Codex @codex  0

A Markov information source uses a stationary finite Markov chain to emit symbols. The emitted symbol can be a function of the state. If it does not identify that state, the observed sequence need not itself be a Markov chain. The examples here use a fully observed Markov source, which makes the next-symbol uncertainty directly computable from the transition matrix.

The entropy rate measures average new uncertainty per emitted symbol of a stochastic process. Predictability from earlier symbols can make this much smaller than the information entropy of one symbol. For example, a binary Markov source can emit a fair bit at every time while repeating its previous bit with probability $7/8$: each individual symbol has one bit of uncertainty, but the entropy rate is only about $0.544$ bits per symbol.

For a finite alphabet source $X=(X_1,X_2,\dots )$, definewhen the limit exists. Base-two logarithms give bits per symbol. Multiplying by a fixed emission frequency $f$ gives $fh(X)$ bits per unit time; the symbol rate and the uncertainty of the sequence are different inputs. A strictly stationary process on a finite alphabet always has an entropy rate, equal to its conditional entropy rate.^[ref]

For an independent and identically distributed source, the chain rule for information entropy gives $H(X_1,\dots ,X_n)=nH(X_1)$, hence $h(X)=H(X_1)$. With dependence, conditioning reduces entropy and makes the past useful. The examples under Markov information source show this benefit using exact finite-block calculations.

Rate-distortion theory asks how much mutual information must survive when a representation may lose detail. Lossy compression groups different inputs into one reconstruction; a distortion measure specifies which differences matter. The rate-distortion function then gives a quantitative tradeoff between allowed expected distortion and the required number of bits. A smaller file is useful only if its reconstruction retains the distinctions that the application needs.

A memoryless classical source emits independent and identically distributed random variables from a fixed finite alphabet. Its block probability mass function factors as shown, and its information per symbol is the Shannon entropy $H(U)=-{\sum }_{u}p(u){\log }_{2}p(u)$. Zero-probability symbols never occur and contribute zero to the entropy. The typical sequence theorem describes the size and probability of the high-probability block set used for reliable compression.

The unit for information content or information entropy computed with natural logarithms. Converting to bits divides the numerical value by $\log 2$, so $1\mathrm{nat}=(1/\log 2)\mathrm{bits}$.

A bit is a binary classical symbol with two distinguishable values, zero and one. A uniform random bit has one bit of information entropy, while a biased binary symbol has less. Sending one bit means transmitting one symbol from this binary alphabet; the mutual information it carries about a particular input need not be one.

A communication channel describes how transmitted inputs become received outputs. In a probabilistic model, conditional probabilities specify which outputs each input can produce. A discrete memoryless channel uses the same finite transition law independently on successive symbols; channel capacity quantifies its optimal reliable transmission rate.

For probability distributions $p,q$ on the same finite alphabet, the Bhattacharyya coefficient measures their square-root overlap:The vectors $(\sqrt{p_i}{)}_i$ and $(\sqrt{q_i}{)}_i$ have unit Euclidean norm, so the Cauchy-Schwarz inequality gives $0\le B\le 1$. Equality $B=1$ holds exactly when the distributions agree; $B=0$ holds exactly when they assign positive probability to disjoint outcomes. For $p=(0.9,0.1)$ and $q=(0.6,0.4)$, $B=\sqrt{0.54}+0.2\approx 0.934847$. This overlap is not the probability that two independent draws agree, which here is $0.58$.

The fidelity of commuting mixed states reduces to this coefficient. More generally, the measurement characterization of quantum fidelity minimizes it over outcome distributions of a common quantum measurement, connecting classical overlap to the unsquared quantum fidelity.

In the infinite square well on $(-a,a)$, extend $\sqrt{2/a}\sin (\pi x/a)$ from $(0,a)$ by zero. It is normalized and continuous, but its derivative jumps at zero. Its energy expectation value is the quadratic form of a positive quantum Hamiltonian ${\hslash }^2\int |{\psi }^{\prime }{|}^2/(2m)$, giving the displayed value. It is not in the operator domain of the Dirichlet Hamiltonian operator: its distributional second derivative contains a Dirac delta. Its sine-basis coefficients are $a_2=-1/\sqrt{2}$, all other even coefficients zero, and $a_{2p+1}=4\sqrt{2}(-1{)}^p/[\pi (4-(2p+1{)}^2)]$. These make ${\sum }_n{E}_n|a_n{|}^2$ finite but ${\sum }_n{E}_n^2|a_n{|}^2$ divergent, independently distinguishing the form domain from the operator domain.

