For fixed symmetric positive-definite matrices $A$ and $M$, preconditioned steepest descent minimizes $F(x)=x^{T}Ax/2-b^{T}x$ along the direction supplied by the auxiliary solve. Given $x_k$, form the linear-system residual $r_k=b-A{x}_k$. If $r_k=0$, stop: $x_k=x_{\ast }$. Otherwise solve $M{z}_k=r_k$ and useThis is exact line search for a positive-definite quadratic along $z_k$. Indeed,whose derivative vanishes at the displayed positive step size. Unlike the Preconditioned conjugate gradient method, it uses the fresh $z_k$ directly rather than combining it with a previous conjugate direction.

For diagonal scaling for a two-dimensional linear system, start at $x_0=0$. Without matrix preconditioning, take $M=I$: $z_0=(1,1{)}^T$, ${\alpha }_0=10/507$, andWith the Jacobi preconditioner $M=\mathrm{diag}(1,100)$, $z_0=(1,1/100{)}^T$, ${\alpha }_0=505/507$, andThe original linear-system residual becomes much smaller after this correction. Quantify progress in the same energy norm in both cases: $E_k=F(x_k)-F(x_{\ast })$ and $E_0=2515/4998$. Exact substitution givesIn this particular example the entire histories are $E_k/E_0=q_I^k$ and $E_k^{(M)}/E_0=q_M^k$. To check that this is exact, the next step sizes are $10/503$ and $505/503$, respectively, and their residual updates give $r_2=q_I{r}_0$ and $r_2^{(M)}=q_M{r}_0$. Multiplying a linear-system residual by a nonzero scalar leaves the next step size unchanged; every pair of steps repeats the same directions at a smaller scale. The two first-step energy ratios then establish these formulas for all $k$.

The curves come from exact rational updates of the specified linear system, converted to decimal values for plotting. The left panel counts iterations. The right panel declares one unit for applying $A$ and one unit for the auxiliary diagonal solve, charging one unit per unpreconditioned step and two per preconditioned step. It excludes setup and vector operations, so it compares a declared cost model rather than measured runtime. Scaling is advantageous even with that extra charge.

These curves compare steepest descent method iterations. On this two-dimensional linear system, both the conjugate gradient method and the Preconditioned conjugate gradient method terminate within two steps in exact arithmetic. The plot therefore establishes no CG iteration-count advantage for this example; the benefit of scaling for larger problems depends on their eigenvalues and auxiliary-solve costs.

Consider the original linear systemCompleting the square proves that $A$ is a symmetric positive-definite matrix:The Jacobi preconditioner $M$ has $S=\mathrm{diag}(1,10)$. Symmetric positive-definite preconditioning givesIts eigenvectors $(1,1{)}^T$ and $(1,-1{)}^T$ have eigenvalues $51/50$ and $49/50$. HenceThe characteristic polynomial of $A$ is ${\lambda }^2-101\lambda +2499/25$. Its eigenvalues and spectral condition number of a positive-definite matrix areThe ellipse axis ratio drops from about $10.0020$ to $1.0202$. The transformation multiplies the second coordinate by ten, bringing the two energy scales together.

Both panels show the same energy level sets, expressed in error coordinates $e=x-x_{\ast }$ and $\eta =M^{1/2}e$. Their coordinate axes have equal scales. The marked displacement $(1,0.1{)}^T$ becomes $(1,1{)}^T$; its energy remains $e^{T}Ae/2=1.02$. These are exact contours of the specified quadratic form.

Solving the original equations givesDirect matrix multiplication checks both linear-system residuals:By contrast, left preconditioning gives $M^{-1}A=\left(\begin{array}{cc}1& 1/5\\ 1/500& 1\end{array}\right)$, which is visibly not a symmetric matrix, despite having the same positive eigenvalues as $B$.

