Q
stringlengths
4
3.96k
A
stringlengths
1
3k
Result
stringclasses
4 values
Theorem 16 (d'Alembert-Gauss). Every algebraic equation of degree \( \geq 1 \) has at least one complex root.
To see this, consider the function \( f\left( z\right) = 1/p\left( z\right) \) where \( p \) is a polynomial not vanishing anywhere. Since \( p\left( z\right) \) is analytic in \( \mathbb{C} \), so also is \( f \) (Chap. II, \( {\mathrm{n}}^{ \circ }{22} \), Theorem 17 - one may also invoke holomorphy, easier to prove ...
No
This is the case of the Riemann function \( \zeta \left( s\right) = \sum 1/{n}^{s} \) in the open set \( \operatorname{Re}\left( s\right) > 1 \) where the series converges.
For every \( \sigma > 1 \), the series converges normally in the half plane \( \operatorname{Re}\left( s\right) \geq \sigma \) since there \( \left| {1/{n}^{s}}\right| \leq 1/{n}^{\sigma } \) . The function is thus holomorphic in \( \operatorname{Re}\left( s\right) > \sigma \) for every \( \sigma > 1 \), so in fact in ...
Yes
Consider similarly the elliptic functions à la Weierstrass of Chap. II, \( {\mathrm{n}}^{ \circ }{23} \), the sums of the series \( \sum {\left( z - \omega \right) }^{-k} \) extended over the periods \( \left( {k > 3}\right) \), or, for \( k = 2 \), the modified series \( \wp \left( z\right) \) .
We have shown that these are analytic by proving by a routine calculation that, in every disc \( \left| z\right| < R \), they are the sum of a finite number of terms, corresponding to the periods situated in this disc, and of an explicitly calculated power series. The preceding theorem yields the result without any cal...
Yes
Consider (Chap. IV, \( {\mathrm{n}}^{ \circ }{20} \) ) Euler’s formula\n\n\[ P\left( z\right) = \mathop{\prod }\limits_{{n \geq 1}}{\left( 1 - {z}^{n}\right) }^{-1} = \mathop{\sum }\limits_{{n \geq 0}}p\left( n\right) {z}^{n} \]\n\nwhich appears in the theory of partitions.
Theorem 18 shows that the product \( \prod \left( {1 - {z}^{n}}\right) \) is holomorphic in the disc \( D : \left| z\right| < 1 \) and never vanishes; the left hand side is therefore also holomorphic in \( D \), whence the existence of an expansion in power series in \( D \) . Since we have \( p\left( n\right) \geq 1 \...
No
Let \( q \) be a constant complex such that \( \left| q\right| < 1 \) and let us consider the infinite product\n\n\[ f\left( z\right) = \prod \left( {1 + {q}^{n}z}\right) \]\n\nwhere the product is extended over all \( n \geq 1 \).
Since \( \sum \left| {q}^{n}\right| < + \infty \), Weierstrass’ theorem applies, the result being an entire function of \( z \) . It is clear that \( f\left( z\right) = \left( {1 + {qz}}\right) f\left( {qz}\right) \), so that the power series \( f\left( z\right) = \sum {a}_{n}{z}^{n} \) which represents \( f \) in the ...
Yes
Consider the infinite product\n\n\[ f\left( z\right) = z\prod \left( {1 - {z}^{2}/{n}^{2}}\right) \]\n\nextended over \( n \geq 1 \) . It satisfies the hypotheses of the theorem, with \( G = \mathbb{C} \) , so represents an entire function having simple zeros at the \( n \in \mathbb{Z} \) and is nonzero elsewhere.
Theorem 18 shows that\n\n\[ {f}^{\prime }\left( z\right) /f\left( z\right) = 1/z + \sum {2z}/\left( {{z}^{2} - {n}^{2}}\right) = \pi \cdot \cot {\pi z} = {\left( \sin \pi z\right) }^{\prime }/\sin {\pi z}. \]\n\nIn the connected open set \( \mathbb{C} - \mathbb{Z} \) where it is defined, the holomorphic function \( f\l...
Yes
Let us consider, with Weierstrass, the infinite product \( \prod \left( {1 - z/\omega }\right) \) extended over the periods \( \omega \in L \) .
\[ 1 - z/\omega = \exp \left( {-z/\omega - {z}^{2}/2{\omega }^{2} - \ldots }\right) = \] \[ = \;\exp \left( {-z/\omega - {z}^{2}/2{\omega }^{2}}\right) \exp \left( {-{z}^{3}/3{\omega }^{3} - \ldots }\right) \] with \( \left| {1 - \exp \left( {-{z}^{3}/3{\omega }^{3} - \ldots }\right) }\right| \leq M\left| {{z}^{3}/3{\o...
Yes
Theorem 19. Let \( f \) be a regulated function on \( \mathbb{T} \). Then\n\n\[ f\left( u\right) = \mathop{\lim }\limits_{\substack{{z \rightarrow 1} \\ {\left| z\right| < 1} }}{H}_{f}\left( {zu}\right) = \mathop{\lim }\limits_{\substack{{r \rightarrow 1} \\ {r < 1} }}\mathop{\sum }\limits_{\mathbb{Z}}{r}^{\left| n\rig...
In this way we immediately recover Weierstrass' theorem: every continuous and periodic function \( f \) is the uniform limit of trigonometric polynomials of the same period. The preceding theorem, with \( J = \mathbb{T} \), shows in fact that \( {H}_{f}\left( {ru}\right) \) converges uniformly on \( \mathbb{T} \) to \(...
Yes
Theorem 20. Every harmonic function \( H \) on a disc \( \left| z\right| < R \) has a series expansion of the form\n\n\[ H\left( z\right) = \sum {c}_{n}{r}^{\left| n\right| }{u}^{n} = {c}_{0} + \mathop{\sum }\limits_{{n > 0}}\left\lbrack {{c}_{n}{\left( x + iy\right) }^{n} + \overline{{c}_{n}}{\left( x - iy\right) }^{n...
with\n\n\[ {c}_{n}{r}^{\left| n\right| } = \int H\left( {ru}\right) {u}^{-n}{dm}\left( u\right) \]\n\nfor every \( r < R \) and every \( n \in \mathbb{Z} \).\n\nThere is no greater problem of convergence for the series (6) than for the power series of \( f \) : they converge normally in every disc of radius \( r < R \)...
Yes
Theorem 21. Let \( G \) an open set in \( \mathbb{C} \) and \( \left( {H}_{n}\right) \) a sequence of harmonic functions in \( G \) which converges uniformly on every compact \( K \subset G \) to a limit function \( H \) . Then \( H \) is harmonic, and, for any \( p \) and \( q \), the partial derivatives \( {H}_{n}^{\...
We know thanks to Borel-Lebesgue (Chap. V, \( {\mathrm{n}}^{ \circ }6 \) ) that uniform convergence on every compact subset is a property of local character: to verify it for every compact \( K \subset G \) it is enough to show that, for every \( a \in G \), it holds on a closed disc of centre \( a \) .\n\nSo choose an...
Yes
Theorem 22. Let \( f \) be a continuous function on \( \mathbb{T} \) . Then the function equal to\n\n(26.3)\n\n\[ \n{H}_{f}\left( z\right) = {\int }_{\mathbb{T}}\frac{1 - {\left| z\right| }^{2}}{{\left| z - u\right| }^{2}}f\left( u\right) {dm}\left( u\right) \n\]\n\nfor \( \left| z\right| < 1 \) and to \( f \) on \( \m...
Uniqueness follows from the maximum theorem; see the beginning of the preceding \( {\mathrm{n}}^{ \circ } \) .
No
Theorem 24. Let \( f \) be a function defined and continuous on \( \mathbb{R} \) such that\n\n(i) the series \( \sum f\left( {x + n}\right) \) converges normally on every compact set,\n\n(ii) \( \sum \left| {\widehat{f}\left( n\right) }\right| < + \infty \) .\n\nThen \( f \) is absolutely integrable on \( \mathbb{R} \)...
In practice, the convergence of the series \( \sum f\left( {x + n}\right) \) is almost always obtained by estimating \( f\left( x\right) \) for \( \left| x\right| \) large. Assume for example\n\n(27.8)\n\n\[ f\left( x\right) = O\left( {\left| x\right| }^{-s}\right) \;\text{ at infinity, with }s > 1. \]\n\nThe continuou...
