Académique Documents
Professionnel Documents
Culture Documents
Appendix K of
A Companion to Analysis
T. W. K
orner
Introduction
Here is a miscellaneous collection of hints, answers, partial answers
and remarks on some of the exercises in the book. I expect that there
are many errors both large and small and would appreciate the opportunity to correct them. Please tell me of any errors, unbridgable gaps,
misnumberings etc. I welcome suggestions for additions. ALL COMMENTS GRATEFULLY RECEIVED. (If you can, please use LATEX 2
or its relatives for mathematics. If not, please use plain text. My e-mail
is twk@dpmms.cam.ac.uk. You may safely assume that I am both
lazy and stupid so that a message saying Presumably you have already
realised the mistake in question 33 is less useful than one which says
I think you have made a mistake in question 33 because not all left
objects are right objects. One way round this problem is to quote Xs
theorem.)
To avoid disappointment note that a number like K15* means that
there is no comment. A number marked K15? means that I still need
to work on the remarks. Note also that what is given is at most a
sketch and often very much less.
Please treat the answers as a last resort. You will benefit more from
thinking about a problem than from reading a solution. I am inveterate
peeker at answers but I strongly advise you to do as I say and not as
I do.
It may be easiest to navigate this document by using the table of
contents which follow on the next few pages.
Contents
Introduction
K1
K2
K3
K4
K5
K6
K7
K8
K9
K10
K11
K12
K13*
K14*
K15
K16
K17
K18
K19
K20
K21
K22
K23*
K24
K25
K26
K27
K28
K29
K30
K31
K32
K33
K34*
K35*
K36
K37*
K38
K39
K40
K41
K42
K43
2
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
K44
K45
K46*
K47
K48
K49
K50
K51
K52
K53
K54
K55
K56
K57
K58
K59
K60
K61
K62
K63
K64
K65
K66
K67
K68
K69
K70
K71
K72
K73
K74
K75
K76
K77
K78
K79*
K80
K81
K82
K83
K84
K85
K86
K87
K88
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
K89
K90
K91
K92
K93
K94
K95
K96
K97
K98
K99
K100
K101
K102
K103
K104
K105
K106
K107
K108
K109
K110
K111
K112
K113
K114
K115
K116
K117
K118
K119
K120
K121
K122
K123
K124
K125
K126
K127
K128
K129
K130
K131
K132
K133
100
101
102
103
104
105
106
107
108
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
K134
K135*
K136
K137*
K138
K139
K140
K141
K142
K143
K144
K145
K146*
K147
K148
K149
K150
K151
K152*
K153
K154
K155
K156
K157
K158
K159
K160
K161
K162
K163
K164
K165
K166
K167
K168
K169
K170
K171
K172
K173*
K174
K175*
K176
K177
K178
146
147
148
149
150
151
152
153
154
155
158
160
161
162
163
164
165
166
167
168
169
170
171
172
173
175
176
177
178
179
180
181
183
184
185
186
187
188
190
191
192
193
194
195
196
K179
K180
K181
K182
K183
K184*
K185
K186*
K187
K188
K189
K190
K191
K192
K193
K194
K195
K196
K197
K198
K199
K200
K201
K202
K203
K204*
K205
K206
K207
K208
K209
K210
K211
K212
K213
K214
K215
K216
K217
K218
K219
K220
K221
K222
K223
198
199
200
202
203
204
205
206
207
208
209
210
211
213
214
215
216
218
219
220
221
222
223
224
226
227
228
229
231
232
234
235
236
237
239
241
242
243
244
245
246
247
248
249
250
K224
K225
K226
K227
K228
K229
K230
K231
K232
K233
K234
K235
K236
K237
K238
K239
K240
K241
K242
K243*
K244
K245
K246
K247
K248*
K249*
K250
K251*
K252*
K253
K254
K255
K256
K257
K258
K259
K260
K261
K262
K263
K264
K265
K266
K267
K268
251
252
253
254
256
257
259
261
263
264
265
267
268
270
272
273
274
275
277
278
279
280
281
282
284
285
286
287
288
289
290
291
292
293
294
296
297
299
301
303
305
307
308
309
310
K269
K270
K271
K272
K273
K274
K275
K276
K277
K278
K279
K280
K281
K282
K283
K284
K285
K286
K287
K288
K289
K290
K291
K292
K293
K294
K295
K296*
K297
K298
K299
K300
K301
K302
K303
K304
K305
K306
K307
K308
K309
K310
K311
K312
K313
311
312
313
314
316
318
319
320
321
322
323
324
325
326
327
328
329
331
332
334
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
352
353
355
356
357
359
361
363
364
365
10
K314
K315
K316
K317
K318
K319
K320
K321
K322
K323
K324
K325
K326
K327
K328
K329*
K330
K331
K332
K333
K334
K335
K336
K337
K338
K339
K340
K341
K342
K343
K344
K345
366
368
369
370
371
373
374
375
377
378
379
381
382
384
385
386
387
388
389
391
392
393
394
395
396
400
401
402
403
404
405
406
11
K1
(ii) Every non-empty set of positive integers has a least member.
Since x is a rational number, qx is an integer for some strictly positive
integer. We can certainly find a k such that k + 1 > x k (ultimately
by the Axiom of Archimedes). But x is not an integer so k +1 > x > k.
Since mx, m and k are integers, m0 = mx mk is. Also
m0 x = mx2 mkx = mN k(mx)
12
K2
Write cn = dn + d1
n . Then
d2n cn dn + 1 = 0
p
cn c2n 4
.
dn =
2
Since the product of the two roots of F is 1, we must take the bigger
root. Thus
p
cn + c2n 4
c + c2 4
dn =
2
2
where c is the limit of cn .
13
K3
(Second part of problem.) Choose a1 = 2, b1 = 1, say.
14
K4
Take x = 0, f (t) = H(t), g(t) = 0. Take f (t) = 0, g(t) = H(t).
(H(t) = 1 for t > 0, H(t) = 0 for t 0.)
15
K5
(i), (ii) and (iii) are true and can be proved by, e.g. arguing by cases.
(iv) is false since f is continuous at 1/2. (f is, however, discontinuous
everywhere else.)
(v) is false. Take e.g. x = 161 + 171/2 , y = 31 171/2 .
16
K6
PN
n=1
by 1.
N
X
n=1
n=1
2n H(x qn ) H(y qn ) 0
n=1
H(x qn )
N
X
n=1
2n H(qm qn ) 2m
N
N
X
2n H(x qn )
2n H(y qn ) 2M
n=1
n=1
17
K7
Observe that an+1 an is decreasing. Thus either
(a) an+1 an > 0 for all n and, if an is unbounded an , or
(b) There exists an N such that an+1 an 0 for all n N and, if
an is unbounded, an .
For the reverse inequality, observe that an satisfies the original
inequality.
18
K8
A decreasing sequence bounded below tends to a limit, so, for any
fixed x (0, 1),
xn = fn (x) y, say.
1
4
x
4
x
2
19
K9
(ii) we need only look at positive integers. If n 2(k + 1)
n
X
n j
n
(1 + ) =
j
j=0
k+1
k+1
(n/2)k+1 k+1
(k + 1)!
nk+1
k+1
k+1
2 (k + 1)!
so
nk (1 + )n n
as n .
k+1
2k+1 (k + 1)!
20
K10
Observe that
x1 + x 2 + + x n
x1 + x2 + + xn1
xn 1
n
=
(n 1)
n n
n
n1
x1 + x 2 + + x n
1 x1 + x2 + + xn1
=
1
n
n
n1
1 (1 0)1 = 0
as n .
Either there exists an N such that m(n) = m(N ) for all n N and
the result is immediate or m(n) and
xm(n)
xm(n)
0
n
m(n)
as n .
If > 1, then
x1 + x2 + + xn xm(n) 1
x1 + x2 + + xn
1 0 = 0.
n
n
n
If < 1, then
x1 + x2 + + xn xm(n) 1
x1 + x2 + + xn
.
n
n
n
21
K11
(i) True. If a + b = c and a, c Q then b = c a Q.
(v) False.
If a 2 is rational, then, since ais irrational,
the statement
22
K12
P
P
j
n1j
j n1j
|.
| |x y| n1
(i) |xn y n | = |(x y) n1
j=0 |x| |y|
j=0 x y
P
j
(ii) If P (t) = N
j=0 aj t , then
|P (x) P (y)| |x y|
N
X
j=0
aj |j|Rj1 .
k=J+1
10k! 10(J+1)!
