarrow \mathbb{R} }$, then $\displaystyle{ Y = g(X) }$ is also a random variable on $\displaystyle{ \Omega }$, since the composition of measurable functions is also measurable. (However, this is not necessarily true if $\displaystyle{ g }$ is Lebesgue measurable.[citation needed]) The same procedure that allowed one to go from a probability space $\displaystyle{ (\Omega, P) }$ to $\displaystyle{ (\mathbb{R}, dF_{X}) }$ can be used to obtain the distribution of $\displaystyle{ Y }$.

f_Y(y) = 2f_X(g^{-1}(y)) \left| \frac{d g^{-1}(y)}{d y} \right|.

f _ y (y) = 2f _ x (g ^ {-1}(y))左 | frac { d ^ {-1}(y)}{ d y }右 | 。

### Example 1

The inverse transformation is

Let $\displaystyle{ X }$ be a real-valued, continuous random variable and let $\displaystyle{ Y = X^2 }$.

x = g^{-1}(y) = \sqrt{y}

X = g ^ {-1}(y) = sqrt { y }

and its derivative is

$\displaystyle{ F_Y(y) = \operatorname{P}(X^2 \le y). }$

\frac{d g^{-1}(y)}{d y} = \frac{1}{2\sqrt{y}} .

Frac { d ^ {-1}(y)}{ d y } = frac {1}{2 sqrt { y }}.

If $\displaystyle{ y \lt 0 }$, then $\displaystyle{ P(X^2 \leq y) = 0 }$, so

Then,

$\displaystyle{ F_Y(y) = 0\qquad\hbox{if}\quad y \lt 0. }$
f_Y(y) = 2\frac{1}{\sqrt{2\pi}}e^{-y/2} \frac{1}{2\sqrt{y}} = \frac{1}{\sqrt{2\pi y}}e^{-y/2}.


f _ y (y) = 2 frac {1}{ sqrt {2 pi } e ^ {-y/2} frac {1}{2 sqrt { y }} = frac {1}{2 pi y } e ^ {-y/2}.

If $\displaystyle{ y \geq 0 }$, then

This is a chi-squared distribution with one degree of freedom.

$\displaystyle{ \operatorname{P}(X^2 \le y) = \operatorname{P}(|X| \le \sqrt{y}) = \operatorname{P}(-\sqrt{y} \le X \le \sqrt{y}), }$

Suppose X is a random variable with a normal distribution, whose density is

so

f_X(x) = \frac{1}{\sqrt{2\pi\sigma^2}}e^{-(x-\mu)^2/(2\sigma^2)}.


F _ x (x) = frac {1}{ sqrt {2 pi sigma ^ 2} e ^ {-(x-mu) ^ 2/(2 sigma ^ 2)}.

$\displaystyle{ F_Y(y) = F_X(\sqrt{y}) - F_X(-\sqrt{y})\qquad\hbox{if}\quad y \ge 0. }$

Consider the random variable Y = X^2. We can find the density using the above formula for a change of variables:

### Example 2

f_Y(y) = \sum_{i} f_X(g_{i}^{-1}(y)) \left| \frac{d g_{i}^{-1}(y)}{d y} \right|.

F _ y (y) = sum _ { i } f _ x (g _ { i } ^ {-1}(y)))左 | frac { g _ { i } ^ {-1}(y)}{ d y }右 | 。

Suppose $\displaystyle{ X }$ is a random variable with a cumulative distribution

In this case the change is not monotonic, because every value of Y has two corresponding values of X (one positive and negative). Differently from the previous example, in this case however, there is no symmetry and we have to compute the two distinct terms:

$\displaystyle{ F_{X}(x) = P(X \leq x) = \frac{1}{(1 + e^{-x})^{\theta}} }$

f_Y(y) = f_X(g_1^{-1}(y))\left|\frac{d g_1^{-1}(y)}{d y} \right| +f_X(g_2^{-1}(y))\left| \frac{d g_2^{-1}(y)}{d y} \right|.

F _ y (y) = f _ x (g _ 1 ^ {-1}(y))左 | frac { d g _ 1 ^ {-1}(y)}{ d y }右 | + f _ x (g _ 2 ^ {-1}(y))左 | frac { d g _ 2 ^ {-1}(y)}{ d y }右 | 。

where $\displaystyle{ \theta \gt 0 }$ is a fixed parameter. Consider the random variable $\displaystyle{ Y = \mathrm{log}(1 + e^{-X}). }$ Then,

The inverse transformation is

$\displaystyle{ F_{Y}(y) = P(Y \leq y) = P(\mathrm{log}(1 + e^{-X}) \leq y) = P(X \geq -\mathrm{log}(e^{y} - 1)).\, }$

x = g_{1,2}^{-1}(y) = \pm \sqrt{y}

X = g {1,2} ^ {-1}(y) = pm sqrt { y }

and its derivative is

The last expression can be calculated in terms of the cumulative distribution of $\displaystyle{ X, }$ so

\frac{d g_{1,2}^{-1}(y)}{d y} = \pm \frac{1}{2\sqrt{y}} .