Yes
Calculate the Fourier transform of the function \( f(t) = 1/(z + t)^s \) where \( z \) is a non real complex parameter and \( s \) an integer \( \geq 2 \), with \( \operatorname{Im}(z) > 0 \).
\[ \int \exp \left( -2\pi int \right) (z + t)^{-s} dt = \left\{ \begin{array}{ll} (-2\pi i)^s u^{s - 1} \exp (2\pi iuz) / (s - 1)! & \text{if } u > 0, \\ 0 & \text{if } u \leq 0 \end{array} \right. \] on condition that \( \operatorname{Im}(z) > 0 \).
Yes
\[ \widehat{f}\left( y\right) = \int \exp \left( {-t\left| x\right| - {2\pi ixy}}\right) {dx} \]
\[ \widehat{f}\left( y\right) = {2t}/\left( {{t}^{2} + 4{\pi }^{2}{y}^{2}}\right) . \]
Yes
Theorem 26. Let \( f \) be a continuous absolutely integrable function on \( \mathbb{R} \). Suppose that \( \widehat{f} \) is absolutely integrable. Then\n\n\[ f\left( x\right) = \int \widehat{f}\left( y\right) \mathbf{e}\left( {xy}\right) {dy}\;\text{ for every }x \in \mathbb{R}. \]
Note that the proof uses only the following facts: (i) formula (2), which we established using a \
No
Theorem 27. Let \( f \) be an absolutely integrable regulated function. Then\n\n\[ \mathop{\lim }\limits_{{N \rightarrow \infty }}{\int }_{-N}^{N}\widehat{f}\left( y\right) \mathbf{e}\left( {xy}\right) {dy} = \frac{1}{2}\left\lbrack {f\left( {x + }\right) + f\left( {x - }\right) }\right\rbrack \]\n\n at every point whe...
And why has the unknown constant \( c \) transformed itself surreptitiously into \( \frac{1}{2} \) ? Because, if one applies the formula to a sufficiently accommodating function, one already knows (Theorem 25) that the right hand side of (15) is equal to \( f\left( x\right) \) . So the constant \( c \) has no choice \(...
No
Let \( f \) be a regulated function such that \( \int \left| {{x}^{p}f\left( x\right) }\right| {dx} < + \infty \) . Then \( \widehat{f} \) is of class \( {C}^{p} \) and\n\n\[ \n{D}^{k}\widehat{f}\left( y\right) = \int {\left( -2\pi ix\right) }^{k}f\left( x\right) \mathbf{e}\left( {xy}\right) {dx}\;\text{ for every }k \...
Now suppose that \( f \) is of class \( {C}^{1} \) (or, more generally, a primitive of a regulated function) and that \( \int \left| {{f}^{\prime }\left( x\right) }\right| {dx} < \infty \) . Integrating by parts, one has, for \( y \neq 0 \) ,\n\n\[ \n{\int }_{-T}^{T}f\left( x\right) {e}^{-{2\pi ixy}}{dx} = {\left. f\le...
Yes
Theorem 1. A holomorphic function \( f \) on a domain \( G \) has primitive on \( G \) if and only if its integral along any admissible path in \( G \) depends only on the latter's endpoints.
In particular, the integral of \( f \) along a closed path, i.e. such that \( \mu \left( 0\right) = \) \( \mu \left( 1\right) \), is zero. In fact this condition is sufficient for ensuring the existence of a primitive. Indeed, if\n\n\[ \n{\mu }_{1},{\mu }_{2} : \left\lbrack {0,1}\right\rbrack \rightarrow G \n\]\n\nare ...
Yes
Theorem 3. Let \( G \) be a domain in \( \mathbb{C}, f \) a holomorphic function on \( G \) and \( {\mu }_{0},{\mu }_{1} \) two admissible paths in \( G \) . If one of the two following conditions is satisfied, then the integrals of \( f \) along \( {\mu }_{0} \) and \( {\mu }_{1} \) are equal:\n\n(a) there is a fixed-...
Exercise 2 (direct proof of Theorem 3). We keep the above construction and notation by, for example, supposing that \( \sigma \) is a fixed-endpoint homotopy; proving this result amounts to showing that integrals along \( {\gamma }_{p} \) and \( {\gamma }_{p + 1} \) are equal for all \( p \) . For this, take a closed p...
No
Let \( f \) be a holomorphic function on a simply connected domain \( G \) ; suppose that \( f \) does not vanish on \( G \) . Then, there is a holomorphic function \( g \) on \( G \) such that \( {e}^{g\left( z\right) } = f\left( z\right) \) for all \( z \in G \) ; it is unique up to addition of a multiple of \( {2\pi...
Since \( f \) does not vanish on \( G \), the function \( {f}^{\prime }/f \) is defined and holomorphic on \( G \), and hence admits a primitive \( g \) ; then \( {\left( {e}^{g}\right) }^{\prime } = {g}^{\prime }{e}^{g} = {e}^{g}{f}^{\prime }/f \), i.e. \( {\left( {e}^{g}\right) }^{\prime }f - {e}^{g}{f}^{\prime } = 0...
Yes
Theorem 5 (Cauchy’s residue formula). Let \( G \) be a domain in \( \mathbb{C}, S \) a closed discrete subset of \( G \) and \( f \) a holomorphic function on \( {G}^{\prime } = G - S \) . Then\n\n\[{\int }_{\mu }f\left( \zeta \right) {d\zeta } = {2\pi i}\mathop{\sum }\limits_{{a \in S}}{\operatorname{Ind}}_{\mu }\left...
The figure below is considered in the classical theory where \( S = \left\{ {{a}_{1},\ldots ,{a}_{n}}\right\} \) is finite. It contains the path \( \mu \) and a path \( \nu \) consisting, on the one hand, of\n\n![38edfff9-77bf-4f67-9514-89cc3c3d5a13_53_0.jpg](images/38edfff9-77bf-4f67-9514-89cc3c3d5a13_53_0.jpg)\n\nFig...
Yes
Theorem 8 (Dixon). Let \( G \) be a domain and \( \mu \) a closed path in \( G \) . The following statements are equivalent:\n\n(i) The integral along \( \mu \) of any holomorphic function on \( G \) is zero.\n\n(ii) The interior of \( \mu \) is contained in \( G \) .
(i) \( \Rightarrow \) (ii). For some \( a \notin G \), consider the function \( f\left( z\right) = 1/\left( {z - a}\right) \) ; by (i), its integral along \( \mu \) is zero, but, up to a factor \( {2\pi i} \), it is also the index of \( a \) with respect to \( \mu \) ; the latter is, therefore, zero, proving (ii).\n\n(...
No
Theorem 9. Let \( I \) be an interval, \( \mu \) a measure on \( I, U \) an open subset of \( \mathbb{C} \) and \( f : I \times U \rightarrow \mathbb{C} \) a function satisfying the following conditions :\n\n(a) \( f \) is continuous on \( I \times U \) ,\n\n(b) \( f\left( {t, z}\right) \) is a holomorphic function of ...
It is sufficient to prove the statements with respect to \( r = 1 \) : the general case will follow by a repeated application of the result.\n\nFirst proof. Theorem 24 bis of Chap. V, \( §7 \) is analogous to the result we need to prove but based on different assumptions: \( {f}^{\prime } \) (instead of \( f \) ) was a...
Yes
Theorem 10. On a simply connected domain, a real harmonic function is the real part of a holomorphic function and is unique up to the addition of a pure imaginary constant.
Uniqueness is obvious. Let \( u \) be a harmonic function; suppose there is a holomorphic function \( f = u + {iv} \) such that \( u = \operatorname{Re}f \) . Then\n\n\[ \n{f}^{\prime } = {D}_{1}u + i{D}_{1}v\;\text{ and }\;{D}_{1}v = - {D}_{2}u, \n\]\n\nand so \( {f}^{\prime } = {D}_{1}u - i{D}_{2}u \) . Conversely, s...
Yes
Theorem 12 (Paley-Wiener). Let \( \varphi \) be an entire function. The following two conditions are equivalent :\n\n(i) There is a number \( a > 0 \) such that \( \varphi \) satisfies \( \left( {PW}\right) \) for all \( n \) ;\n\n(ii) \( \varphi \) is the complex Fourier transform of a \( {C}^{\infty } \) function van...