N
X
k=J+1
10k(J+1)
10
10(J+1)! .
9
23
K13*
No comments.
24
K14*
No comments.
25
K15
Apply the intermediate value theorem.
an converges (as in lion hunting) to c say.
Suppose f 0 (c) > 0. We can find an > 0 such that
f 0 (c)
f (t) f (c)
0
<
f
(c)
tc
2
and so
f (t) f (c)
f 0 (c)
>
tc
2
for all t with |t c| < . Thus f (t) > f (c) for c < t < c + and
f (t) < f (c) for c < t < c
Choose n such that 2n < . We have f (an ) < f (bn ) which is absurd.
26
K16
Repeat the lion hunting argument for the intermediate value theorem
with f (an ) g(an ), f (bn ) g(bn ). If an c, say, then, since f is
increasing this gives
g(bn ) f (bn ) f (c) f (an ) g(an ).
27
K17
Let > 0. Choose k > 0 such that |fk (x) f (c)| < for all
|x c| < k . Set = min1kN k . We claim that |g(x) g(c)| < for
all |x c| < .
To see this, suppose, without loss of generality, that fN (c) = g(c).
Then
and
g(x) g(c) +
if x 0,
0
fn (x) = nx
if 0 x < 1/n,
1
if 1/n x.
28
K18
(i) Observe that
{x R : E (, x] is countable}
29
K19
(i) False. Consider xj = 1.
(ii) False. Consider xj = j.
30
K20
(i) Let > 0. There exists an N such that an lim supr ar +
for n N . There exists an M such that n(p) N for p M . Now
lim supr ar + an(p) for p M . Since was arbitrary,
lim sup ar a.
r
so
lim sup ar + lim sup br + 2 > an + bn .
r
Thus
lim sup ar + lim sup br + 2 > lim sup(an + bn ).
r
But is arbitrary so
lim sup ar + lim sup br lim sup(an + bn ).
r
and so
an + bn an + lim inf br .
r
Thus
lim sup(an + bn ) lim sup an + lim inf br
n
31
K21
Thus
2
2
x
y
= kxk 2 x y + kyk
kxk2 kyk2
kxk2
kxkkyk kyk2
kyk2 2x y + kxk2
=
kxk2 kyk2
2
kx yk
.
=
kxkkyk
x
kx yk
y
32
K22
(i), (ii) and (iii) are true. Proof by applying the definitions.
(iv) is false. Take, for example, n = 1, U = {x : |x| < 1}.
33
K23*
No comments.
34
K24
(i) If (xn , yn ) E and k(xn , yn ) (x, y)k 0, then xn x and
yn y. Thus yn = 1/xn x and x = y so (x, y) E.
1 (E) = {x : x > 0} is not closed since 1/n 1 (E) and 1/n 0
/
1 (E) as n 0.
(ii) Let E = {(x, 1/x) : x > 0} {(1, 0)}. Then E is closed (union
two closed sets), 1 (E) = {x : x > 0} is not closed but 2 (E) = {y :
y 0} is.
(iii) If xn 1 (E) and xn x we can find yn such that (xn , yn ) E.
Since yn is a bounded sequence there exists a convergent subsequence
yn(j) y as j . Since xn(j) x the argument of (i) shows that
(x, y) E and x 1 (E). (It is worth looking at why this argument
fails in (ii).)
35
K25
(i) If G [K, K]n+m , then E [K, K]n .
(ii) Take n = m = 1, E = (0, ), f (x) = 1/x See question K.24.
(iii) Suppose k(xj , f (xj )) (x, y)k 0. Then kxj xk 0. Since
E is closed, x E. Since f is continuous, kf (xj ) f (x)k 0. Thus
y = f (x) and (x, y) G.
(iv) Take n = m = 1, E = R and f (x) = x1 for x 6= 0, f (0) = 0.
36
K26
If we write f (x) = 0 for x 6= 0 and f (0) = 1, then f is upper
semicontinuous but not continuous.
If f is not bounded we can find xn E such that f (xn ) n. Extract
a convergent subsequence and use upper semicontinuity to obtain a
contradiction. To show that the least upper bound is attained, take a
sequence for which f approaches its supremum, extract a convergent
subsequence and use upper semicontinuity again.
Take n = 1, K = [0, 1], f (x) = 1/x for x 6= 0 , f (0) = 0 to obtain
an upper semicontinuous function not bounded below.
Take n = 1, K = [0, 1], f (x) = x for x 6= 0 , f (0) = 1 to obtain an
upper semicontinuous function bounded below which does not attain
its infimum.
(As usual it is informative to see how the proof of the true result
breaks down when we loosen the hypotheses.)
37
K27
(i) f is continuous on [0, a] so is bounded and attains its bounds on
[0, a]. By periodicity f is everywhere bounded and attains its bounds.
(ii) Let f (x) = (x [x])1 ) if x is not an integer and f (x) = 0 if x is
an integer. (Here [x] is the integer part of x.)
(iii) By considering g(x) kx we see that we may suppose k = 0.
Given > 0 there exists an X > 0 such that |g(x + 1) g(x)| < (and
so |g(x + n) g(x)| < n for x > X. By hypothesis there exists a K
such that |g(y)| K for 0 y X + 1. Write n(x) for the integer
such that X < x n(x) X + 1. If x > X + 1 we have
x
x
n(x) K
+
x
x
as x . Since was arbitrary, g(x)/x 0 as x .
38
K28
No comments.
39
K29
Observe that xn can lie in at most two of the three intervals
[an1 , an1 + kn1 ], [an1 + kn1 , an1 + 2kn1 ]and [an1 + 2kn1 , bn1 ]
where kn1 = (bn1 an1 )/3.
40
K30
(i) Lots of different ways. Observe that the map J given by (x, y) 7
x y is continuous since
|(x + h) (y + k) x y| |x k| + |h y| + |h k|
(iv) If we assume the result true for Rn1 , then, since U has dimension n 1 and |U is self adjoint, U has an orthonormal basis e2 , e3 ,
. . . , en of eigenvectors for |U . Taking e1 = e, gives the desired result.
41
K31
(i) Observe that
|g(u) g(v)| kuvk
x
/ B(y, R) and
g(x) > R = g(z) g(x0 ).
x (b x)b E
(x (b x)b) = 0
x = b xb.
42
K32
(i) Argument as in K31.
(ii) Suppose x E and x y > 0. Then, if is small and positive,
we have
x = (1 )0 + x E
but the same kind of argument as in K30 and K31 shows that kyk >
ky xk.
(iii) Translation.
43
K33
Write B for the closed unit ball. If kzk < 1 then
z = (1 kzk)z + kzk0
z = x + (1 )y
and kzk = 1. Thus kxk = kyk = 1 and x, (1)y and z are collinear.
Inspection now shows that x = y = z.
A simpler pair of arguments gives the extreme points of the cube.
(ii) Since K is closed and bounded the continuous function x 7 x a
attains a maximum on K.
Suppose u0 is an extreme point of K 0 . If v, w K, 1 > > 0 and
then
x0 + u0 = v + (1 )w
x0 a = (x0 + u0 ) a
= v a + (1 )w) a
x0 a + (1 )x0 a
= x0 a.
44
K34*
No comments.
45
K35*
No comments
46
K36
(i) If neither K1 nor K2 have the finite intersection property, we can
find K1 , K2 , . . . KN , KN +1 , . . . , KN +M K such that
N
\
j=1
and so
Kj [a, c] = and
N\
+M
j=1
M
\
j=N +1
Kj [c, b] =
Kj [a, b] = .
a = a 0 a1 a2 a n bn b 2 b1 b0 = b
47
K37*
No comments
48
K38
Observe that, if [aj , bj ] K, then
N
\
j=1
1jN
49
K39
(i) f10 (t) = 1 cos t 0 for all t. Thus f1 is everywhere increasing.
But f1 (0) = 0 so f1 (t) 0 for t 0 so t sin t for t 0.
(ii) f20 (t) = f1 (t) 0 for t 0 and f2 (0) = 0.
(iv) We have
2N
X
(1)j t2j+1
j=0
(2j + 1)!
sin t
2N
+1
X
(1)j t2j+1
(2j + 1)!
2N
+1
X
(1)j t2j+1
(2j + 1)!
j=0
for t 0.
