Define the continuous linear functionals . For each fixed , weak convergence makes a convergent, hence bounded, scalar sequence. The Uniform boundedness principle states that a pointwise bounded family of bounded linear operators from a Banach space to a normed space has uniformly bounded operator norms. Since a Hilbert space is a Banach space, it applies to .
The Cauchy-Schwarz inequality gives , and evaluation at gives equality when ; equality also holds at zero. Consequently
From , one is an eigenvalue. Zero is also an eigenvalue: the nonzero continuous function has both endpoint values zero, so .
For , the linear projection identity gives an explicit bounded inverse:
Multiplying on either side by yields . There are no other points in the spectrum, and thus the spectrum of a nontrivial bounded projection here is
Let , as calculated in part (ii). Its endpoint values ensure . Since for all ,
This verifies the spectral radius formula explicitly:
Let . Since both linear functionals have operator norm one, for every the triangle inequality gives
Taking the essential supremum proves .
For the reverse bound, continuity and the extreme value theorem give with . Choose complex of modulus one so that , taking if . The continuous function
has supremum norm one by the triangle inequality and its endpoint values. Since ,
The output is continuous, so its L-infinity norm equals its supremum norm, including when is an endpoint. Therefore , and the exact operator norm is
This is the norm of a two-endpoint interpolation projection.
Because the linear functionals agree with endpoint point evaluations on continuous functions,
The endpoint values also make linearly independent. By its definition, the image of a linear map lies in . Conversely , so acts as the identity operator on . Thus
For every , , whence . It is a linear projection, and an annihilating polynomial is
Write ; its value at , if that point is in , may be assigned arbitrarily because Lp spaces identify functions equal almost everywhere. Its L-infinity norm is one.
The supplied estimate shows that cannot agree almost everywhere with any continuous function : choosing would give . Hence
is a direct sum, so every element has a unique expression .
For any complex with , define the linear functional
If , the supplied estimate, applied to , and absolute homogeneity of a norm give
For , the same bound follows from point evaluation. Thus , and makes its operator norm exactly one. The complex Hahn-Banach theorem extends to a norm-one linear functional on all of , with value at .
Conversely every norm-one extension satisfies . We have therefore determined all possible values:
In particular, taking and proves nonuniqueness of norm-preserving point evaluation extensions.
For a continuous function on , the L-infinity norm of its restriction to equals its supremum norm on . Indeed, values arbitrarily close to a maximum occur on an interval of positive Lebesgue measure, even if the maximum lies at an endpoint. Also, continuous functions equal almost everywhere on are equal everywhere on . Thus restriction embeds isometrically as a linear subspace of .
The point evaluation functional is complex-linear and satisfies . The constant function one attains equality, so . The complex Hahn-Banach theorem extends this bounded linear functional from that linear subspace to , preserving its operator norm. Therefore
The Neumann series criterion states that for a bounded linear operator on a Banach space implies
with convergence in the operator norm. In particular, suffices for invertibility of .
If , apply this criterion to . It gives
Thus no such belongs to the spectrum, and
For each integer , if , then is invertible by the same argument. The commuting factorization
then gives a bounded two-sided inverse for , namely the displayed sum multiplied by . Hence every satisfies . Taking the supremum over the spectrum and then the infimum over proves
Writing for the Banach space of bounded linear operators on , the spectrum is
Here is the identity operator. Since is a Banach space, the bounded inverse theorem makes this equivalent to failure of bijectivity of .
The spectral radius is
For a nonzero complex Banach space, the spectrum of a bounded linear operator is nonempty and compact, so the supremum is a maximum. If the zero space is allowed, its empty spectrum is assigned spectral radius zero.
When , and . Take the bounded measurable function
It represents an element of the Lp space , with L-infinity norm at most one. Since
where the lower endpoint contributes zero, its symmetric Lebesgue integral gives
For and , both sequences tend to zero through , but and . Therefore
Use , restricted to and regarded as a linear subspace of . The density of continuous functions in Lp on an interval says that is a dense subspace whenever .