It suffices to prove that (i) \( \Rightarrow \) (ii). The function \( {z}^{n}f\left( z\right) \) being bounded on every horizontal and in particular on \( \mathbb{R} \), the Fourier transform\n\n\[ \widehat{f}\left( t\right) = \int f\left( x\right) \mathbf{e}\left( {-{tx}}\right) {dx} \]\n\nof \( f \) on \( \mathbb{R} ...
Yes
Theorem 14. Let \( {\mathcal{S}}_{ + } = \mathcal{S}\left( {\mathbb{R}}_{ + }\right) \) be the set of functions defined and infinitely differentiable for \( x \geq 0 \) and that are together with their derivatives rapidly decreasing at infinity. For any \( f \in \mathcal{S}\left( {\mathbb{R}}_{ + }\right) \), the Melli...
(a) Assertions (i) and (ii) for \( \varphi = {\Gamma }_{f} \), where \( f \in \mathcal{S} \). Assertion (i) was proved before the theorem. So was the formula\n\n(13.13)\n\n\[ \n\operatorname{Res}\left( {\varphi , - k}\right) = {a}_{k} = {f}^{\left( k\right) }\left( 0\right) /k!.\n\]\n\n\( {}^{63} \) it is in fact bound...
Yes
Lemma 1. Let \( f \) be a function defined and infinitely differentiable on an open interval \( 0 < x < b \) . \( f \) can be extended to an infinitely differentiable function on the interval \( 0 \leq x < b \) if and only if the following conditions hold:\n\n(a) \( f \) has an asymptotic expansion\n\n\[ f\left( x\righ...
First of all, the relation \( f\left( x\right) = {a}_{0} + {a}_{1}x + o\left( x\right) \) shows both that \( f\left( x\right) \) tends to \( {a}_{0} \) as \( x \) approaches 0 and that if we define \( f\left( 0\right) = {a}_{0} \), then the function \( f \) thus extended has a derivative equal to \( {a}_{1} \) at the o...
Yes
Lemma 2. Let \( \varphi \) be a function defined and holomorphic on an open set\n\n\[ U : a < \operatorname{Re}\left( s\right) < b,\;\operatorname{Im}\left( s\right) > c \]\n\nand \( m \) be a real number. If the function \( {s}^{m}\varphi \left( s\right) \) is bounded at infinity on the closed vertical strip of finite...
To see this, argue as in \( {\mathrm{n}}^{ \circ }4 \) ,(iv). Take a closed strip\n\n\[ B : \operatorname{Im}\left( s\right) \geq {c}^{\prime } > c,{a}^{\prime } \leq \operatorname{Re}\left( s\right) \leq {b}^{\prime }\text{with}a < {a}^{\prime } < {b}^{\prime } < b \]\n\ncontained in \( U \) and choose some \( r > 0 \...
Yes
Theorem 1. Let \( f \) be a \( {\mathbb{R}}^{q} \) -valued \( {C}^{s} \) map defined in the neighbourhood of \( 0 \) in \( {\mathbb{R}}^{p} \) and such that \( f\left( 0\right) = 0 \) . Suppose that the rank \( r \) of \( f \) is constant in the neighbourhood of 0 . Then there are \( {C}^{s} \) local charts \( \left( {...
To prove the theorem, first note that, up to a permutation of the canonical coordinates in \( {\mathbb{R}}^{p}\; \) and \( \;{\mathbb{R}}^{q},\;D\left( {{f}^{1},\ldots ,{f}^{r}}\right) /D\left( {{x}^{1},}\right. \) \( \left. {\ldots ,{x}^{r}}\right) \neq 0 \) may be assumed to hold at 0, hence also in a neighbourhood o...
Yes
Theorem 3. Let \( G \) be a domain in a Cartesian space \( E \) and \( \omega \) a closed differential form of class \( {C}^{1} \) on \( G \) . The integrals of \( \omega \) along the two admissible paths \( {\mu }_{0} \) and \( {\mu }_{1} \) in \( G \) are equal if one of the following conditions holds:\n\n(a) There i...
From this, we can deduce that if all closed paths in \( G \) are homotopic to a constant path, then any closed form on \( G \) has a primitive; a domain with this property is said to be simply connected.\n\nHomotopy being an equivalence relation, it is then obvious that condition (b) of the theorem always holds. The sa...
Yes
Theorem 4. Let \( U \subset {\mathbb{R}}^{n} \) be an open bounded set, \( A \) its closure and \( \varphi \) a diffeomorphism of class \( {C}^{1} \) from \( A \) to a compact subset \( B \subset {\mathbb{R}}^{n} \) . Suppose that the borders of \( A \) and \( B \) have measure zero. Then, for any integrable \( {}^{39}...
We will prove the formula for continuous functions. By \( {\mathrm{n}}^{ \circ }{10} \) and 11 of Chapter V, \( §2 \) it can then be immediately generalized to lsc (resp. usc) functions; for this, simply observe that if there is an increasing (resp. decreasing) philtre \( \Phi \) for continuous functions, then the func...
No
Lemma 1. Let \( U \) be an open subset of \( {\mathbb{R}}^{n} \) containing 0 and \( \psi : U \rightarrow {\mathbb{R}}^{n} \) a \( {C}^{1} \) map such that \( \psi \left( 0\right) = 0,{\psi }^{\prime }\left( 0\right) = 1 \) . Given a number \( q \) such that \( 0 < q < 1 \) , let \( r \) be a number \( > 0 \) such that...
The derivative \( {\psi }^{\prime }\left( x\right) \) being a continuous function of \( x \) equal to 1 at \( x = 0 \) , the existence of \( r \) for given \( q > 0 \) is obvious. Having said that, suppose \( x \in K\left( r\right) \), so that \( {tx} \in K\left( r\right) \) for \( 0 \leq t \leq 1 \) . The derivative o...
Yes
For all \( q > 0 \), there exists \( r > 0 \) such that, for \( x, y \in A \), \[ \left| {x - y}\right| \leq r \Rightarrow \begin{Vmatrix}{{\varphi }^{\prime }\left( x\right) - {\varphi }^{\prime }\left( y\right) }\end{Vmatrix} \leq q/\begin{Vmatrix}{{\varphi }^{\prime }{\left( x\right) }^{-1}}\end{Vmatrix}. \]
The map \( x \mapsto {\varphi }^{\prime }\left( x\right) \) being continuous on \( A \) and \( {\varphi }^{\prime }\left( x\right) \) being invertible for all \( x \in A \), the map \( x \mapsto {\varphi }^{\prime }{\left( x\right) }^{-1} \) is also continuous on \( A \) : indeed Cramer’s formulas \( {}^{41} \) tell us...
Yes
Lemma 1. Let \( A \) and \( B \) be two disjoint closed subsets in a metric space \( X \) . There is a function \( f \) defined and continuous on \( X \) satisfying \( f\left( x\right) = 1 \) on \( A, f\left( x\right) = 2 \) on \( B \) and \( 1 \leq f\left( x\right) \leq 2 \) everywhere.
The lemma is trivial if \( A \) is empty (take \( f = 2 \) everywhere) or if \( B \) is empty (take \( f = 1 \) everywhere). Otherwise, choose a distance function \( d\left( {x, y}\right) \) defining the topology on \( X \) and set\n\n\[ f\left( x\right) = \inf \left\lbrack {d\left( {x, A}\right) ,{2d}\left( {x, B}\rig...
Yes
Lemma 2. Let \( U \) be an open set and \( A \) a closed one contained in \( U \) . There is an open set \( V \) such that \( A \subset V \subset \bar{V} \subset U \) . If \( X \) is locally compact \( {}^{69} \) and \( A \) is compact, \( \bar{V} \) may be assumed to be compact.
Here too, the first statement is trivial if \( U = X \) (take \( V = X \) ) or if \( A \) is empty (take \( V = \{ x\} \) with \( x \in U \) ). Otherwise, set \( B = X - U \), choose a function \( f \) by lemma 1 and take \( V = \{ f\left( x\right) < 3/2\} \), an open set containing \( A \) trivially and whose closure,...