(2j + 1)!
sin t
j=0
(v) Thus
as N .
j 2j+1
X
(1) t
|t|N +1
0
sin t
50
K40
(i) Observe that g 0 is increasing since g 00 is positive. If g 0 (x1 ) > 0
then g 0 (t) g 0 (x1 ) > 0 for all t [x1 , x2 ] so 0 = g(x2 ) > g(x1 ) = 0
which is absurd. Thus g 0 (x1 ) 0 and g 0 (x2 ) 0. By the intermediate
value theorem there exist a c [x1 , x2 ] such that g 0 (c) = 0. Since g 0
is increasing g 0 (t) 0 and g is decreasing on [x1 , c] whilst g 0 (t) 0
and g is increasing on [c, x2 ]. Thus g(t) g(x1 ) = 0 on [x1 , c] and
g(t) g(x2 ) = 0. We use a similar argument to get the result on
[c, x2 ].
(ii) Look at g(t) = f (t) A Bt with A and B chosen to make the
hypotheses of (i) hold.
(iii) We may assume that 1 > n+1 > 0. The key algebraic manipulation in a proof by induction runs as follows.
n+1
!
!
n
X
X
j
Pn
j xj = f n+1 xn+1 + (1 n+1 )
f
xj
k
k=1
j=1
j=1
!
n
X
j
Pn
xj
n+1 f (xn+1 ) + (1 n+1 )f
k=1 k
j=1
n+1 f (xn+1 ) + (1 n+1 )
n+1
X
j f (xj )
n
X
j=1
Pn j
k=1
j=1
since
n
X
j=1
Pn j
k=1
= 1 and
n
X
k=1
k = (1 n+1 ).
f (xj )
51
K41
The area of the inscribed quadrilateral is
=
4
X
a2 sin j cos j =
1
2
4
X
a2 sin 2j .
j=1
j=1
P4
j=1 j
= .
j=1
The area is attained when the j are all equal (and with a little thought,
observing that sin00 t < 0 on (0, )) only then.
Consider a circumscribing n-gon A1 A2 . . . An . If Aj Aj+1 touches the
circle at Xj let P
2j1 be the angle Aj1 OXj and 2j be the angle
Xj OAj . Then r=1 2nr = 2 and the area of the circumscribed
polygon is
2n
!
2n
X
X
= 21 a2
tan r 21 a2 2n tan
r /2n = na2 tan(/n)
r=1
r=1
52
K42
Since g is continuous on [0, 1], it attains a maximum M say. Let
E = {x [0, 1] : f (x) = M }. Since E is non-empty it has a supremum
. Since f is continuous, g() = M . If = 0 or = 1 then M = 0. If
not then we can find a k with k, + k (0, 1) and
M = g() = 12 (g( k) + g( + k)) M
53
K43
Observe that
f (h) f (0)
= |h sin h4 | |h| 0
h
as h 0.
54
K44
Since g is continuous on [a, b], it attains a maximum at some c [a, b].
Since g 0 (c) = 0, we have c (a, b) and f 0 (c) = k.
For the final paragraph, note that f 00 (a) 0 f 00 (b).
55
K45
(i) The statement f must jump by at least 1 at z is not well
defined. If the reader feels that there should be some way of making this statement well defined in a useful way she should consider
f (x) = 21 sin(1/x) for x 6= 0, f (0) = 0.
However f can not be continuous at z. Observe that
max(|f (z) f (xn )|,|f (z) f (yn |)
(ii) The behaviour of f 0 (zn(j) ) does not control the value of f 0 (z).
Proposed result is false, see K43.
56
K46*
No comments.
57
K47
Since g 0 is constant, the conclusions of intermediate value theorem
hold trivially for g 0 .
58
K48
Observe that
cos n = <(cos + i sin )n
X n
=
(1)r cosn2r sin2r
2r
02rn
X n
(1)r cosn2r (1 cos2 )r
=
2r
02rn
Thus tTn (t) = (Tn+1 (t) + Tn1 (t))/2 for all t [1, 1].
59
K49
Two polynomials with the same roots and the same leading coefficient are equal.
Use the fact that |Tn (t)| 1 for T [1, 1].
Choose and so that + = 1, + = 1. If P is the
interpolating polynomial of degree n for f on [a, b], then taking g and
Q such that
g(t) = f (t + ), Q(t) = P (t + ),
we see that Q is the interpolating polynomial of degree n for g on
[1, 1]. If |f n (x)| A on [a, b] then |g n (x)| An = A((b a)/2)n on
[1, 1] so
(b a)n A
|f (t) P (t)|
22n1
on [a, b].
60
K50
(ii) P (t) =
n
X
f j (a)
j=0
j!
tj .
61
K51
If we set
P (t) =
n
X
j=0
Aj tj (1 t)n+1 +
n
X
j=0
Bj tn+1 (1 t)j ,
then the resulting system of equations for Aj is triangular with nonvanishing diagonal entries and thus soluble. The same holds for Bj .
If y (0, 1), set
xn+1 (1 x)n+1
.
y n+1 (1 y)n+1
F has vanishing r-th derivative at 0 and 1 for 0 r n and vanishes at
y. By Rolles theorem F 0 vanishes at least twice in (0, 1), F 00 vanishes
at least three times in (0, 1), F 000 vanishes at least four times in (0, 1),
. . . , F (n+1) vanishes at least n + 2 times in (0, 1), F (n+2) vanishes at
least n + 1 times in (0, 1), F (n+3) vanishes at least n times in (0, 1),
. . . , F 2n+2 vanishes at least once (at say) and this gives the result.
F (x) = f (x) P (x) E(y)
62
K52
(i) g(b) g(a) = g 0 (c)(b a) 6= 0 for some c (a, b).
(ii) A = (f (b) f (a))/(g(b) g(a)).
(iv) Observe that
with ct a as t a.
f 0 (ct )
f (t) f (a)
= 0
g(t) g(a)
g (ct )
63
K53
(i) We have
f (b) = f (a) + f 0 (a)h + f 00 (c)
h2
2!
h2
2!
and so
|f 0 (a)||h| 2M0 + 21 M2 |h|2
2M0 M2
+
h
2h
for all h > 0 and, choosing h = 2(M0 M1 )1/2 , we have the required
result.
|f 0 (t)|
(iii) Take f (t) = t to see that (a) and (b) are false.
(c) is true by the method of (ii), provided that L 2(M0 M1 )1/2 .
(d) True. g(t) = sin t.
1/2
1/2
M2 t) will do.
64
K54
(i) True. Choose a > 0 such that |f (t) f (s)| < 1 whenever
|t s| < 2 and an integer N > 2 + 1 . If x (0, 1) we can find
M N and x1 , x2 , . . . , xM (0, 1) such that x1 = x, xM = 1/2,
|xj+1 xj | for 1 j M 1. Since |f (xj+1 ) f (xj )| 1 we have
|f (x) f (1/2)| N and |f (x)| N + |f (1/2)|.
(ii) False. If f (t) = t then f does not attain its bounds.
(iii) False. Set f (t) = sin(1/t).
65
K55
(i) True. By the intermediate value theorem, we see that f (x) l
for some l as |x| . Given > 0, we can find a K such that
|f (x) l| < /2 for all |x| K. Since a continuous function on
a closed bounded set is uniformly continuous, we can find a with
1 > > 0 such that |f (x) f (y)| < for all x, y [K 2, K + 2]
with |x y| < . By going through cases, we see that |f (x) f (y)| <
for all x, y R with |x y| < .
(ii) False. Take f (t) = sin t.
(iii) False. Take f (t) = sin t2 .
(iv) False. Take f (t) = t.
(v) False. Take f (z) = exp(i|z|).
(vi) True. Observe that ||f (z)| |f (w)|| |f (z) f (w)|.
(vii) True. Let > 0. Choose > 0 such that |x y| < implies
||f (x)| |f (y)|| < /2.
Now suppose y > x and |x y| < . If f (x) and f (y) have the same
sign, then automatically, |f (x)f (y)| < /2. If they have opposite sign,
then by the intermediate value theorem, we can find u with y > u > x
and f (u) = 0. Since |x u|, |y u| < we have
|f (y) f (x)| |f (y) f (u)| + |f (u) f (x)| < /2 + /2 = .
66
K56
Observe that
so that
Thus
k(x + u) yk kx yk + kuk
f (x + u) f (x) + kuk.
kf (a) f (b)k ka bk
kx yk ky yn k
kx yk f (x)
as j . Thus kx yk = f (x).