For , the supremum norm gives
In particular , with zero limit. Part (c), together with being a dense subspace, implies . To identify the limit throughout , for and use
First choose in the dense subspace to make the first term arbitrarily small, then let . This is convergence on a dense subspace under a uniform operator bound, and proves
The zero linear functional belongs to the continuous dual space .
The answer is
Indeed, Lebesgue measure on is a sigma-finite measure, and duality of Lp spaces gives . The restricted Lebesgue measure is a non-atomic measure and has positive measure, so the Dual of L infinity is larger than L1: the canonical integral representation does not exhaust .
For , the Holder conjugate exponent also lies strictly between one and infinity. Duality of Lp spaces then holds in both directions, in agreement with Reflexivity of Lp spaces. For , the first required identification already fails.
For a normed space , let be its continuous dual space and its bidual space. Its canonical embedding into the bidual is
The definition of a reflexive space in this setting is
The Hahn-Banach theorem makes a linear map that is an isometry. For any normed space , its continuous dual space is a Banach space: a Cauchy sequence in the operator norm has scalar limits , and the uniform Cauchy bound proves both boundedness of and . Apply this with to see that is complete. If is onto, the isometry transfers this completeness to . Therefore
The kernel is complete: the map is an isometry from with its absolute-value norm onto with the supremum norm.
Define a linear map by
The fundamental theorem of calculus gives and . Also,
Thus is a bounded right inverse, and is a Banach space. Applying part (c),
Choose the continuously differentiable functions
The estimate for gives
Thus by uniform convergence, and the triangle inequality gives . The sequence is a Cauchy sequence in , but its unique possible supremum norm limit is , which is not differentiable at zero. That limit is outside , so
By the mean value theorem, a continuously differentiable function with zero derivative on is constant. Thus the kernel of the differentiation operator is
It is a closed set by the finite-dimensional subspace is closed theorem, applied in the normed space . It is proper because the function has nonzero derivative.
For , take . Its supremum norm is at most one, whereas and . Hence
The bounded linear functional from part (i) satisfies , where is its operator norm. Hence
These are equivalent norms. Applied to , the inequalities show that a sequence is a Cauchy sequence for either norm exactly when it is one for the other. Applied to , they show that the two norms give the same limits in . Thus every Cauchy sequence has a limit in one norm precisely when it does in the other:
Since is proper, is a nonzero linear functional. Choose with , by rescaling any vector outside its kernel. For every ,
If , applying gives . Consequently this is a direct sum, and the quotient vector space has dimension one. This computes the codimension of a vector subspace:
To prove boundedness without assuming completeness, use the linear functional with closed kernel criterion directly. Since is a closed set and , the distance
is positive. If , then , so
For the resulting estimate is immediate. Therefore
Here denotes the continuous dual space.
Put and form its period-one periodization of an integrable function
The function is not a Schwartz function, so justify the needed Poisson summation formula directly. For every , its derivative satisfies as . Thus the Weierstrass M-test gives uniform convergence on of this series and each differentiated series. It follows that is a smooth period-one periodic function.
Absolute integrability and Fubini's theorem allow its Fourier coefficients to be computed by splitting the real line into unit intervals:
These Fourier coefficients are absolutely summable. Their Fourier series defines a continuous periodic function with the same Fourier coefficients as ; uniqueness of Fourier coefficients in L1 identifies the two almost everywhere, and continuity then identifies them everywhere. Evaluating at zero gives the rigorously justified Poisson summation formula
Rearranging proves
The last series is a geometric series with sum . Equivalently, the left side is , where is the hyperbolic cotangent.
Splitting the Fourier transform integral at zero gives two absolutely convergent integrals:
Both and are integrable functions. Let , so . The duality property of the Fourier transform, or Fourier inversion in tempered distributions, gives
These Fourier transforms have continuous representatives by the Riemann-Lebesgue lemma, so the identity holds at every . Since is an even function,

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