No
Lemma 3. Let \( {U}_{0},\ldots ,{U}_{n} \) be open sets and \( X \) their union. There exist open sets \( {V}_{0},\ldots ,{V}_{n} \) whose union is \( X \) and such that \( {\bar{V}}_{i} \subset {U}_{i} \) for all \( i \) .
The \( {V}_{p} \) are constructed by induction on \( p \) by requiring them to also satisfy\n\n\[ \n{V}_{0} \cup \ldots \cup {V}_{p} \cup {U}_{p + 1} \cup \ldots \cup {U}_{n} = X.\n\]\n\nAs \( {V}_{0} \) only needs to satisfy\n\n\[ \nX - \left( {{U}_{1} \cup \ldots \cup {U}_{n}}\right) \subset {V}_{0} \subset {\bar{V}}...
No
Lemma 4. Let \( X \) be a locally compact metric space, \( A \) a compact subset of \( X \) and \( {\left( {U}_{i}\right) }_{1 \leq i \leq n} \) a finite open covering of \( A \) . Then, there exist positive valued continuous functions \( {f}_{i} \) on \( X \), with compact support and such that\n\n\[ \operatorname{Sup...
Set \( {U}_{0} = X - A \) . Lemma 3 gives open sets \( {V}_{i}\left( {0 \leq i \leq n}\right) \) covering \( X \) and such that \( {V}_{i} \subset \overline{{V}_{i}} \subset {U}_{i} \) . For \( 0 \leq i \leq n \), lemma 1 proves the existence of continuous functions \( {g}_{i} \) on \( X \), with values in \( \left\lbr...
Yes
Theorem 1. Let \( X \) be a compact Riemann surface.\n\n(a) Any function defined and holomorphic on \( X \) is a constant.
(a) is obvious: a function \( h \) everywhere holomorphic reaches its maximum somewhere, and so is constant in the neighbourhood of its maximum. As \( X \) is by definition connected, classical arguments apply verbatim. (Corollary: the only entire elliptic functions on \( \mathbb{C} \) are the constants).
No
Lemma 2. Let \( P\left( {X, Y}\right) \) be a polynomial with complex coefficients and \( E \subset \) \( {\mathbb{C}}^{2} \) the set of simple points \( {}^{7} \) of the curve \( P\left( {z,\zeta }\right) = 0 \) . Then \( E \) is a submanifold of \( {\mathbb{C}}^{2} = {\mathbb{R}}^{4} \) and, for all \( \left( {a,\alp...
Equivalently, there exists a unique holomorphic function \( f \) on \( V \) satisfying\n\n(2.3)\n\n\[ f\left( a\right) = \alpha \& P\left\lbrack {z, f\left( z\right) }\right\rbrack = 0\text{ for all }z \in V. \]\n\n\( f \) is called a local uniform branch at \( a \) of the algebraic function defined by \( P \) .\n\nSin...
Yes
Choose \( B = {\mathbb{C}}^{ * }, X = \mathbb{C} \) and \( p \) to be the map\n\n\[ \mathbf{e} : \zeta \mapsto \exp \left( {2\pi i\zeta }\right) \]\n\nfrom \( X \) onto \( B \) . As \( {\mathbf{e}}^{\prime }\left( \zeta \right) \neq 0 \) everywhere, \( p \) is a local homeomorphism (Chap. VIII, \( {\mathrm{n}}^{ \circ ...
The periodicity of the exponential function shows that\n\n\[ {p}^{-1}\left( D\right) = \mathop{\bigcup }\limits_{\mathbb{Z}}{D}^{\prime } + n = \bigcup {D}_{n}, \]\n\nand if \( {D}^{\prime } \) (i.e. \( D \) ) is sufficiently small. the translates \( {D}_{n} \) of \( {D}^{\prime } \) are pairwise disjoint and homeomorp...
Yes
Theorem 2. Let \( \left( {X, B, p}\right) \) be a covering and \( \gamma : I \rightarrow B \) a path in \( B \) . There is a unique lifting \( \mu \) from \( \gamma \) to \( X \) with a given initial point. If two paths \( {\gamma }_{0} \) and \( {\gamma }_{1} \) in \( B \) are fixed-endpoint homotopic and if \( {\mu }...
We first prove the uniqueness of \( \mu \) . In the neighbourhood of some \( {t}_{0} \in I \) , a lifting \( \mu \left( t\right) \) of \( \gamma \), being a continuous function of \( t \), takes its values in a neighbourhood of \( \mu \left( {t}_{0}\right) \) mapped homeomorphically by \( p : X \rightarrow B \) onto a ...
No
Corollary 1. Let \( \left( {X, B, p}\right) \) be a covering and suppose that \( X \) is connected and simply connected. \( {}^{11} \) Let \( {\gamma }_{0} \) and \( {\gamma }_{1} \) be two paths in \( B \) with the same endpoints and let \( {\mu }_{0} \) and \( {\mu }_{1} \) be liftings to \( X \) of \( {\gamma }_{0} ...
If the condition holds, the two liftings are homotopic, since \( X \) is simply connected, and thus that is also the case of the given paths in \( B \) . The converse, which makes no assumptions on \( X \), is the second statement of theorem 2 .
No
For any closed path \( \gamma \) in \( {\mathbb{C}}^{ * } \), there is a unique integer \( n \) such that \( \gamma \) is homotopic to \( n\mathbf{u} : t \mapsto \mathbf{e}\left( {nt}\right) \) .
Indeed, the map \( \mathbf{e} : \mathbb{C} \rightarrow {\mathbb{C}}^{ * } \) transforms \( \mathbb{C} \) into a connected and simply connected covering space of \( {\mathbb{C}}^{ * }\left\lbrack {\left( \mathrm{i}\right) \text{, Example 1}}\right\rbrack \) . A lifting of \( \gamma \) is a path \( \mu \) in \( \mathbb{C...
Yes
Theorem 3. Every covering \( \left( {X, B, p}\right) \) of a simply connected and locally connected space is trivial.
It all amounts to showing the existence of a global section with given value \( \alpha \in X \) at a given point \( a \in B \) . To define it at an arbitrary \( z \in B \), connect \( a \) to \( z \) by a path \( \gamma \) and consider the lifting \( \mu \) of \( \gamma \) with initial point \( \alpha \) . If \( \gamma...
Yes
Theorem 4. Every connected and locally simply connected space \( B \) has a connected and locally simply connected covering \( \left( {X, B, p}\right) \) . It is unique up to isomorphism. If \( \left( {Y, B, q}\right) \) is a connected covering of \( X \) and if \( \alpha \in X \) and \( \beta \in Y \) are such that \(...
(a) Choose a point \( a \in B \) and consider the set of all paths \( \gamma : I \rightarrow B \) with initial point \( a \) . Two homotopic (understood to be fixed-endpoint homotopic) paths are considered to be equivalent. Let \( X \) be the set of classes of these paths. Define \( p : X \rightarrow B \) by associatin...
No
Theorem 5. Let \( \left( {Y,{D}^{ * }, q}\right) \) be a connected covering of order \( k < + \infty \) of the pointed disc \( {D}^{ * } : 0 < \left| z\right| < 1 \) . Then there is a holomorphic function \( \varphi \) in \( Y \) such that (i) \( \zeta \mapsto \varphi \left( \zeta \right) \) is a conformal representati...
This means that \( \left( {Y,{D}^{ * }, q}\right) \) is isomorphic to the covering of \( {D}^{ * } \) obtained by taking \( Y = {D}^{ * } \) and \( q\left( z\right) = {z}^{k} \) [(i), Example 2], or to the covering obtained by constructing the Riemann surface of the algebraic function \( {z}^{1/k} \) using the method o...
No
Let \( {\zeta }_{i}\left( {1 \leq i \leq n}\right) \) be the roots of an equation\n\n\[{\zeta }^{n} + {c}_{1}{\zeta }^{n - 1} + \ldots + {c}_{n} = 0\]\n\nwith complex coefficients. Then\n\n\[\sup \left| {\zeta }_{i}\right| \leq \max \left( {1,\left| {c}_{1}\right| + \ldots + \left| {c}_{n}\right| }\right) .