67
K57
(i) Given > 0 we can find a > 0 such that, if u, v Q, then
|u v| < implies |f (u) f (v)| < . We can now find an N such
that |xn x| < /2 for n N . If n, m N then |xn xm | < so
|f (xn )f (xm )| < . Thus the sequence f (xn ) is Cauchy and converges.
(ii) Using the notation of (i), we can find an M such that |xm
x| < /2 and |ym x| < /2 for m M . Thus |xm ym | < and
|f (xm ) f (ym )| < for m M . Since was arbitrary, f (xm ) and
f (ym ) tend to the same limit.
(iii) Observe that (ultimately by the Axiom of Archimedes) any x
R is the limit of xn Q.
(iv) Take xn = x.
(v) We use the notation of (i). If |xy| < /3 we can find xn , yn Q
with xn x, yn y and |xn x| < /3, |yn y| < /3. Then
|xn yn | < for all n so |f (xn ) f (yn )| < and |F (x) F (y)| .
68
K58
PN
P
n
n
(i) If |z| < min(R, S) then N
n=0 bn z tends to limit
n=0 an z and
and so
N
N
N
X
X
X
bn z n
an z n +
(an + bn )z n =
n=0
n=0
n=0
tends to limit as N .
P
n
R < S If R
< |z| < S then
n=0 an z converges and
PSuppose
P
n
n
n=0 bn z diverges so
n=0 (an + bn )z diverges. Thus the sum has
radius of convergence R.
(ii) The argument of the first paragraph of (i) still applies.
n
If |z| < 1 then, provided n is sufficiently large,
n=0 an z diverges.
P
2
n
n
|z| (R/2) so n=0 an z n converges. The radius of convergence is
1.
If R = 0, then any radius of convergence with 0 1 is possible.
2
n
For = 0 take an = nn . For 0 < < 1 take an = n . For = 1
2
take an = n
where n tends to 1 very slowly from below.
n
Similar ideas for R = .
Pn
Pn
n
n
(vi) Let |cn | = max(|an |, |bn |). Since
j=0 |an z |
j=0 |cn z |
radius of convergence R0 at most R. Thus R0 min(R, S). But |cn |
|an | + |bn | so R0 min(R, S). Thus R0 = min(R, S).
If |cn | = min(|an |, |bn |) then similar arguments show that the radius
of convergence R00 max(R, S). But every value A max(R, S) is
possible for R00 . (If A < and R, S > 0 can take eg. a2n = R2n ,
b2n = A2n , a2n+1 = A2n1 , b2n = S 2n1 .
69
K59
Let > 0. We can find an N such that |h(n) | < for n N .
Observe that if n N then h(n) < + and so
Thus if |z| > 22(+)1 we have |an z n | . Thus the radius of convergence R satisfies R 22(+)1 . Since was arbitrary R 221 .
A similar argument shows that, if |z| < 22()1 |an z n | 0, and thus
R 221 so R = 221 .
Suppose 1 0. Define E inductively by taking an+1 = an /2 if
an /2 212 , an+1 = 2an otherwise. Then h(n) and |an z n | 1
whenever |z| = 221 and n is large. Thus we have divergence everywhere on the circle of convergence.
Suppose 1 0. Define E inductively by taking an+1 = 2an if
2an 2(12)n n2 , an+1 = an /2 otherwise. Then |an z n | n2 whenever |z| = 221 and n is large. Thus we have convergence everywhere
on the circle of convergence. When n is large we have
so
whence
an 4 2(12)n n2
2n2h(n)n 22+(12)n n2
(log 2)(n 2h(n)n) (log 2)(2 + (1 2)n 2 log n
70
K60
If |an |1/n is unbounded, the radius of convergence is zero.
We deal with the case (lim supn |an |1/n )1 = R with 0 < R < .
The other cases are similar.
Suppose R > > 0. Then we can find an N such that |an |1/n
(R /2)1 for all n N . Thus if |z| < R we have
n
R
n
|an z | <
R /2
P
for n N and n=0 |an z n | converges by comparison with a convergent
geometric sum.
We can also find n(j) such that |an |1/n(j) (R + /2)1 . Thus,
if |z| > R + we have
n(j)
R+
n(j)
|an z | >
.
R + /2
Since was arbitrary, R is the radius of convergence as required.
71
K61
Write = lim supn an+1 /an . If > 0, we can find an N with
an+1 /an + for n N . Thus, by induction, an A( + )n for
n N where A = aN ( + )N . Thus, if n N ,
a1/n
A1/n ( + ) +
n
2n+1
2n+1
2n
2n+1
72
K62
(iii) Could write bn = cn + dn with dn = 0 for n large and |bn | .
Thus lim supn |Cn | .
(v) If bj = (1)j limit is 0.
Ar(j+1)
X
n=0
n
(1 rj+1 )rj+1
+
= (1
and
M (j)+1
rj+1 )
N (j)
rj+1
n=N (j+1)
n
(1 rj+1 )rj+1
< 2j + 2j = 2j+1
M (j+1)
As(j+1)
n=N (j+1)
N (j+1)
= sj+1
(1 sj+1 )snj+1
M (j+1)+1
sj+1
73
K63
(ii) C can be identified with R2 .
P
n
N +1
(iii) (a) If z 6= 1, N
)/(1 z) tends to a limit if
n=0 z = (1 z
and only if |z| < 1. The limit is 1/(1 z). If z = 1 we have divergence.
(b) Observe that, if z 6= 1,
Cn = (n + 1)
= (n + 1)1
k
n X
X
zj
k=0 j=0
n
X
k=0
1 z k+1
1z
1 z n+1
1
1
(n + 1) + z
=
n+1
1z
1z
z 1 z n+1
1
+
.
=
1z n+1 1z
Thus we have convergence if and only if |z| 1, z 6= 1 and the limit is
1/(1 z).
74
K64
(v) If k then, using part (i), Bk b. Since this is true for
every such sequence, B b as .
(vi) We seek to Borel sum z j . Observe that, if z 6= 1
X
1
n 1 z n+1
e
=
(1 ze(1z) ).
n!
1
z
1
z
n=0
We have convergence if and only if <(1 z) < 1 and the Borel sum is
then 1/(1 z).
If z = 1,
X
X
n X n
n
B =
(n + 1) e =
e +
e = 1 .
n!
n!
n!
n=0
n=0
n=0
75
K65
(i) True. Since the sequence is bounded, there exists a b and a strictly
increasing sequence n(j) such that bn(j) b. Set ujn(j) = 1, ujk = 0
otherwise.
(ii) True. We can extract subsequences with different limits and use
(i).
(iii) False. Let bj = 1 for 2n n j 2n 1, bj = 1 for
2 + 1 j 2n + n [n 4], bj = 0 otherwise Observe that, if unk = 1
n
for 2P
N k 2n , unk = 1 for 2n + 1 k 2n + N , then U G
but
k=0 ujk bk = 2N for j sufficiently large. But N is arbitrary.
n
X
(1)r+1
uj(r)k bk
k=0
= (1)r+1
as r .
X
k=0
X
k=0
N (r)
N (r+1)
uj(r)k bk +
k=0
uj(r)k 2
k=N (r)+1
N (r)
X
k=0
|uj(r)k | 2
uj(r)k 2r+3 1.
uj(r)k bk +
k=N (r+1)+1
k=N (r+1)+1
|uj(r)k |
uj(r)k bk
76
K66
(i) Try bk = 1 for k = N , bk = 0 otherwise. Try bk = 1 for all k.
77
K67
(i) implies (ii) since absolute convergence implies convergence.
For the converse observe that, writing xn = (xn1 , xn2 , . . . , xnm ), we
have
kxj k |xj1 | + |xj2 | + + |xjm |.
P
Thus, if j=1 kxj k diverges, we must be able to find a k with 1 k m
P
such that
j=1 |xjk | diverges. Choose j so that j xjk 0. Then
N
N
N
X
X
X
j x j
j xjk =
|xjk | .
j=1
as N .
j=1
j=1
78
K68
If
n=1
an converges, then
X
X
X
an
an
an
nS1 ,nN
nN
n=1
nS2 ,nN
nS1 ,nN
an
nS2 ,nN
1/2
n=1
n2
2
1 2
2
an = (a1/2
n ) < (n ) = n
P
1/2
converges, so does nS2 an /n. Hence the result.