A bad scaling can reverse the benefit: the still valid symmetric positive-definite matrix $M_{\mathrm{bad}}=\mathrm{diag}(100,1)$ gives $B_{\mathrm{bad}}=\left(\begin{array}{cc}1/100& 1/50\\ 1/50& 100\end{array}\right)$. Its Rayleigh quotients on the coordinate unit vectors are $0.01$ and $100$, so ${\kappa }_2(B_{\mathrm{bad}})\ge 10000$. Cheapness and positive definiteness do not guarantee improved conditioning.

Let $A$ and $M$ be real symmetric positive-definite matrices. The principal square root of a positive semidefinite matrix construction gives an invertible matrix for each. SetThen $Ax=b$ is equivalent to $By=c$. This is a matrix congruence, and $B$ is a symmetric positive-definite matrix: $B^T=B$, and for every nonzero vector $v$,since $S^{-1}v\ne 0$. This short proof establishes both requirements for applying the conjugate gradient method to the transformed linear system.

The left transformed matrix has the same eigenvalues becauseIts correct geometry uses the weighted inner product $⟨u,v{⟩}_M=u^{T}Mv$: $⟨u,M^{-1}Av{⟩}_M=u^{T}Av=⟨M^{-1}Au,v{⟩}_M$. Thus it is symmetric in this inner product, even when it is not a symmetric matrix in ordinary coordinates. The spectral condition number governing the transformed symmetric problem is ${\kappa }_2(B)$, not generally ${\kappa }_2(M^{-1}A)$.

The Preconditioned conjugate gradient method realizes this transformation using auxiliary solves $M{z}_k=r_k$, without constructing $S$ or its matrix inverse. With fixed $M$, initialize $r_0=b-A{x}_0$ and $p_0=z_0$; while $r_k\ne 0$, usethen solve $M{z}_{k+1}=r_{k+1}$ and setStop at a zero linear-system residual before forming another ratio. Positive definiteness makes the denominators positive before termination.^[ref]

An invertible matrix alone is insufficient: with $A=I$ and $M=\mathrm{diag}(-1,1)$, the left transformed matrix has eigenvalues $-1,1$, and $M$ has no real symmetric positive-definite matrix square root. Even a valid symmetric positive-definite matrix can worsen conditioning if its scaling is unsuitable, as the next worked example shows.

Right preconditioning substitutes $x=M^{-1}u$ into the linear system $Ax=b$. Applying its transformed linear map first solves $Mz=v$, then computes $Az$. Recover the solution by solving $Mx=u$. Its linear-system residual $b-A{M}^{-1}u$ is exactly the original linear-system residual $b-Ax$, which is useful when monitoring the imbalance in the original equations.

The left preconditioning and right preconditioning transformed matrices are similar matrices:They therefore have the same eigenvalues, although their Euclidean norm condition numbers and residual histories need not agree. A similarity transformation preserves eigenvalues; a nonorthogonal change of coordinates need not preserve lengths.

Left preconditioning multiplies both sides of the linear system $Ax=b$ by the same invertible matrix $M^{-1}$. It leaves the unknown vector $x$ unchanged. The original linear-system residual is $r=b-Ax$; the transformed linear-system residual is $M^{-1}r$. Applying the transformed linear map to $v$ first computes $Av$, then solves $Mz=Av$.

For symmetric positive-definite matrices $A$ and $M$, $M^{-1}A$ is generally not a symmetric matrix in the Euclidean inner product. Its positive eigenvalues alone do not justify ordinary Euclidean conjugate gradient method recurrences. The remedy is symmetric positive-definite preconditioning, implemented through the Preconditioned conjugate gradient method.

For an invertible matrix $P$, induction gives $(PA{P}^{-1}{)}^k=P{A}^k{P}^{-1}$, including $k=0$. The power series for the matrix exponential is absolutely convergent in every matrix norm, so multiplication by the fixed matrices $P,P^{-1}$ can be taken through the sum. This proves the displayed identity without any diagonalizability assumption. In particular, if $A=PD{P}^{-1}$ with $D$ diagonal, then $e^A=P\mathrm{diag}(e^{D_{11}},\dots ,e^{D_{nn}})P^{-1}$.