Let \( M \) be the left hand side of (6). Each root of (5) satisfies\n\n\[{\left| {\zeta }_{i}\right| }^{n} \leq \left| {c}_{1}\right| \cdot {\left| {\zeta }_{i}\right| }^{n - 1} + \ldots + \left| {c}_{n}\right| \leq \left| {c}_{1}\right| \cdot {M}^{n - 1} + \ldots + \left| {c}_{n}\right| ,\]\n\nwhence\n\n\[{M}^{n} \le...
Yes
Theorem 7. The Riemann surface of an irreducible algebraic equation is connected.
It suffices to prove this for the open subset \( X \) of \( \widehat{X} \) because the point \( {\eta }_{Y} \) adjoined to \( X \) can obviously be connected to points of \( X \) by paths.\n\nAs seen in section (ii) of the previous \( {\mathrm{n}}^{ \circ } \), like \( X \), any connected component \( {X}^{\prime } \) ...
No
Theorem 8. All meromorphic functions on \( \widehat{X} \) are rational functions of \( z \) and \( \zeta \) .
We will assume a result from the general theory of commutative field extensions though it is not hard to show. Otherwise the proof given will be complete.\n\nLet \( \varphi \) be a meromorphic function on \( \widehat{X} \) . It has finitely many poles projecting onto points \( {a}_{i} \in \widehat{\mathbb{C}} \) . Let ...
No
Lemma 2. Let \( M \) be a commutative field, \( K \) a subfield of \( M \) and \( \zeta \) an element of \( M \) satisfying an irreducible algebraic equation of degree \( n \) over \( K \). Then the subfield \( L \) of \( M \) generated by \( K \) and \( \zeta \) has dimension \( n \) over \( K \) and admits \( 1,\zeta...
Since \( {\zeta }^{n} \) is a linear combination of \( 1,\zeta ,\ldots ,{\zeta }^{n - 1} \) with coefficients in \( K \), the same is true for \( {\zeta }^{p} \) for all \( p > n \) :\n\n\[ \n{\zeta }^{n + 1} = \zeta \left( {{c}_{0} + \ldots + {c}_{n - 1}{\zeta }^{n - 1}}\right) = {c}_{0}\zeta + \ldots + {c}_{n - 1}{\z...
Yes
Lemma 3. Let \( M \) be a commutative field of characteristic \( 0, K \) a subfield of \( M \) and \( n \) an integer \( \geq 1 \) . Suppose that all \( x \in M \) satisfy an algebraic equation of degree \( \leq n \) with coefficients in \( K \) . Then, the dimension of \( M \) as a vector space over \( K \) is \( \leq...
This lemma is itself based on the primitive element theorem due to Dedekind for the field of algebraic numbers and valid for all fields of characteristic 0: if all \( x \in M \) are algebraic over \( K \), then for any finitely number of elements \( {x}_{1},\ldots ,{x}_{p} \in M \), there exists \( x \) such that the s...
No
Lemma 1. Let \( \mu \) be a positive linear functional on \( L\left( X\right) \) . Then\n\n(1.1)\n\n\[ \left| {\mu \left( f\right) }\right| \leq \mu \left( \left| f\right| \right) \text{ for all }f \in L\left( X\right) . \]
This is obvious if \( f \) is real since \( f = {f}^{ + } - {f}^{ - },\left| f\right| = {f}^{ + } + {f}^{ - } \) . If \( f = g + {ih} \) is complex-valued, multiplying \( f \) by a complex factor with absolute value 1, \( \mu \left( f\right) \) may be assumed to be real and \( > 0 \) . Then \( \mu \left( h\right) = 0 \...
Yes
For any compact set \( K \subset X \), there is a constant \( {M}_{K}\left( \mu \right) \) such that\n\n\[ \left| {\mu \left( f\right) }\right| \leq {M}_{K}\left( \mu \right) \parallel f\parallel \;\text{ for all }f \in L\left( {X, K}\right) . \]
Indeed, there is a positive-valued continuous function \( p \) on \( X \) equal to 1 on \( K \) and zero outside a compact set \( {K}^{\prime } \supset K \) . Then \( \left| {f\left( x\right) }\right| \leq \parallel f\parallel p\left( x\right) \) for all \( f \in L\left( {X, K}\right) \), and so\n\n\[ \left| {\mu \left...
Yes
Lemma 3 (Dini’s Theorem). Let \( \Phi \subset {L}_{\mathbb{R}}\left( X\right) \) be an increasing philtre \( \varphi = \sup \left( \Phi \right) \) . Suppose that \( \varphi \in L\left( X\right) \) . Then\n\n\[ \mathop{\lim }\limits_{\Phi }\parallel \varphi - f{\parallel }_{X} = 0\;\text{ and }\;\mu \left( \varphi \righ...
The proof is the same as in Chap. V, \( {\mathrm{n}}^{ \circ }{10} \) . We first suppose that \( \Phi \subset \) \( {L}_{ + }\left( X\right) \) . Let \( r \) be a number \( > 0 \) and \( K \) the support of \( \varphi \) . As \( \varphi \left( x\right) = \sup f\left( x\right) \) is everywhere \( < + \infty \), for all ...
Yes
Lemma 4. Let \( \Phi \) be an increasing philtre of continuous functions with compact support and \( \varphi \) its upper envelope. Then\n\n(1.6)\n\n\[ \n{\mu }^{ * }\left( \varphi \right) = \mathop{\sup }\limits_{\Phi }\mu \left( f\right) = \mathop{\lim }\limits_{\Phi }\mu \left( f\right) .\n\]
The first expression is obviously greater than the second one. So, it suffices to prove the reverse inequality. For any \( g \in {L}_{\inf }\left( X\right) \), let \( {\Phi }_{g} \subset {L}_{\mathbb{R}}\left( X\right) \) be the set of functions of the form \( \inf \left( {f, g}\right) \) where \( f \in \Phi \) . Since...
Yes
Theorem 1. The function \( \varphi \mapsto {\mu }^{ * }\left( \varphi \right) ,\varphi \in \mathcal{I} \), has the following properties:\n\n(i) Additivity:\n\n\[ \n{\mu }^{ * }\left( {\varphi + \psi }\right) = {\mu }^{ * }\left( \varphi \right) + {\mu }^{ * }\left( \psi \right) \n\] \n\n(ii) Passage to the limit for in...
The proofs are the same as in Chap. V, \( {\mathrm{n}}^{ \circ }{10} \) and 11. To prove (i), choose two increasing philtres \( \Phi \) and \( \Psi \) in \( {L}_{ + }\left( X\right) \) converging to \( \varphi \) and \( \psi \) and note that the functions \( f + g \), with \( f \in \Phi \) and \( g \in \Psi \), form an...
Yes
Theorem 3. All subsets of a null set are null; the finite or countable union of a family of null sets is null.
Obvious.
No
Theorem 4. Let \( f \) be a function with values in \( \left\lbrack {0, + \infty }\right\rbrack \) ; then\n\n(2.8)\n\n\[{\mu }^{ * }\left( f\right) = 0 \Leftrightarrow f\left( x\right) = 0\text{ ae. }\]\n\n(2.9)\n\n\[{\mu }^{ * }\left( f\right) < + \infty \Rightarrow f\left( x\right) < + \infty \text{ ae. }\]
Supposing that \( f \geq 0 \) and \( {\mu }^{ * }\left( f\right) = 0 \), let us consider the set \( {N}_{p} \) of \( x \) where \( f\left( x\right) > 1/p \) and its characteristic function \( {\chi }_{p} \) . Then \( f \geq {\chi }_{p}/p \), whence\n\n\[{\mu }^{ * }\left( {N}_{p}\right) = {\mu }^{ * }\left( {\chi }_{p}...
Yes
Theorem 5. Let \( \sum {f}_{n}\left( x\right) \) be a series of complex-valued functions such that \( \sum {N}_{p}\left( {f}_{n}\right) < + \infty \) for given \( p \geq 1 \) . Then\n\n(3.8)\n\n\[ \sum \left| {{f}_{n}\left( x\right) }\right| < + \infty \text{ ae. } \]\n\nand any function \( f \) such that\n\n(3.9)\n\n\...
Let us start with the case of a series \( \sum {f}_{n} \) whose terms are \( \geq 0 \) and denote the partial sums by \( {s}_{n} \) . The increasing sequence of functions \( {s}_{n}^{p} \) converges to \( {f}^{p} \), where for all \( x \), we set \( f\left( x\right) = \sum {f}_{n}\left( x\right) \leq + \infty \) . Then...