By Cauchy Schwarz,
!2
N
N
X
X
X
X
X a1/2
1
1
n
an
an
2
n
n
n2
n=1
n=1
n=1
n=1
n=1
79
K69
(i) Diverges
2N
X
1 + (1)n
n=1
as N .
N
X
1
=
n
n=1
2
(ii) Converges by comparison since p2
n n .
N (j)
an
X
n=1
an A
n=1
P
since an increasing sequence bounded above converges
n=1 an converges.
P
For the converse observe that if
n=1 an converges then, if X N ,
Z X
Z N
N
1
N
1 Z n+1
X
X
X
an .
an
f (t) dt K
f (t) dt
f (t) dt =
1
n=1
n=1
n=1
(vi) an = (1)n .
P N
(vii) Let M
a tend to b. Given > 0 we can find an N0 such
PM N n=1 n
that k n=1 an bk /2 for N N0 . We can also find an N1 such
that kan k /2M for n N1 . Let N2 = (M + 1)(N1 + N0 ). If
m N2 we can write m = N M + u with N N0 , 0 u M 1 and
M N + r N1 for all r 0. We then have
k
m
X
n=1
an bk k
MN
X
n=1
an bk +
MX
N +u
n=M N +1
kan k <
u
+
2 2M
80
as desired.
(viii) Set N (j) = 22j , an = j 1 if 22j n < 2 22j , an = j 1 if
2 22j n < 3 22j , an = 0 otherwise.
81
K70
(i) Observe that (an bn )1/2 (an + bn )/2 and use comparison.
(ii) Take bn = an+1 in (i).
(iii) Observe that an+1 (an an+1 )1/2 and use comparison.
(iv) a2n = 1, a2n+1 = n4 .
82
K71
(i) True. a4n 0 so there exists an N with |an | 1 for n N . Thus
|a5N | for n N and the result follows by comparison and the fact
that absolute converge implies convergence.
a4n
(ii) False. Take an = (1)n1/6 and use the alternating series test.
(iii) False. Take a2k = 2k , ar = 0 otherwise.
(iv) False. Same counterexample as (iii).
(v) True. Given > 0 we can find an N such that
for M N . Thus, if n 2N + 2
nan 2(n N 1)an 2
M
X
PM
j=N
aj < /2
aj < .
j=N
(vi) False. Take an = (n log n)1 for n 3 and use the integral
comparison test.
(vii) True. Abels test.
(viii) True. By CauchySchwarz,
N
!2
N
N
X
X
X
X
X
3/4
2
3/2
2
|an |n
|an |
n
|an |
n3/2 .
n=1
n=1
n=1
n=1
n=1
5/4
an is absolutely convergent and so convergent.
n=1 n
83
K72
akn 0 so there exists an N with an 1 for n N . Thus akn |ak+1
n |
for n N and the result follows by comparison.
Set f (n) = log(n + 1)n and observe that, if k 2
3N
X
j=1
akj
N
X
j=1
(Use comparison and the fact that n (log n)k for n sufficiently large.)
84
K73
(i) Write
Sn =
n
X
1
j bj .
j=1
j=1
n1
X
j=1
(j j+1 )Sj + n Sn
so that
n
n1
X X
bj
(j+1 j )kSj k + n kSn k
j=1
j=1
n1
X
j=1
n
X
j=1
bj k
1
n
1
n
N
1
X
j=1
N
1
X
j=1
kbj k +
Pn
1
n k
j=N
n
X
j=N
1
j bj k for
bj k
kbj k2 2
85
K74
Consider the set An of integers between 10n and 10n+1 1 inclusive
whos decimal
contain the integer 9. We observe
Sn expansion does not n+1
that {0} j=0 Aj contains (9/10)
10n+1 = 9n+1 elements. Thus
An contains at most 9n+1 elements and
X
j 1 10n 9n+1 .
jAn
The partial sums of the given series thus can not exceed
9n+1 = 90 and the series converges.
n=0
10n
86
K75
(i) Observe that 1/x 1/(n + 1) for n x n + 1 and integrate.
Tn Tn+1
=
n+1
n+1
n
1
dx.
x
Z r+1
n1
X
1
1
Tn
dx .
r
x
r
r=1
Decreasing sequence bounded below tends to a limit.
(v) Given l we can choose integers pn , qn 1 such that pn+1 > 2pn ,
qn+1 > 2qn , pn /qn is strictly decreasing and log 2+(1/2) log(pn /qn ) l.
At the n-th stage, if the ratio of positive terms to negative exceeds
pn /qn use pn positive terms followed by qn + 1 terms if the ratio is less
than pn /qn , use pn + 1 positive terms followed by qn negative terms,
otherwise usepn positive terms followed by qn negative terms. Switch
from the n-th stage to the n + 1-th stage when the size of each unused
terms is less than (pn+1 + qn+1 + 1)1 2n and the ratio of positive terms
to negative differs from pn /qn by at most 2n .
87
K76
P
(ii) Observe that, if
j=0 bj converges, we can find an N () such
PN
that | j=M ) bj | for N M N () so
j
b
x
xM
j
j=M
for M N ().
M 1
X
X
X X
bj xj
bj
|bj xj bj | +
bj +
bj xj
j=0
j=0
j=0
M
1
X
j=0
j=M
j=M
|bj xj bj | + 2 2
X
X
X
j
j
aj x
c j xj .
bj x =
j=0
Now let x 1.
j=0
j=0
88
K77
P
P
n m
converges for |x| = |y| = , we have
(a) Since
m=0 cn,m x y
n=0
n+m
n m
|cn,m |
= |cn,m x y | 0 for n, m so there exists a K with
|cn,m | n+m < K for all n, m 0. Set = 1 .
P
P
P
(b) (i) nN, mN |xn y m | = nN |x|n mN |y|m .
E is the square (1, 1) (1, 1).
n+m n m
P
PP
r
|x y | = N
(ii)
r=0 (|x| + |y|)
n+mN
n
E = {0}.
P
P
(vi) nN |xn y n | = nN |xy|n .
(c) Observe that |cnm xn y m | |cnm xn0 y0m | when |x| |x0 |, |y| |y0 |.
89
K78
(i) Choose Nj () so that |aj aj,n | < /M for n Nj () and set
N (, M ) = max1jM Nj ().
Observe that
X
j=1
aj
M
X
j=1
aj
M
X
aj,n
j=1
for all M .
P
(ii) If
j=1 aj converges with value A, then given > 0 we can find
P
an M such that M
j=1 aj > A . By (i) we can find N (, M ) such that
X
j=1
and so
X
j=1
aj,n
aj
X
j=1
M
X
j=1
aj
aj,n A 2
P
for all n N (, M ). Thus
j=1 aj,n A.
P
If
j=1 aj fails to converge, then given any K > 0, we can find
PM
an M such that
j=1 aj > 2K. But we can find an N such that
PM
PM
P
aj,n j=1 aj K for n N and so
j=1 P
j=1 aj,n K for n N .
(iii) Yes we can drop the condition. Consider bj,n = aj,n aj,1 instead.
90
K79*
No comments
91
K80
It is possible to do this by fiddling with logarithms but P
I prefer to
nz
copy the proof P
of the radius of convergence. Show that if
n=0 bn e
converges then n=0 bn enw converges absolutely when <w < <z.
P
nz
Next observe that the sum
converges everywhere (set
n=0 bn e
X =P
) or nowhere (set X =P
) or we can find z1 and z2 such
nz1
nz2
that
b
e
converges
and
diverges. Explain why
n=0 n
n=0 bn e
this means that
X
bn enz converges}
X = sup{<z :
n=0
inz
inz
converges
and
converges
if
both
converge,
c
(e
+
e
)
and
n=0 n
P
thus n=0 cn cos nz converges if =z < Y and diverges if =z > Y .
92
K81
(i) By integration by parts,
In+1 = [ sin
=n
x cos x]/2
0
/2
0
+n
/2
sinn1 x cos2 x dx
0
= nIn1 nIn+1 .
(ii) Observe that sinn1 x sinn x sinn+1 x for x [0, /2]. Integrating gives In+1 In In1 so
In+1
In
n
=
1.
n+1
In1
In1
so In /In1 1 as n .
(iii) I0 = /2, I1 = 1
(2n 1)(2n 3) . . . 1
2n(2n 2) . . . 2
I2n =
, andI2n+1 =
.
