The conjugate transpose also reverses factor order. If and are unitary matrices,
Thus their product is unitary. This is closure under multiplication in the unitary group.
The transpose of a matrix product reverses factor order:
For the requested two-dimensional check, put and . Both sides are
More generally, the entry formula is exactly .
Here , so obeys . An integrating factor is , giving
The nonzero solutions are therefore
The additional solution is . Each formula is interpreted on an interval where is defined and its displayed denominator is nonzero.
For the Bernoulli differential equation, work on a branch where and the required powers are defined. Multiplication by makes the chain rule useful:
Thus , and . The case is already linear and must be treated separately, because the proposed substitution would be constant. Solutions lost by dividing by a power of , particularly when it solves the original equation, must also be checked separately.
The second gradient is . Combining its vanishing with gives
The zero vector always works. A nonzero common stationary vector exists exactly when have a common eigenvector with respective eigenvalues ; both stationary values are then zero. Commuting symmetric matrices have a common orthonormal eigenbasis, so commutation is sufficient to find such vectors for the paired eigenvalues in that basis. It does not guarantee an arbitrary prescribed pair.
Commutation is also not necessary for one common eigenvector. For example,
are real symmetric matrices sharing with eigenvalues , but . This separates the question's common-vector condition from the stronger simultaneous diagonalization of the entire space.
Because is a real symmetric matrix, differentiating the quadratic form gives
Thus the stationary vectors are precisely , and at every one of them
The zero vector is always stationary. Nonzero stationary vectors exist exactly when is an eigenvalue of , and are exactly its nonzero eigenvectors with that eigenvalue. There is no norm constraint here; introducing one would change the stationary-value problem.
For a diagonalizable matrix, write . Relative to an existing fixed point,
Convergence to the same fixed point for every initial vector therefore holds exactly when
Here is the spectral radius. Necessity follows by choosing an initial displacement in each eigenspace; sufficiency follows because every displayed scalar power tends to zero. The condition also makes an invertible matrix, so that fixed point exists uniquely for every .
If instead different initial vectors may converge to different constant solutions, an eigenvalue is allowed: all other eigenvalues must have modulus less than one, and the component of in the eigenspace for must vanish. That component of the initial vector then remains unchanged. No other unit-modulus eigenvalue is compatible with convergence for arbitrary initial vectors.
Repeated substitution in the affine iteration gives
Therefore the required matrix polynomials are
These formulas do not require an inverse. If is an invertible matrix, multiplying the finite geometric sum by gives .
A fixed point obeys . Thus the necessary and sufficient existence condition is
If is not an eigenvalue of , then is an invertible matrix and the unique constant solution is . If is singular but the equation is consistent, all constant solutions are for any particular solution . For an existing fixed point, ; the successive difference in the question likewise obeys .
If are linearly independent, is the plane through with normal . Its equation is
so it contains the origin exactly when . If are proportional, is the line through in direction , containing the origin exactly when is proportional to . The linear span of two nonzero vectors has dimension one or two, so neither a single point nor all of occurs.
The affine subspace is . If are linearly independent, their linear span is all of , so is the whole plane. If they are proportional, is the line through in direction . This line passes through the origin exactly when . Because is nonzero, the set cannot be a single point.
In one dimension the nonzero vector already spans the real line. Hence , and translating it by leaves
This is an affine subspace containing the origin for every allowed choice of the three vectors.
For distinct positive integers , the product-to-sum identity gives
Both cosine integrals vanish over , so
If one index is zero, the corresponding function is identically zero and the same answer follows immediately.
For a positive integer , the identity gives
The stated nonnegative indexing includes as a special case: , whose squared norm is zero. Thus ; it cannot belong to a basis.
A norm is nonnegative, so it cannot equal a negative . The locus is therefore , irrespective of .
First assume . The first equation is a plane, and the second is a sphere centred at . Define its signed normal separation and projected centre by
The intersection of a plane and a sphere is
The circle lies in the given plane. Its radius follows from the Pythagorean theorem, splitting into orthogonal components. If , the first equation is the identity , so the locus is the entire sphere, rather than a planar section.
For , the norm equation requires . Thus the locus is
This includes , when the first condition imposes no restriction.
The orthogonal projection of the origin onto the line occurs at , giving closest point . Hence the distance from a point to a line in three dimensions is
Eliminating from the two plane equations gives , followed by . Taking gives the point . The entire intersection line is , where .
A direction of the intersection line must be orthogonal to both normal vectors. Their cross product is
Thus a convenient parallel vector is .

There are unlisted articles, also show them or only show them.