1,2} ^ {-1}(y)}{ d y } = pm frac {1}{2 sqrt { y }}.

\displaystyle{ Then, 然后, \begin{align} F_Y(y) & = 1 - F_X(-\log(e^y - 1)) \\[5pt] f_Y(y) = \frac{1}{\sqrt{2\pi\sigma^2}} \frac{1}{2\sqrt{y}} (e^{-(\sqrt{y}-\mu)^2/(2\sigma^2)}+e^{-(-\sqrt{y}-\mu)^2/(2\sigma^2)}) . f _ y (y) = frac {1}{ sqrt {2 pi sigma ^ 2} frac {1}{2 sqrt { y }}(e ^ {-(sqrt { y }-mu) ^ 2/(2 sigma ^ 2)} + e ^ {-(- sqrt { y }-mu) ^ 2/(2 sigma ^ 2)}). & = 1 - \frac{1}{(1 + e^{\log(e^y - 1)})^\theta} \\[5pt] & = 1 - \frac{1}{(1 + e^y - 1)^\theta} \\[5pt] This is a noncentral chi-squared distribution with one degree of freedom. 这是一个单自由度的非中心卡方分布。 & = 1 - e^{-y \theta}. \end{align} }

which is the cumulative distribution function (CDF) of an exponential distribution.

### Example 3

There are several different senses in which random variables can be considered to be equivalent. Two random variables can be equal, equal almost surely, or equal in distribution.

Suppose $\displaystyle{ X }$ is a random variable with a standard normal distribution, whose density is

In increasing order of strength, the precise definition of these notions of equivalence is given below.

$\displaystyle{ f_X(x) = \frac{1}{\sqrt{2\pi}}e^{-x^2/2}. }$

Consider the random variable $\displaystyle{ Y = X^2. }$ We can find the density using the above formula for a change of variables:

If the sample space is a subset of the real line, random variables X and Y are equal in distribution (denoted X \stackrel{d}{=} Y) if they have the same distribution functions:

$\displaystyle{ f_Y(y) = \sum_{i} f_X(g_{i}^{-1}(y)) \left| \frac{d g_{i}^{-1}(y)}{d y} \right|. }$

\operatorname{P}(X \le x) = \operatorname{P}(Y \le x)\quad\text{for all }x.

{ p }(x le x) = { p }(y le x) quad text { for all } x。

In this case the change is not monotonic, because every value of $\displaystyle{ Y }$ has two corresponding values of $\displaystyle{ X }$ (one positive and negative). However, because of symmetry, both halves will transform identically, i.e.,

To be equal in distribution, random variables need not be defined on the same probability space. Two random variables having equal moment generating functions have the same distribution. This provides, for example, a useful method of checking equality of certain functions of independent, identically distributed (IID) random variables. However, the moment generating function exists only for distributions that have a defined Laplace transform.

$\displaystyle{ f_Y(y) = 2f_X(g^{-1}(y)) \left| \frac{d g^{-1}(y)}{d y} \right|. }$

The inverse transformation is

Two random variables X and Y are equal almost surely (denoted X \; \stackrel{\text{a.s.}}{=} \; Y) if, and only if, the probability that they are different is zero:

$\displaystyle{ x = g^{-1}(y) = \sqrt{y} }$

and its derivative is

\operatorname{P}(X \neq Y) = 0.

{ p }(x neq y) = 0。

$\displaystyle{ \frac{d g^{-1}(y)}{d y} = \frac{1}{2\sqrt{y}} . }$

For all practical purposes in probability theory, this notion of equivalence is as strong as actual equality. It is associated to the following distance:

Then,

d_\infty(X,Y)=\operatorname{ess} \sup_\omega|X(\omega)-Y(\omega)|,

D _ infty (x，y) = { ess } sup _ omega | x (omega)-y (omega) | ,

$\displaystyle{ f_Y(y) = 2\frac{1}{\sqrt{2\pi}}e^{-y/2} \frac{1}{2\sqrt{y}} = \frac{1}{\sqrt{2\pi y}}e^{-y/2}. }$

where "ess sup" represents the essential supremum in the sense of measure theory.

This is a chi-squared distribution with one degree of freedom.

### Example 4

Finally, the two random variables X and Y are equal if they are equal as functions on their measurable space:

Suppose $\displaystyle{ X }$ is a random variable with a normal distribution, whose density is

X (omega) = y (omega) qquad hbox { for all } omega.

$\displaystyle{ f_X(x) = \frac{1}{\sqrt{2\pi\sigma^2}}e^{-(x-\mu)^2/(2\sigma^2)}. }$

This notion is typically the least useful in probability theory because in practice and in theory, the underlying measure space of the experiment is rarely explicitly characterized or even characterizable.