2n(2n 2) . . . 2 2
(2n + 1)(2n 1) . . . 1
Thus
n
Y
4k 2 2
I2n+1
=
1
2
4k 1
I2n
k=1
and the Wallis formula follows.
93
K82
(i) Power series or set f (x) = exp x x and observe f (0) = 0,
f 0 (x) 0 for all x 0.
(iii) Observe that
m
n
n
m
Y
Y
Y
Y
(1 + aj ) 1
(1 + aj ) = (1 + aj )
(1 + aj )
j=n+1
j=1
j=1
j=1
!
m
n
Y
Y
(1 + |aj |) 1 .
(1 + |aj |)
j=n+1
j=1
an
2|an |.
|bn | =
1 + an
P
Thus
n=1 bn converges absolutely.
n
n
Y
Y
Y
Y
1=
(1 + aj ) (1 + bj )
(1 + aj ) (1 + bj )
j=1
so
j=1 (1
+ aj )
j=1
j=1 (1
j=1
j=1
+ bj ) = 1.
(1 + aj /Rj ) =
j=1
N
Y
Rj1
j=1
Rj
R0
0
RN
94
K83
(i) Left to reader.
(ii) Let an = n , an,N = n if all the prime factors of N lie among
the first N primes, an,N = 0 otherwise. Observe that
0 an,N an,N +1 an ,
P
an as N and that
n=1 an converges.
that an,N
(iii) If
1
K
1 p1
1
K
1 p
pP, pN
pP, pN
pP
and
X
1
K
n
n=1
for all
1
n=1 n
1
K
1 p
N
X
1
K
n
n=1
N
X
1
K
n
n=1
which is absurd.
rE,r2N
for all N .
N 1
N 1
N
1
X
1 X X 1 X j
A
=
2 M (j) A
2n(1) =
r
r
1 21
j=0 rE
j=0
j=0
j
95
K84
Q
Suppose that Pn > 1 for all n N . Then Pm = PN m
j=N (1 an ).
P
If j=1 an diverges then (see K82) Pm 0 which is impossible by the
P
definition of N . Thus
j=1 an converges and Pm tends to a limit. A
similar argument applies if Pn 1 for all n N .
Thus we need only consider the case Pn 1 changes sign (or is zero)
infinitely often. But an 0 and if (A) PN 1 1 0 PN 1 or
PN 1 1 0 PN 1 and (B) 0 an for n N 1 then we
have 1 + Pn 1 for n N . Thus Pn 1.
P
If
j=1 an diverges we can say nothing about l. (If l 1 consider
a1 = l 1, an = 0 for n 2. If l 1 consider a1 = 1, a2 = 1 l/2,
an = 0 for n 3.)
96
K85
Let an = z n , an,N = z n if 0 n 2N +1 1 and an,N = 0, otherwise.
Observe that 0 |an,N |an |, and that an,N an as N . By
dominated convergence (Lemma 5.25)
N
Y
j=1
2j
(1 + z ) =
X
n=0
an,N
X
n=0
an = (1 z)1 .
97
K86
Observe that
2 cot 2 =
cos2 sin2
= cot + tan
cos sin
as n .
1 2n
1
1
cot
=
cos 2n
n
n
n
2
2
sin 2
98
K87
Very similar to K86. To obtain the formula for un , observe that
2 cos2 = 1 + cos 2 and cos(/4) = 21/2 .
99
K88
The square of a rational is rational.
(ii) Observe, by induction or Leibnizs rule, that f (k) is the sum of
powers of x with integral coefficients plus terms of the form
xr (1 x)s
M
(min(r, s))!
with M an integer and r, s 1. Thus f (k) (0) and f (k) (1) are always
integers. Since bn 2nr is an integer, G(0) + G(1) is always an integer.
as n .
an
0
n!
100
K89
(a)(i) Given c C we can find b B with g(b) = c and a A with
f (a) = b.
(ii) g f (a) = g f (a0 ) implies f (a) = f (a0 ) and so a = a0 .
101
K90
If f satisfies the conditions of the first sentence of the second paragraph, we can write f = g J where J(x) = x and g satisfies the
conditions of the first paragraph. Thus
for all x E.
102
K91
(i) Suppose that f is continuous. If f is not strictly monotonic we
can find x1 < x2 < x2 so that f (x1 ), f (x2 ) and f (x3 ) are not a strictly
increasing or a strictly decreasing sequence. Without loss of generality
(or by considering f (1 x) and 1 f (x) if necessary), we may suppose
f (x3 ) f (x1 ) f (x2 ). If either of the inequalities are equalities then
f is not injective. Thus we may suppose f (x3 ) > f (x1 ) > f (x2 ). But
by the intermediate value theorem we can find c (x2 , x3 ) so that
f (c) = f (x1 ) and f is not injective.
Suppose f is strictly monotonic. Without loss of generality we suppose f increasing. We shall show that f is continuous at t (0, 1)
(the cases t = 0 and t = 1 are handled similarly). Observe that
f (1) > f (t) > f (0). Let > 0. Set 0 = min(1 f (t), f (t), )/2.
We can find s1 and s2 with f (s1 ) = f (t) 0 f (s2 ) = f (t) + 0 . If
O/ta = min(t s1 , t + s2 ), then |s t| > implies s2 > s > s1 so
f (s2 ) > f (s) > f (s1 ) and |f (s) f (t)| 0 < .
(ii) The same proof as in (i) shows that, if g is continuous and injective, it must be strictly monotonic. The example of g(t) = t/2 for
t 1/2, g(t) = t/2 + 1/2 shows that the converse is false.
(iii) The same proof as in (i) shows that, if g is strictly monotonic
and surjective, it must be continuous. The example g(t) = sin t shows
that the converse is false.
103
K92
Write f (x) = (log x)/x. By looking at f 0 we see that f is strictly
increasing from to e1 on the interval (0, e], has a maximum value
of e1 at e and is strictly decreasing from e1 to 0 on the interval
(0, ). The function f has a unique zero at 1,
We have xy = y x if and only if f (x) = f (y). The set
{(x, y) : xy = y x , x, y > 0}
is thus the union of the straight line {(x, x) : x > 0} and a curve with
reflection symmetry in that line, having asymptotes x = 1 and y = 1
and passing through (e, e).
If f (x) = f (y) and x < y, then x e so we need only examine the
cases m 2 to see that the integer solutions of nm = mn are n = m,
(m, n) = (2, 4) and (m, n) = (4, 2).
Since f () < f (e) we have e > e .
104
K93
Observe that
xy =
1
4
(x + y)2 (x y)2 .
105
K94
(i) Choose with > > 1. If we set f (x) = x x , then (recalling
that x = exp(x log ))
f 0 (x) = x1 x + x (log ) x = x1 x ( + x) > 0
for x sufficiently large (x x0 , say). Thus
(ii) Set
(
1/2
1/2 )
1
1
1
, f3 (x) =
, f2 (x) = exp
f4 (x) = exp
log
x
x
x
and
f1 (x) =
1
log
x
Observe that log f2 (x)/ log f1 (x) = (log y)1/2 /(3 log log y) as
x .
106
K95
(ii) If Q has a root 6= 0 of order n 1 then differentiating the given
equation n 1 times and setting z = we get P ()Q(n) () = 0 so
P () = 0 and P and Q have a common factor. Thus Q(z) = Az N for
some N 1, A 6= 0 and it is easy to check that this too is impossible.
(iv) If the degree of P is no smaller than that of Q then
P (x)
log x .
Q(x)
If the degree of P is smaller than that of Q then
P (x)
log x 0.
Q(x)
If the degree of P is no smaller than that of Q then
P (x)
log x .
Q(x)
107
K96
Set f (x) = xlog(1+x). Then f (0) = 0 and f 0 (x) = 1(1+x)1 0
for 0 < x and f 0 (x) 0 for x > 0. so f (x) 0 and x log(1 x)
for 0 x 1.
Set g(x) = log(1 + x) x + x2 . Then g(0) = 0 and
kn
kn
Y
r
r X
log 1 2
1 2 =
n
n
r=1
r=1
kn
kn
kn
kn
X
X
Y
X
r
r
r
r2
log
.