Yes
Lemma 1. If \( f \) and \( g \) are integrable, so is \( {\alpha f} + {\beta g} \) for all \( \alpha ,\beta \in \mathbb{C} \), and\n\n\[ \mu \left( {{\alpha f} + {\beta g}}\right) = {\alpha \mu }\left( f\right) + {\beta \mu }\left( g\right) .
This follows from the inequality\n\n(4.4)\n\n\[ {N}_{1}\left\lbrack {\left( {f + g}\right) - \left( {{f}_{n} + {g}_{n}}\right) }\right\rbrack \leq {N}_{1}\left( {f - {f}_{n}}\right) + {N}_{1}\left( {f - {f}_{n}}\right) \]\n\nand from the linearity of the integral of continuous functions.
No
Lemma 2. Let \( \left( {f}_{n}\right) \) be a sequence of functions in \( {L}^{p} \) and \( f \) a function such that \( \lim {N}_{p}\left( {f - {f}_{n}}\right) = 0 \) . Then \( f \) is in \( {L}^{p} \) . If \( p = 1 \), then
(4.6)\n\n\[\n\mu \left( f\right) = \lim \mu \left( {f}_{n}\right)\n\]\n\nObvious.
No
Theorem 6. Let \( \left( {f}_{n}\right) \) be a sequence of functions in \( {L}^{p} \) such that \( \sum {N}_{p}\left( {f}_{n}\right) < \) \( + \infty \) . Then \( \sum \left| {{f}_{n}\left( x\right) }\right| < + \infty \) ae. Any function \( f \) satisfying \( f\left( x\right) = \overline{\sum }{f}_{n}\left( x\right) ...
This is theorem 5 applied to functions in \( {L}^{p} \) . By lemma 2, the (class of the) limit function \( f \) is still in \( {L}^{p} \) and (8) follows from (4) and (7).
Yes
Theorem 7 (Riesz-Fischer). The normed vector space \( {L}^{p} \) is complete. \( A \) sequence of functions \( {f}_{n}\left( x\right) \) in \( {L}^{p} \) converging in mean to a limit \( f \) contains a subsequence converging almost everywhere to \( f \) . If, moreover, the sequence \( {f}_{n}\left( x\right) \) converg...
Being a closed subspace of a complete space, \( {L}^{p} \) is complete. The second proposition of the statement is here too theorem 5 in the special case of \( {L}^{p} \) .
No
For any \( f \in {L}^{p} \), there is a series of continuous functions with compact support such that\n\n\[ \sum {N}_{p}\left( {f}_{n}\right) < + \infty ,\;f\left( x\right) = \sum {f}_{n}\left( x\right) \text{ ae. }, \]\n\n\[ \lim {N}_{p}\left( {f - {f}_{1} - \ldots - {f}_{n}}\right) = 0. \]
If \( f \) is real-valued, the \( {f}_{n} \) may be assumed to be real. Writing \( {\varphi }^{\prime }\left( x\right) = \) \( \sum {f}_{n}^{ + }\left( x\right) ,{\varphi }^{\prime \prime }\left( x\right) = \sum {f}_{n}^{ - }\left( x\right) \) defined lsc functions with values in \( \left\lbrack {0, + \infty }\right\rb...
Yes
For all \( f \in {L}^{p} \), there is a sequence of functions \( {f}_{n} \in L\left( X\right) \) and a function \( F \in {L}^{p} \) such that\n\n\[ f\left( x\right) = \lim {f}_{n}\left( x\right) \text{ ae.,}\;\left| {{f}_{n}\left( x\right) }\right| \leq \left| {F\left( x\right) }\right| \text{ ae. } \]
To prove the first point, it suffices to consider the partial sums \( {s}_{n}\left( x\right) \) of the series whose existence is ensured by corollary 1 and to choose \( F\left( x\right) = \) \( \sum \left| {{f}_{n}\left( x\right) }\right| \) . By theorem 6 applied to the \( \left| {f}_{n}\right|, F \in {L}^{p} \) .
No
Lemma 3. A lsc or usc function \( \varphi \) is integrable if and only if \( {N}_{1}\left( \varphi \right) < + \infty \) .
As \( - \varphi \) is lsc if \( \varphi \) is usc, we need only consider the case when \( \varphi \) is lsc and show that condition \( {N}_{1}\left( \varphi \right) < + \infty \) is sufficient. As \( \varphi \) is lsc, \( {\varphi }^{ + } \) is lsc, \( {\varphi }^{ - } \) is usc and these two functions satisfy the cond...
Yes
Lemma 4. A function \( f \) with values in \( \left\lbrack {-\infty , + \infty }\right\rbrack \) is integrable if and only if, for all \( r > 0 \), there exist integrable functions \( \varphi \) and \( \psi \), lsc and usc respectively, such that\n\n\[ \psi \leq f \leq \varphi \;\& \;{\mu }^{ * }\left( {\varphi - \psi ...
If \( f \) is integrable, there is a function \( u \in {L}_{\mathbb{R}}\left( X\right) \) such that \( {\mu }^{ * }\left( \left| {f - u}\right| \right) < r \) , hence a lsc function \( \theta \) such that\n\n\[ \left| {f - u}\right| \leq \theta ,\;{\mu }^{ * }\left( \theta \right) \leq r. \]\n\nAs \( - \theta \leq f - ...
Yes
Lemma 1. The upper (resp. lower) envelope of a finite family of functions of \( {L}^{p} \) is in \( {L}^{p} \) .
For real functions, inequality \( \left| {{f}^{ + } - {g}^{ + }}\right| \leq \left| {f - g}\right| \) shows that \( {N}_{p}\left( {{f}^{ + } - {g}^{ + }}\right) \leq \) \( {N}_{p}\left( {f - g}\right) \) . If \( f \) is in \( {\mathcal{L}}^{p} \), so are \( {f}^{ + } \) and \( {f}^{ - } \) . We conclude the proof using...
No
Theorem 8. Let \( \left( {f}_{n}\right) \) be an increasing sequence of positive functions in \( {L}^{p}.f = \sup {f}_{n} = \lim {f}_{n} \) is in \( {L}^{p} \) if and only if \( \sup {N}_{p}\left( {f}_{n}\right) < + \infty \) . Then \( \lim {N}_{p}\left( {f - {f}_{n}}\right) = 0. \)
The condition is clearly necessary since \( 0 \leq {f}_{n} \leq f \) for all \( n \) . To obtain the converse, it suffices to show that the sequence \( \left( {f}_{n}\right) \) satisfies Cauchy’s criterion in \( {L}^{p} \) . This is easy if \( p = 1 \) . Indeed, since \( {f}_{j} - {f}_{i} \) is integrable and positive ...
Yes
Lemma 2. Any decreasing sequence \( \left( {f}_{n}\right) \) of positive functions in \( {L}^{p} \) converge in mean.
Apply theorem 8 to functions \( {f}_{1} - {f}_{n} \) .
No
Lemma 3. The lower envelope of a countable family \( \left( {f}_{n}\right) \) of positive functions in \( {L}^{p} \) is in \( {L}^{p} \) .
Apply the previous result to the functions \( \inf \left( {{f}_{1},\ldots ,{f}_{n}}\right) \) ; we already know (lemma 1) they are in \( {L}^{p} \) .
No
Lemma 4. The upper envelope of a countable family of positive functions \( {f}_{n} \in {L}^{p} \) is in \( {L}^{p} \) if and only if there is function \( F \geq 0 \) such that \( {}^{11} \)\n\n\[ \n{N}_{p}\left( F\right) < + \infty \;\& \;{f}_{n}\left( x\right) \leq F\left( x\right) \text{ ae. for all }n.\n\]
The condition is necessary: for \( F \) take the upper envelope of the functions \( {f}_{n} \) . If it holds, the functions \( {g}_{n} = \sup \left( {{f}_{1},\ldots ,{f}_{n}}\right) \in {L}^{p} \) form an increasing sequence and satisfy \( {g}_{n} \leq F \) ae., whence \( {N}_{p}\left( {g}_{n}\right) \leq {N}_{p}\left(...