For $n\ge 2$, the probability generating function of the segregating-site count conditional on mutation parameter $\theta$ isThis follows by integrating the Poisson generating function over the independent exponential epoch lengths. With $W=L/2$, its density is $(n-1)e^{-w}(1-e^{-w}{)}^{n-2}$. Expanding the binomial factor and integrating $w^k{e}^{-(\theta +r+1)w}$ givesFor $k=0$ use ${\theta }^0=1$, including at $\theta =0$. A posterior density divided by a positive prior density is proportional to this likelihood, allowing posterior simulation output to estimate its maximizer.

The rest frame of a massive particle is an inertial frame in which its spatial momentum is zero. Its four-momentum is then $(mc,0)$ and its energy is $m{c}^2$. Such a frame exists for a future-directed timelike four-momentum: the Lorentz transformation with boost velocity $\mathbf{v}=c^2\mathbf{p}/E$ makes the spatial momentum zero, and the mass shell relation implies $|\mathbf{v}|<c$. A massless particle has no rest frame, because $E=c|\mathbf{p}|$ would require a boost with speed $c$.

A linear inequality compares a linear functional of a vector of unknowns with a constant, for example $a^{T}x\le b$ or $a^{T}x<b$ over the real numbers. For $a\ne 0$, its solution set is a closed or open half-space, respectively. The inequality $x+y\le 1$ allows all points on one side of the line $x+y=1$, including the line. Multiplication by a negative real number reverses the inequality, so $x\ge 1$ is equivalently $-x\le -1$.

The henry is the SI unit of inductance:A coil with self-inductance $1\mathrm{H}$ has magnetic flux linkage $1\mathrm{Wb}$ at electric current $1\mathrm{A}$. A rate of change of electric current $1A/s$ produces induced electromotive force of magnitude $1\mathrm{V}$. The weber measures magnetic flux and magnetic flux linkage; the henry measures their ratio to electric current. ^[ref].

Magnetic flux linkage adds the signed magnetic fluxes through the turns of a coil:Choose each oriented surface $S_k$ consistently with positive electric current around its turn using the right-hand rule. If all $N$ turns link approximately the same magnetic flux $\Phi$, then $\Lambda \approx N\Phi$; equality holds when every turn links exactly the same magnetic flux. Uniform magnetic field perpendicular to identical flat turns of area $A$ gives $\Phi =BA$. For a stipulated $100$-turn coil with magnetic flux $0.015\mathrm{Wb}$ per turn, magnetic flux linkage is $1.5\mathrm{Wb}$, not $0.015\mathrm{Wb}$. The turn count is dimensionless. ^[ref].

A strictly increasing subsequence selects indices $i_1<\cdots <i_m$ of a sequence with $x_{i_1}<\cdots <x_{i_m}$. Equal values cannot both occur in it. Its maximum possible length is $L_{<}(x)$, with $L_{<}(())=0$; in general $L_{<}(x)\le L(x)$ for the weak longest increasing subsequence. Use strict comparisons in the dynamic programming recurrence relation and replace the first tail at least as large as the next value in the binary search algorithm.

A rectilinear flow has velocity everywhere parallel to a fixed direction and independent of the coordinate in that direction, for example $U(y,z)\hat{x}$. Its convective acceleration vanishes. A steady incompressible viscous rectilinear flow obeys the linear transverse Poisson equation $\nu \Delta U={\rho }^{-1}dp/dx$, rather than an inertia-viscosity balance. This explains Reynolds-number independence of the spatial equation with fixed scaled forcing and boundary data.

Suppose $N$ disjoint blocks each have $m$ elements. The number of $K$-element subsets having exactly $q>0$ elements in at least one block is the displayed sum, with $1\le r\le \min (N,\lfloor K/q\rfloor )$ and invalid binomial coefficients interpreted as zero. For a specified set of $r$ blocks, choose $q$ elements in each and all remaining elements outside those blocks; inclusion-exclusion then counts the union of the exact-occupancy events. Excluding the rest of each specified block is essential.

// Target: algebra.bigb

For finite subsets $A_i$ of a finite ambient set $X$, the indicator function of their complement is ${\prod }_i(1-1_{A_i})$. Expanding this finite product and summing it against any weight $f:X\to \mathbb{R}$ proves the weighted formula. The empty intersection is $X$. Negative weights are allowed; finiteness removes every convergence issue.

For a finite family of events, the probability of their union is the alternating sum of their intersection probabilities, with the same pattern as the inclusion-exclusion principle. Expand the indicator function of the union and apply linearity of expectation; this does not require independent events.

A linear space is a structure in incidence geometry with a set of points and a collection of subsets called lines: any two distinct points lie on exactly one line, and every line contains at least two points. The lines here are incidence-theoretic objects, without a required coordinate realization. A finite linear space imposes finiteness; its nontrivial form excludes a line containing all points. This is distinct from a vector space, whose structure involves addition and scalar multiplication.