1
2
2
2
4
n
n
n
n
r=1
r=1
r=1
r=1
But
and
as n so
kn
X
1
k2
kn(kn + 1)
k
r
k
+
=
=
n2
2n2
2
n
2
r=1
kn
X
r2
(kn)3
k
=
0
4
4
n
n
n
r=1
kn
Y
k2
r
log
1 2
n
2
r=1
108
K97
Suppose that x 1. Then xn 1 and
n n
(1 + n )(1 + n )
= 1+
.
1 + n + n
1 + n + n
Since 2n n tends to a limit we can find an A such that |n A2n .
Similarly we can find a B such that |n B2n . Thus we can find a
C such that
n n
2n
1 + n + n C2 .
It follows that
2n n
(1 + )(1 + ) 2n
X
2
n
n
C r 22nr
1
r=1 r
1 + n + n
n
2
X
2nr C r 22nr
r=1
2
X
2nr C r 2nr
r=1
as n .
C2n
0
1 C2n
a2 ) = 2n c2
n 1
(b a)
2n ((1 + n )(1 + n ) (1 + n + n ) 0
109
that is to say
2n ((xy)2
so writing z = xy
and
(xn + yn 1)) 0
Result (ii) shows that we need only prove (iii), (iv), (v) and (vi) at
the point x = 1. To do this observe that a simple version of Taylors
theorem
|f (1 + ) f (1) f 0 (1)| sup |f 00 (1 + )|| 2 |
applied to f (x) = x
2n
||||
yields
n
|(1 + )2 1 2n | 2n 4 ||2
for all || 101 and n 3. (We can do better but we do not need
to.) Thus if || 101 and we set x = 1 + we obtain
for n 3 so that
| log(1 + ) | 4||2 .
110
K98
If
f (x + )
f (x)
1/
so log f is differentiable and, by the chain rule, f = exp log f is differentiable at x. Since
d
f 0 (x)
log l =
log f (x) =
dx
f (x)
we have
f 0 (x)
l = exp
.
f (x)
111
K99
If
1
x1
x a
xa
tends to a limit as x a then (x a)f (x) 0. Since
f (x) =
as x a this gives
x a
a1
xa
1=0
so = .
If = then f (x) = F (x)/G(x) with
F (x) = x ax1 x + a
112
K100
We only consider x when x is real and we define it as exp( log x)
where log is the real logarithm.
113
K101
(iii) Set f (x) = log g(x) and apply (i) to get g(x) = xb for some real
b. Easy to see that this is a solution.
(iv) Observe that, if x > 0, g(x) = g(x1/2 )2 > 0. Thus, by (iii)
g(x) = bx for all x > 0 and some b > 0.
Now g(1)2 = g(1) = 1 so g(1) = 1 and either g(x) = b|x| for all
x or g(x) = sgn(x)b|x| for all x. (Recall sgn x = 1 if x > 0, sgn x = 1
if x < 0.) It is easy to see that these are solutions.
114
K102
(i) We know that
(t/2)2 = t
but the equation (exp is)2 = exp iu, where u R, has two real solutions
in s with < s (the solutions being s/2 and one of s/2 ).
As a specific example, if (t) = exp 2it, then (3/4) = /2 but
(3/8) = 3/4.
However since is continuous we can find a > 0 such that |(t)
1| 1001 for |t| . If |t| then |(t)| /4 and |(t/2)| /4
so (t/2) = (t)/ (since |(t/2) | > /4).
(iii) The same ideas show that the continuous homomorphisms are
precisely the maps 7 n for some integer n.
115
K103
(i) f (0) = 2f (0) so f (0) = 0.
k
(ii) Set F (2k p1
j ) = 2 j whenever j and k are integers with j 1
and F (t) = 0 otherwise. (By the uniqueness of factorisation F , is well
defined.)
0.
1
Since F (p1
j ) and pj 0 as j , F is not continuous at
(iii)
f (t2n ) f (0)
f 0 (0)t
t2n
as n so f (t) = f 0 (0)t for all t 6= 0 and so for all t.
f (t) = 2n f (t2n ) = t
k 1
(iv) Set F (2k p1
j ) = 2 pj whenever j and k are integers with j 1
and F (t) = 0 otherwise. Since |F (t) F (0)| = |f (t)| |t|, we have F
continuous at 0. Since
F (p1
F (21/2 j 1 ) F (0)
j ) F (0)
0
=
1
1
but
21/2 j 1
p1
j
as j , F is not differentiable at 0.
116
K104
Consider the case x 6= 0. Observe that
1
= 1 + ()||
1/2
(1 + )
2
with () 0 as 0. (Use differentiation or Taylor series.) Observe
also that |x h| kxkkhk. Thus
x+h
f (x + h) =
(kxk2 + 2x h + khk2 )1/2
x+h
=
1/2
2x h + khk2
x+h
+ 1 (h)khk
1
=
kxk
2kxk2
h
(x h)x
= f (x)
+ 2 (h)khk
kxk
kxk3
where kj (h)k 0 as khk 0. Thus f is differentiable at x and
Df (x)h =
(x h)x
h
.
kxk
kxk3
kxk2
hx
xh
= 0.
(Df (x)h) x =
kxk
kxk3
Since
kf (x) f (0)k = 1 9 0
117
K105
Without loss of generality take z0 = 0.
(i)(ii) If f is complex differentiable at 0, then writing f 0 (0) =
A + Bi with A and B real, and taking h and k real
f (h + ik) = f (0) + (A + Bi)(h + ik) + (h + ik)|h + ik|
with (h + ik) 0 as |h + ik| 0. Taking real and imaginary parts,
1 (h, k)
Ah Bk
u(0.0)
u(h, k)
(h[ 2 + k 2 )1/2)
+
+
=
2 (h, k)
Ak + Bh
v(0, 0)
v(h, k)
with
1 (h, k)
2 (h, k) 0
A B
B A
If A = B = 0 take = 0 and = 0. Otherwise take = (A2 + B 2 )1/2
and such that cos = A, sin = B.
(ii)(iii)(iv) Just a matter of correct interpretation.
(iv)(i) Carefully reverse the first proof.
118
K106
(i) Observe that
g(x + h) = f (x + h, c x h)
= f (x, c x)f,1 (x, c x)h + f,2 (x, c x)(h) + (h, h)(h2 + h2 )1/2
x
u(x) =
.
cx
1
.
1
1
f,1 (x, c x) f,2 (x, x c)
= (f,1 (x, c x) f,2 (x, c x)).
1
119
K107
(i) Differentiating with respect to and applying the chain rule, we
have
m
X
1
f (x) =
xj f,j (x).
j=1
j=1
Thus
d
( v()) = 0
d
and by the constant value theorem
v() = v(1)
and f (x) = f (x).
120
K108
This is really just a question for thinking about.
(i) We seek to maximise the (square root of)
2
a b
cos
s=1
m
X
x2s = 1
s=1
121
K109
If e1 , e2 , . . . , em is the standard basis (or any orthonormal basis)
and (aij ) is the associated matrix of , then
* n
+
m
X
X
aij =
arj er , ei
arj her , ei i =
r=1
r=1
Conversely if aij = aji for all i, j, then essentially the same calculation
shows that
hej , ei i = hei , ej i
and so by linearity
! m
* n
+
X
X
xr er ,
ys e s
r=1
s=1
=
=
=
m X
m
X
r=1 s=1
m X
m
X
r=1 s=1
* n
X
xr ys her , es i
xr ys her , es i
xr er ,
r=1
m
X
s=1
ys e s
!+
X
X
X
2 2
2
x2j
xj j 1
xj j u j =
xj u j =
j=1
j=1
j=1
j=1
122
K110
(ii) Let x =
Pn
j=1
xj ej . Then
yk =
n
X
j=1
unless x1 = 0.
kj xj
(
k
2 1/2
r=1 (r xr ) )
Pn
e1
123
K111
(ii) hx, xi = hx, xi = kxk2 .
Thus
k kkxk2 = kxkk xk |hx, xi| = kxk2 kk2 kkxk2
Thus 0.
1 1
.
(vii) We have A A =
1 5
A has eigenvalues 1 and 2. A A has eigenvalues 4 51/2 . kAk =
(4 + 51/2 )1/2 .
124
K112
Follow the proof of Rolles theorem. If kc|| < 1 and f has a maximum
at c then f,1 (c) = f,2 (c) = 0 so Df (c) is the zero map.
g(x, y) = x2 will do. Observe that cos2 a cos b sin is not
identically zero (consider = /2 and other values of ).
We can not hope to have a Rolles theorem proof in higher dimensions.