If $A$ is Skew-Hermitian, then $-iA$ is a Hermitian matrix. The finite-dimensional spectral theorem gives $A=Q\mathrm{diag}(i{\theta }_j)Q^{†}$, with real ${\theta }_j$ and a unitary matrix $Q$. Since the matrix exponential respects similarity, $e^A=Q\mathrm{diag}(e^{i{\theta }_j})Q^{†}$. Each diagonal entry has modulus one, so $e^A$ is a unitary matrix. Thus a skew-Hermitian matrix generates norm-preserving linear evolution.

A Hermitian matrix has a unitary spectral decomposition $A=Q\mathrm{diag}({\lambda }_j)Q^{†}$ with real eigenvalues ${\lambda }_j$. The matrix exponential respects similarity, giving $e^A=Q\mathrm{diag}(e^{{\lambda }_j})Q^{†}$. Its eigenvalues are positive and its conjugate transpose equals itself. In particular $v^{†}{e}^{A}v={\sum }_j{e}^{{\lambda }_j}|(Q^{†}v{)}_j{|}^2>0$ for $v\ne 0$, so it is a Hermitian positive-definite matrix.

If every eigenvalue of a finite square matrix $A$ has negative real part, its matrix exponential decays and its integral converges. The Jordan normal form shows that polynomial factors in $t$ are dominated by exponential decay. Integrating $d(e^{tA})/dt=A{e}^{tA}$ gives $A{\int }_0^{\infty }{e}^{tA}dt=-I$, proving the displayed matrix inverse formula. In particular, for a Hermitian matrix with negative eigenvalues, the spectral theorem reduces the integral to scalar exponentials with integrals $-1/{\lambda }_j$. The minus sign is essential: the integral is positive definite in this Hermitian case, while $A^{-1}$ has negative eigenvalues.

The following original model isolates amplification by inverted electric dipoles. Let $N=N_u-N_l$ be the excess number in the upper member of one equally degenerate resonant pair. Let $a$ be a slowly varying cavity amplitude with photon number $n=|a{|}^2$, and let $p$ describe the collective molecular dipole component in the amplifying phase convention. On exact resonance, stipulate the linear equationsThe dipole relaxation rate is $\Gamma$, and $g_0=|d_{ul}{\mathcal{E}}_{\mathrm{vac}}|/\hslash$ is a single-molecule coupling in the chosen mode normalization. Here $d_{ul}$ is an electric dipole moment matrix element, ${\mathcal{E}}_{\mathrm{vac}}$ is the field's zero-point amplitude, and $\hslash =h/(2\pi )$. Field overlap and dipole orientation enter this coupling. The characteristic polynomial for growth rate $\lambda$ isFor $N>0$, one root becomes positive precisely when $g_0^{2}N>\kappa \Gamma /2$. If dipole relaxation is fast enough to eliminate $p$, then $p\approx g_{0}Na/\Gamma$ andThus positive population inversion alone is insufficient: the gain rate $GN$ must exceed the passive decay rate. This reduction describes the onset near resonance; it does not predict output phase noise, actual transit dynamics or an exact historical threshold. Nikolay Basov and Aleksandr Prokhorov's molecular-generator theory likewise treats self-excitation as competition between molecular gain and resonator-plus-load losses.^[ref]

To include replenishment, suppose upper-state molecules enter at rate $F$ and both levels leave with residence rate $\gamma$. A deliberately simple ordinary differential equation model isOne downward transition removes two units of the difference $N_u-N_l$, so $\dot{N}=F-\gamma N-2GNn$. Below onset the unsaturated inversion is $N_0=F/\gamma$, givingFor fixed mode volume $V$, the equivalent number density condition is $N_0/V>\kappa /(GV)$. Raising selected beam intensity raises gain; increasing quality factor lowers the loss rate at fixed frequency. These statements hold with $G$, residence time and mode overlap fixed.