Yes
Theorem 11 (Hölder). Let \( p, q > 1 \) and \( r \geq 1 \) be real numbers such that \( 1/p + 1/q = 1/r \) . Then\n\n(5.5)\n\n\[ f \in {L}^{p}\;\& \;g \in {L}^{q} \Rightarrow {fg} \in {L}^{r} \]\n\nand\n\n(5.6)\n\n\[ {N}_{r}\left( {fg}\right) \leq {N}_{p}\left( f\right) {N}_{q}\left( g\right) \]
Relation (6) is a particular case of (3.4'), which holds without any integrability assumptions. It, therefore, suffices to prove (5). To this end, we choose sequences \( \left( {f}_{n}\right) \) and \( \left( {g}_{n}\right) \) in \( L\left( X\right) \) such that\n\n\[ \lim {N}_{p}\left( {f - {f}_{n}}\right) = 0\;\& \;\...
Yes
If \( f, g \in {L}^{2} \), then \( {fg} \) is integrable and\n\n\[ \left| {\mu \left( {fg}\right) }\right| \leq {N}_{2}\left( f\right) {N}_{2}\left( g\right) \]
This corollary shows that a Hilbert inner product\n\n\[ \left( {f \mid g}\right) = \mu \left( {f\bar{g}}\right) = \int f\left( x\right) \overline{g\left( x\right) }{d\mu }\left( x\right) \]\n\ncan be defined on \( {L}^{2} \). It is obviously linear \( {}^{14} \) in \( f \), satisfies the condition \( \left( {f \mid g}\...
No
Corollary 2. If \( f \in {L}^{p} \), then \( f{\chi }_{K} \in {L}^{1} \) for every compact set \( K \subset X \) and \( {fg} \in {L}^{1} \) for all \( g \in L\left( X\right) \) .
Indeed, the function \( {\chi }_{K} \) is in \( {L}^{1}\left( {{\mathrm{n}}^{ \circ }4\text{, lemma 7}}\right) \) and hence is in \( {L}^{q} \) for all \( q \) (theorem 10), whence \( f{\chi }_{K} \in {L}^{1} \) . Use the same argument for the product \( {fg} \) .
No
Lemma 5. The following two properties are equivalent for all functions \( \mathbf{j} \) :\n\n(LI 1) the function \( \mathbf{j}\left( x\right) {\chi }_{K}\left( x\right) \) is integrable for all compact sets \( K \) ;\n\n(LI 2) the function \( \mathbf{j}\left( x\right) f\left( x\right) \) is integrable for all \( f \in ...
(LI 1) \( \Rightarrow \) (LI 2): if \( f \) vanishes outside a compact set \( K \), then \( f\mathbf{j} = f \cdot \mathbf{j}{\chi }_{K} \) . Hence, corollary 2 applied to \( f{\chi }_{K} \) gives the result.\n\n(LI 2) \( \Rightarrow \) (LI 1): if \( f\mathbf{j} \in {L}^{1} \) for all \( f \in L\left( X\right), f \) can...
Yes
Theorem 13. All functions belonging to a \( {L}^{p} \) space are measurable.
Because it is the limit almost everywhere of a sequence of continuous functions with values in \( \mathbb{C} \), the archetype of a metrizable and separable space.
No
Theorem 14. Let \( P \) be a metrizable and separable set. The inverse image of every Borel set \( B \subset P \) under a measurable map \( f : X \rightarrow P \) is measurable.
Let \( \mathcal{T} \) be the set of subsets \( E \) of \( P \) such that \( {f}^{-1}\left( E\right) \) is measurable. Elementary formulae about the inverse image of an intersection or a complement show that, like the family of measurable subsets of \( X,\mathcal{T} \) is a tribe. However, by definition, \( \mathcal{T} ...
Yes
Lemma 1. Let \( {f}_{1},\ldots ,{f}_{n} \) be maps from \( X \) to metrizable and separable sets \( {P}_{1},\ldots ,{P}_{n} \) . Then the map\n\n\[ f : x \mapsto \left( {{f}_{1}\left( x\right) ,\ldots ,{f}_{n}\left( x\right) }\right) \]\n\nfrom \( X \) to the product space \( P = {P}_{1} \times \ldots \times {P}_{n} \)...
First, the product space is metrizable and separable since if, for each \( i \) , \( {D}_{i} \) is everywhere dense in \( {P}_{i} \), then the product of the sets \( {D}_{i} \) is everywhere dense in \( P \) . If \( f \) is measurable, then, for any Borel set \( B \subset {P}_{1} \), the set\n\n\[ {f}^{-1}\left( {B \ti...
Yes
Lemma 2. Let \( P \) and \( Q \) be two separable and metrizable spaces, \( \varphi \) a Borel map from \( P \) to \( Q \) and \( f \) a measurable map from \( X \) to \( P \) . Then \( g = \varphi \circ f \) : \( X \rightarrow Q \) is measurable.
It suffices to write that \( {g}^{-1}\left( B\right) = {f}^{-1}\left( {B}^{\prime }\right) \), where \( {B}^{\prime } = {\varphi }^{-1}\left( B\right) \).
Yes
Theorem 15. Let \( {P}_{1},\ldots ,{P}_{n} \) and \( Q \) be separable and metrizable spaces, \( \varphi \) a Borel map from \( {P}_{1} \times \ldots \times {P}_{n} \) to \( Q \) and \( {f}_{i} : X \rightarrow {P}_{i} \) measurable maps. Then the map\n\n\[ x \mapsto \varphi \left\lbrack {{f}_{1}\left( x\right) ,\ldots ...
This is a direct consequence of the previous two lemma.
No
Theorem 17 (Lusin). A function \( f \) with values in a metrizable and separable space \( P \) is measurable if and only if \( f \) satisfies condition (LUS).
(a) Let us first suppose that there is a countable partition of \( X \) into measurable sets \( {E}_{n} \) such that \( f \) is constant on each of them. We will then say that \( f \) is a step function. This is a less trivial analogue of the step functions of Chap. V, \( {\mathrm{n}}^{ \circ } 1. All functions of this...
No
Theorem 18. A function \( f \) with values \( i{n}^{25}\mathbb{C} \) or \( \left\lbrack {-\infty , + \infty }\right\rbrack \) is in \( {L}^{p} \) if and only if it is measurable and \( {N}_{p}\left( f\right) < + \infty \) .
(a) As \( \left| {f\left( x\right) }\right| < + \infty \) ae., we need only consider the case of finite-valued function. By Lusin, there is a sequence of compact sets \( {K}_{n} \), which may be assumed to be increasing, such that \( f \) is continuous on each \( {K}_{n} \) and zero ae. outside the union of these \( {K...
Yes
Corollary 1. The product of a function of \( {L}^{p} \) and a bounded measurable function is in \( {L}^{p} \) .
Obvious.
No
Corollary 3. Let \( f \) be a reasonable and measurable function with values in \( \left\lbrack {0, + \infty }\right\rbrack \) . Then there is an increasing sequence \( \left( {f}_{n}\right) \) of bounded positive integrable functions such that\n\n\[ f\left( x\right) = \lim {f}_{n}\left( x\right) \;\text{ for all }x \i...
Indeed, let \( {A}_{n} \) be an increasing sequence of integrable sets whose union is the measurable set \( \{ f\left( x\right) \neq 0\} \) . We set\n\n\[ {f}_{n}\left( x\right) = \begin{array}{l} \inf \left\lbrack {f\left( x\right), n}\right\rbrack \;\text{ if }x \in {A}_{n}, \\ 0\;\text{ otherwise. } \end{array} \]\n...
Yes
Lemma 2. A locally compact metric space \( X \) is separable if and only if it is countable at infinity.
Indeed if \( X \) is the countable union of compact sets \( {K}_{n} \), each \( {K}_{n} \) contains an everywhere dense countable subset \( {D}_{n} \) . This gives the result for \( X \) by considering the union of all \( {D}_{n} \) . Conversely, if \( \left( {U}_{n}\right) \) is a countable basis for the topology on \...
Yes
Lemma 3. Every open subspace \( P \) of a Polish space \( X \) is Polish.
Let \( P = X - F \), where \( F \) is closed, and we set\n\n\[ f\left( x\right) = d\left( {x, F}\right) \]\n\nassuming that \( X \) is complete with respect to \( d \) . The relation \( f\left( x\right) \neq 0 \) is equivalent to \( x \in P \) . In the Cartesian product \( \mathbb{R} \times X \), which is obviously\n\n...