Consider the random variable $\displaystyle{ Y = X^2. }$ We can find the density using the above formula for a change of variables:

$\displaystyle{ f_Y(y) = \sum_{i} f_X(g_{i}^{-1}(y)) \left| \frac{d g_{i}^{-1}(y)}{d y} \right|. }$

A significant theme in mathematical statistics consists of obtaining convergence results for certain sequences of random variables; for instance the law of large numbers and the central limit theorem.

In this case the change is not monotonic, because every value of $\displaystyle{ Y }$ has two corresponding values of $\displaystyle{ X }$ (one positive and negative). Differently from the previous example, in this case however, there is no symmetry and we have to compute the two distinct terms:

There are various senses in which a sequence X_n of random variables can converge to a random variable X. These are explained in the article on convergence of random variables.

$\displaystyle{ f_Y(y) = f_X(g_1^{-1}(y))\left|\frac{d g_1^{-1}(y)}{d y} \right| +f_X(g_2^{-1}(y))\left| \frac{d g_2^{-1}(y)}{d y} \right|. }$

The inverse transformation is

$\displaystyle{ x = g_{1,2}^{-1}(y) = \pm \sqrt{y} }$

and its derivative is

$\displaystyle{ \frac{d g_{1,2}^{-1}(y)}{d y} = \pm \frac{1}{2\sqrt{y}} . }$

Then,

$\displaystyle{ f_Y(y) = \frac{1}{\sqrt{2\pi\sigma^2}} \frac{1}{2\sqrt{y}} (e^{-(\sqrt{y}-\mu)^2/(2\sigma^2)}+e^{-(-\sqrt{y}-\mu)^2/(2\sigma^2)}) . }$

This is a noncentral chi-squared distribution with one degree of freedom.

## Some properties

• The probability distribution of the sum of two independent random variables is the convolution of each of their distributions.

## Equivalence of random variables

There are several different senses in which random variables can be considered to be equivalent. Two random variables can be equal, equal almost surely, or equal in distribution.

In increasing order of strength, the precise definition of these notions of equivalence is given below.

### Equality in distribution

If the sample space is a subset of the real line, random variables X and Y are equal in distribution (denoted $\displaystyle{ X \stackrel{d}{=} Y }$) if they have the same distribution functions:

$\displaystyle{ \operatorname{P}(X \le x) = \operatorname{P}(Y \le x)\quad\text{for all }x. }$

To be equal in distribution, random variables need not be defined on the same probability space. Two random variables having equal moment generating functions have the same distribution. This provides, for example, a useful method of checking equality of certain functions of independent, identically distributed (IID) random variables. However, the moment generating function exists only for distributions that have a defined Laplace transform.

### Almost sure equality

Two random variables X and Y are equal almost surely (denoted $\displaystyle{ X \; \stackrel{\text{a.s.}}{=} \; Y }$) if, and only if, the probability that they are different is zero:

$\displaystyle{ \operatorname{P}(X \neq Y) = 0. }$

For all practical purposes in probability theory, this notion of equivalence is as strong as actual equality. It is associated to the following distance:

$\displaystyle{ d_\infty(X,Y)=\operatorname{ess} \sup_\omega|X(\omega)-Y(\omega)|, }$

where "ess sup" represents the essential supremum in the sense of measure theory.

Category:Statistical randomness

This page was moved from wikipedia:en:Random variable. Its edit history can be viewed at 随机变量/edithistory

1. Blitzstein, Joe; Hwang, Jessica (2014). Introduction to Probability. CRC Press. ISBN 9781466575592.
2. Steigerwald, Douglas G. "Economics 245A – Introduction to Measure Theory" (PDF). University of California, Santa Barbara. Retrieved April 26, 2013.
3. "List of Probability and Statistics Symbols". Math Vault (in English). 2020-04-26. Retrieved 2020-08-21.
4. "Random Variables". www.mathsisfun.com. Retrieved 2020-08-21.
5. Yates, Daniel S.; Moore, David S; Starnes, Daren S. (2003). The Practice of Statistics (2nd ed.). New York: Freeman. ISBN 978-0-7167-4773-4.
6. "Random Variables". www.stat.yale.edu. Retrieved 2020-08-21.
7. L. Castañeda; V. Arunachalam; S. Dharmaraja (2012). Introduction to Probability and Stochastic Processes with Applications. Wiley. p. 67. ISBN 9781118344941.
8. Bertsekas, Dimitri P. (2002). Introduction to Probability. Tsitsiklis, John N., Τσιτσικλής, Γιάννης Ν.. Belmont, Mass.: Athena Scientific. ISBN 188652940X. OCLC 51441829.
9. 脚本错误：没有“Footnotes”这个模块。