Above onset, the nonzero steady state hasThe inversion is clamped while excess supply supports output. A finite initial seed is understood: the noiseless equations leave $n=0$ exactly unchanged even above onset. Actual startup has spontaneous emission and field fluctuations.

In the saturated isotropic magnetostrictive strain convention, saturation magnetostriction is the longitudinal relative length change from a random-domain reference to alignment. A parallel-to-perpendicular length difference instead equals $3{\lambda }_s/2$ in that model.

Volume magnetostriction is the relative volume change associated with a magnetic-state change. To first order it is the trace of the magnetostrictive strain. Zero trace is an additional restriction, not a consequence of isotropy alone.

In 1949, Clifford Shull and J. Samuel Smart compared powdered manganese(II) oxide at room temperature and $80\mathrm{K}$. Powder preparation supplies many crystal orientations. Their Fig. 1 records scattered-neutron intensity versus counter angle. Nuclear reflections remained, while the cold pattern gained strong reflections forbidden by the chemical unit cell. They indexed these with a magnetic unit cell having twice the chemical cell edge. The warm pattern already contained diffuse magnetic scattering: loss of sharp magnetic peaks does not mean loss of individual magnetic moments.^[ref]

The source was Oak Ridge's graphite-moderated reactor. In Early Development of Neutron Scattering (1994), Clifford Shull describes how the early two-axis instruments used a monochromating crystal to select the incident-neutron wavelength; a shielded neutron counter rotated around the specimen to record angular intensity. This specifies the source–monochromator–sample–counter path, without identifying which instrument revision produced the MnO trace.^[ref]

The control was the same compound above and below magnetic ordering, with retained nuclear peaks identifying the underlying chemical periodicity. The original paper labels an aluminium impurity peak and defers a complete magnetic-structure account. It supplies neither a chemical synthesis recipe nor detailed cold-stage construction. The inference is periodic magnetic order, rather than a claim that every possible structural change was excluded at arbitrary resolution.^[ref]

The connection to the measured angle is the elastic Bragg scattering condition. If the incident and outgoing wave vectors have equal magnitude $2\pi /\lambda$ and make an angle $2\theta$, their difference has magnitudeA longer real-space period supplies additional, smaller reciprocal lattice spacings, which can put magnetic reflections at new counter angles. In the one-dimensional toy model above, doubling the period changes the allowed spacing from $2\pi /a$ to $\pi /a$. The actual MnO indexing is three-dimensional; the toy row does not assert its reflection positions or relative peak strengths.

Louis Néel supplied the ordering theory; the 1949 diffraction work was a later experimental test. His 1970 Nobel Prize in Physics recognized antiferromagnetism and ferrimagnetism. Clifford Shull's neutron-diffraction award followed in 1994.^[ref]

A uniformly charged solid sphere of total charge $Q$, radius $R$ and angular velocity $\mathbf{\omega }$ has magnetic dipole momentFor a uniformly charged thin spherical shell the coefficient is $1/3$ instead of $1/5$. At large distances the magnetic field is a magnetic dipole field, $\mathbf{B}={\mu }_0[3(\mathbf{m}\cdot \hat{\mathbf{r}})\hat{\mathbf{r}}-\mathbf{m}]/(4\pi r^3)$. Its angularly averaged magnitude and root mean square magnitude both scale as $r^{-3}$; its angularly averaged vector is zero. Steady rotation of this axisymmetric charge distribution produces magnetostatics, not radiation.