Yes
Corollary 1. Let \( P \) a metrizable space. Suppose that \( P \) is complete with respect to some distance compatible with its topology. Then \( P \) is a \( {G}_{\delta } \) set in its completion with respect to any distance compatible with its topology.
Obvious.
No
Every Polish space is homeomorphic to a \( {G}_{\delta } \) set in a compact Polish space.
To prove this, let us consider the infinite-dimensional cube \( {I}^{\mathbb{N}} \), where \( I = \left\lbrack {0,1}\right\rbrack \) . It is the set of maps or sequences \( u : \mathbb{N} \rightarrow I \) equipped with the distance\n\n\[ d\left( {u, v}\right) = \sum {2}^{-n}\left| {{u}_{n} - {v}_{n}}\right| \]\n\nwhich...
No
Theorem 22. Let \( X \) and \( Z \) be locally compact spaces, \( {\mu }_{x} \) a family of positive measures on \( Z \) depending on a parameter \( x \in X \), and \( \lambda \) a positive measure on \( X \) . Suppose that condition (INT) holds and that this is also the case of one of the following conditions:\n\n( \(...
Proposition (i) in the statement is obvious. \( {}^{44} \)
No
Theorem 23. Let \( X \) and \( Z \) be two locally compact spaces, \( \lambda \) a positive measure on \( X \) and \( p \) a measurable map from \( X \) to \( Z \) . Suppose that, for all \( f \in L\left( Z\right) \), the function \( f \circ p \) is \( \lambda \) -integrable. Let\n\n\[ \nu \left( f\right) = \lambda \le...
(i) A reasonable function \( f \) on \( Z \) is \( \nu \) -integrable if and only if \( f \circ p \) is \( \lambda \) -integrable; then\n\n(14.3)\n\n\[ \nu \left( f\right) = \lambda \left( {f \circ p}\right) .\]
Yes
Lemma 1. Let \( X \) and \( Z \) be two locally compact spaces and \( p \) a continuous proper map from \( X \) to \( Z \) . The image \( p\left( F\right) \) of every closed set \( F \subset X \) is closed in \( Z \) .
The restriction of \( p \) to \( F \) being proper, we may assume that \( X = F \) . Let \( z \in Z \) be a closure point of \( p\left( X\right) \) . Since every compact neighbourhood \( W \) of \( z \) has non-trivial intersection with \( p\left( X\right) \), the compact sets \( {p}^{-1}\left( W\right) \) are non-empt...
No
Lemma 2. Let \( X \) and \( Z \) be two locally compact spaces, \( p \) a continuous proper surjective map from \( X \) to \( Z \) and \( f \) a map from \( Z \) to a topological space. \( f \) is continuous if (and only if) \( f \circ p \) is continuous.
By lemma \( 1, p \) maps every closed subset of \( X \) onto a closed subset of \( Z \) . Hence for a set \( A \subset Z \) to be closed it is sufficient (and necessary since \( p \) is continuous) for \( {p}^{-1}\left( A\right) \) to be closed. Hence \( V \subset Z \) is open if and only if \( {p}^{-1}\left( V\right) ...
Yes
Lemma 1. Let \( Z \) be a separated topological space and \( R \subset Z \times Z \) the graph \( {}^{56} \) of an equivalence relation on \( Z \) such that the map \( p : Z \mapsto Z/R \) is open. \( Z/R \) is separated if and only if \( R \) is closed in \( Z \times Z \) .
Let us set \( {Z}^{\prime } = Z/R \) . The map \( \left( {x, y}\right) \mapsto \left( {p\left( x\right), p\left( y\right) }\right) \) from \( Z \times Z \) to \( {Z}^{\prime } \times {Z}^{\prime } \) is continuous. However, the set \( R \) of ordered pairs such that \( p\left( x\right) = p\left( y\right) \) is the inve...
No
Lemma 2. Let \( G \) be a group acting on a separated topological space \( Z \) . The quotient space \( G \smallsetminus Z \) is separated if and only if the following condition holds: for every pair of points \( c,{c}^{\prime } \in Z \) such that \( {Gc} \neq G{c}^{\prime } \), there are neighbourhoods \( U \) and \( ...
We denote by \( R \) the equivalence relation defined by \( G \) . As was seen above, the map \( p \) is open. Let \( \left( {c,{c}^{\prime }}\right) \) be a closure point of \( R \) . Any neighbourhood of \( \left( {c,{c}^{\prime }}\right) \) contains a set \( U \times {U}^{\prime } \), where \( U \) and \( {U}^{\prim...
No
Lemma 4. The map \( f \mapsto {f}^{G} \) from \( {L}_{ + }\left( Z\right) \) to \( {L}_{ + }\left( X\right) \) is surjective.
First note that any compact set \( K \subset X \) is the image under \( p \) of a compact subset of \( Z \) . For all \( x = p\left( z\right) \) and all compact neighbourhoods \( W \) of \( z \), the image \( p\left( W\right) \) is indeed a compact neighbourhood of \( x \) . By BL, \( K \) can be covered by finitely ma...
Yes
Lemma 6. Let \( \Gamma \) be a discrete group acting on a locally compact space \( Z \) . The quotient space \( \Gamma \smallsetminus Z \) is locally compact, if the following condition holds:\n\n(GPD) for all compact sets \( A, B \subset Z \), the set of \( \gamma \in \Gamma \) such that \( {\gamma A}\# B \) is finite...
If \( z,{z}^{\prime } \) are two points of \( Z \) and \( U,{U}^{\prime } \) are sufficiently small neighbourhoods of \( z \) and \( {z}^{\prime } \), then\n\n(15.19)\n\n\[ {\gamma U}\# {U}^{\prime } \Leftrightarrow {\gamma z} = {z}^{\prime } \]\n\nThe first proposition is clear. If \( {\Gamma z} \neq \Gamma {z}^{\prim...
Yes
Theorem 25. Let \( X \) be a locally compact Polish space, \( \lambda \) a positive measure on \( X \) and \( \mathbf{j} \geq 0 \) a locally \( \lambda \) -integrable function. A function \( f \) with values in \( \mathbb{C} \) or \( \left\lbrack {-\infty , + \infty }\right\rbrack \) is integrable with respect to the m...
(a) As observed at the end of section (ii) of \( {\mathrm{n}}^{ \circ }{12} \), it suffices to set \( Z = X \) and\n\n(16.4)\n\n\[{\mu }_{x}\left( f\right) = \mathbf{j}\left( x\right) f\left( x\right) \;\text{ for all }f \in L\left( X\right)\]\nin order to put \( \nu \) into the form\n\n(16.5)\n\n\[\nu \left( f\right) ...
Yes
Lemma 2. Let \( f \) be a function with values in \( \mathbb{C} \) or \( \left\lbrack {-\infty , + \infty }\right\rbrack \) . Then \( f \) is measurable with respect to \( {d\nu } = \mathbf{j}{d\lambda } \) if and only if \( f\mathbf{j} \) is \( \lambda \) -measurable.
The function \( {\mathbf{j}}^{\prime }\left( x\right) \) equal to \( 1/\mathbf{j}\left( x\right) \) on \( S \) and 0 on \( X - S \), being \( \lambda \) -measurable like \( \mathbf{j} \) and constant outside \( S \), is \( \nu \) -measurable by theorem 26. So, if \( f\mathbf{j} \) is \( \lambda \) - measurable, and hen...
Yes
Theorem 27. Let \( \lambda \) and \( \nu \) be two positive measures on a locally compact Polish space \( X{.}^{68} \) There is a locally \( \lambda \) -integrable function \( \mathbf{j} \) such that \( {d\nu }\left( x\right) = \) \( \mathbf{j}\left( x\right) {d\lambda }\left( x\right) \) if and only if every \( \lambd...
(a) Let us first suppose that \( \nu \leq \lambda \) and that there is a compact set \( K \) such that \( \nu \left( {X - K}\right) = 0 \) . The first assumption shows that \( {\nu }^{ * }\left( f\right) \leq {\lambda }^{ * }\left( f\right) \) for any function \( f \geq 0 \) . Then, for all \( f \in L\left( X\right) \)...
No