For a centred magnetic dipole moment pointing toward geographic south, the field at northern latitude $\lambda$ has northward component $B_N=B_{\mathrm{eq}}\cos \lambda$ and downward component $B_D=2B_{\mathrm{eq}}\sin \lambda$. These follow by resolving the magnetic dipole field into local radial and meridional directions. The downward inclination is therefore $\tan \iota =B_D/B_N=2\tan \lambda$, while the magnitude is $B_{\mathrm{eq}}\sqrt{1+3{\sin }^2\lambda }$. This ideal aligned-dipole model has no eastward component and does not capture actual geographic departures of the geomagnetic field.

For coaxial circular loops with radii $a_1,a_2\ll d$, the first loop's on-axis magnetic field of a circular current loop has far value $B_0\simeq {\mu }_0{I}_1{a}_1^2/(2d^3)$. Its near-axis radial magnetic field at the second loop is $B_{\rho }\simeq 3{\mu }_0{I}_1{a}_1^2{a}_2/(4d^4)$. Integrating the axial Lorentz force from $I_2{a}_{2}d\varphi \hat{\mathbf{\varphi }}\times B_{\rho }\hat{\mathbf{\rho }}$ gives the displayed magnitude. Equal current senses attract; opposite senses repel. This is the leading dipole limit and is not an exact close-coil force law.

An autophagosome is the membrane-enclosed carrier formed around macroautophagy cargo. Enclosing a mitochondrion does not immediately digest it: subsequent delivery must expose the cargo to digestive enzymes in lysosomes.^[ref]

In a local London equation description, the measured London penetration depth satisfies ${\lambda }_L^2=m/({\mu }_0{n}_s{e}^2)$ when $n_s$ counts superconducting electrons and $m$ is their effective mass. Thus screening measures $n_s/m$. If instead one counts Cooper pairs, use pair mass $2m$ and pair electric charge $-2e$, giving pair density $n_s/2$ for the same screening length. Identifying this superconducting carrier density with the total conduction-electron density requires the additional assumption that the relevant carriers all participate in screening at the measurement temperature.

For a real location parameter $\mu$ and positive scale parameter $s$, the logistic distribution has probability density function and cumulative distribution functionDifferentiating $F$ recovers $f$, and its limits at infinity show normalization of a probability distribution. The density is symmetric about $\mu$ with exponentially decaying tails, so its expected value exists and equals $\mu$. Allowing $s<0$ in the displayed expression makes $f$ negative and its integral $-1$; a signed scale must instead be replaced by $|s|$ in the prefactor to define the same valid density.

The University of California, Berkeley guide LLS - Measuring the Light Signal from a Diode, in Experimental Setup and Experiments X–XI, documents an experiment using a red light-emitting diode amid room-light background. An optical chopper, controlled by an SR540, periodically interrupts its light. A PIN-10DP photodiode sends electric current to an SR570 preamplifier, which converts it to an electric potential difference for an SR760 spectrum analyzer and an SR830 lock-in amplifier. The SR540 supplies the external reference. Preparation includes aligning the photodiode and its shielding cone and choosing gain to avoid overload. One comparison first searches for the source frequency without the chopper reference, then checks against that external reference; switching off the LED tests the candidate spectral peak. The measurement procedure varies chopper frequency and preamplifier filtering, then records repeated lock-in readings with selected averaging settings. These are prescribed observations to make, not published numerical results.^[ref]

The guide notes nearly triangular chopping because slit and emitter sizes are comparable. Its interpretation requires detector conversion, preamplifier gain and waveform geometry to infer optical power from the selected component.^[ref]

The inference is therefore specific: synchronized electric potential difference supports the presence of a modulated optical component at the reference frequency. It does not by itself give total emitted optical power. The photodiode's spectral responsivity, collected fraction of the light, and conversion from the measured Fourier coefficient to the chopped waveform must enter that calibration. The finite-window lock-in response also explains why a frequency-search reading can drift when the reference is not synchronized. An LED-off comparison tests background at the same timing; it cannot guarantee that every synchronized disturbance disappears when the desired source is present. Overload before multiplication destroys information that later averaging cannot restore.