125
K113
Since (x, y) 7 x y, (x, y) 7 xy, x 7 x1 and x 7 f (x) are all
continuous we see that F which can be obtained by composition of such
functions is continuous except perhaps at points (x, x).
If F is continuous at (x, x) then F (x + h, x) must tend to the limit
F (x, x) as h 0 so f is differentiable and F (x, x) = f 0 (x). Since
F (x + h, x + h) F (x, x) as h 0, f 0 must be continuous.
Conversely if f 0 exists and is continuous then, setting F (x, x) =
f 0 (x), we saw above that F is continuous at all points (x, y) with x 6= y.
If h 6= k the mean value theorem gives
f (x + h) f (x + k)
F (x + h, x + k) F (x, x) =
f 0 (x)
xy
0
= f x + (h,k h + (1 h,k )k) f 0 (x)
for some h,k with 0 < h,k < 1. But it is also true that, if we set
h,h = 1/2,
F (x + h, x + h) F (x, x) = f 0 (x + h) f 0 (x)
Thus
as (h2 + k 2 )1/2 0.
126
K114
If k 6= 0, then we can find an X > 0 such that |f 0 (t) k| < |k|/4 and
so |f 0 (t)| > 3|k|/4 for t X. Thus, using the mean value inequality,
+
x
xX
x
x
x x
xX
3|k| x X f (X)
4 x x
|k|
|k| |k|
2
4
4
1
whenever x max(3X, (4|f (X)|) ).
g(x) = x2 sin x4 will do.
127
K115
The local form of Taylors theorem
f 00 (x) 2
h + (h)h2
2
applied to log (or other arguments) show us that
f (x + h) = f (x) + f 0 (x)h +
log(1 h) = h
h2
+ (h)h2
2
with (h) 0 as h 0.
Thus since qn we can find an N such that
128
K116
If a = b = 0 maximum c.
If a = 0 maximum c if b 0, 10b + c if b 0
If a > 0 then maximum c if b/2a 5, 100a + 10b + c if b/2a 5.
129
K117
(i) Take x1 = 1, xj = 0 otherwise.
(ii) To see that B is positive definite if A is, take
1
1
x1 = (a1
11 a12 x2 + a11 a13 x3 + + a11 a1n xn ).
130
K118
(i) One way of seeing that the j are continuous is to solve the
characteristic equation. Now use the intermediate value theorem.
cos t sin t
will do.
(iii) B(t) =
sin t cos t
131
K119
The reduction to Figure K2 can not be rigorous unless we state
clearly what paths are allowed. However if we have a path which is not
symmetric under reflection in the two axes of symmetry joining mid
points of opposite sides then by considering the reflected path we can
reconstruct a shorter symmetric path.
[A more convincing route involves the observation that the shortest
path system for three points M , N , P is three paths M Z, N Z, P Z
making angle 2/3 to each other provided the largest angle of the
triangle M N P is less than 2/3 and consists of the two shortest sides
otherwise. (See Courant and Robbins What is Mathematics?.)]
One we have the diagram we see that the total path length is
f () = 2a sec + (a a tan )
where a is the length of the side of the square and is a base angle of
the isosceles triangles.
1 2 sin
f 0 () =
cos2
so f increases on [0, /6] and decreases on [/6, /4] so the best angle is
/6 (check that this gives an arrangement consistent with the statement
in square brackets). IN PARTICULAR = /4 is not best.
There are two such road patterns the other just being a rotation
through /4
132
K120
The time taken is
h x (1 + x2 )1/2
+
u
v
for 0 x h. We observe that
x
1
f 0 (x) = +
u v((1 + x2 )1/2
so f 0 is negative and f is decreasing for 0 x with 0 x (1
v 2 /u2 )1/2 and f 0 is positive and f is increasing for x (1v 2 /u2 )1/2 .
Thus if (1 v 2 /u2 )1/2 < h (i.e. (1 v 2 /u2 )h2 > 1) he should take
x = (1 v 2 /u2 )1/2 . However, if (1 v 2 /u2 )h2 < 1 his best course if
0 x h is to take x = h. Since choosing x outside this range is
clearly worse he should take x = h.
f (x) =
If h is large his initial run is almost directly towards the tree and his
greater land speed makes this desirable. If h is small his run is almost
perpendicular to the direction of the tree and does him very little good.
133
K121
We wish to find the maxima and minima of f (x, y) = y 2 13 x3 + ax
in the region x2 + y 2 1. We first examine the region x2 + y 2 < 1.
The stationary points are given by
0 = f,1 (x, y) = x2 + a, 0 = f,2 (x, y) = 2y.
2x 0
0
2
1/2
x = 1 + (1 + a2 ) y = 1 (1 + (1 + a2 ))2
) and decreases as x runs
2
from from 1 (1 + a ) to 1. Thus g has minima at x = 1 and 1 and
(by evaluating g at these points) a global minimum at 1 if a 1/3.
Since a 1/3 we know that this gives a global minimum for f at
(0, 1). If a 1/3, g has a global minimum at 1 and if 0 a 1/3
we know that it gives a global minimum for f at (0, 1).
If 1 > a > 0 we observe that we know that the global minimum
occurs when y = 0 and by looking at h(x) = f (x, 0) = 13 x3 + ax we
see that it occurs at (a1/2 , 0).
134
K122
The composition of differentiable functions is differentiable and the
product of differentiable functions (with values in R) are differentiable
so since the maps (x, y) 7 r and (x, y) 7 are differentiable except
at 0, f is. (To see that (x, y) 7 r is differentiable observe that r =
(x2 + y 2 )1/2 and use basic theorems again. To see that (x, y) 7 is
differentiable on
S = {(x, y) : |y| > 9x}
so
If f is differentiable at 0 then
f (r cos , r sin ) f (0, 0)
cos f,1 (0, 0) + sin f,2 (0, 0)
r
g() = g(0) cos + g(/2) sin
ie
g() = A sin + B cos
for some constants A and B so
f (x, y) = Ax + By.
The necessary condition is sufficient by inspection.
135
K123
(i) Observe that.
f (t, t) = t2 ( a2 t)( b2 t)
0 0
f,11 (0) f,12 (0)
=
0 2
f,21 (0) f,22 (0)
which is singular.
136
K124
(i) The second sentence holds by Exercise 8.22 which says that if f
is Riemann integrable so is |f |.
(ii) First sentence false. Take f = g = F .
Second sentence false. Take
f (x) = F (x), g(x) = 0
f (x) = 0,
for x 1/2.
g(x) = F (x)
f (x) = 0,
for x 1/2.
g(x) = F (x)
f (x) = 0,
for x 1/2.
g(x) = 1
and functions like f (1 x) and g(x) we see that the product of two
functions that are not Riemann integrable may or may not be Riemann
integrable and that the product of one that is Riemann integrable with
one that is not may or may not be Riemann integrable.
137
K125
(i) Consider the dissection DN = {r/N : 0 r N } where N M 2 .
At most 2M 2 intervals [(r 1)/N, R/N ] contain points of the form p/q
with p and q coprime and 1 q M . Thus
0 = s(f, DN ) S(f, DN ) 2M 2 N 1 + M 1 M 1
as
R 1 N . Since M is arbitrary f is Riemann integrable with
f (t) dt = 0.
0
=1
p/q x
1/p
but choosing xn irrational such that xn x we have
(f (xn ) f (x)
= 0 0.
xn x
138
K126
Observe that, if 0 u < v 1, then un v n n(y x) (by the mean
value theorem or algebra). Write fn (x) = f (xn ). If D is a dissection
write
Dn = {y 1/n : y D}.
Suppose that |f (x)| K. Given > 0, we can find a 1 > > 0 such
that |f (x) f (0)| /2 for 0 x . Since 1/n 1 as n we
can find an N such that 1 1/n (K + |k|)1 /2 for n N .
If N n then |fn (x) k| /2 for x [0, 1/n ] and |fn (x)| K for
x [ 1/n , 1] so
Z 1/n
Z 1
0
1/n
so that
f (x) dx k .
139
K127
Thus
Z b
Z
|f (s)|
a
(a+b)/2
(b+a)/2
+
(a+b)/2
2
K (b a)/2
(a+b)/2+
+
(a+b)/2
b
(a+b)/2+
2
+ (K )( b a)/2
|f (s)| ds
(b a)/2
as required.
f (x) = C (x (a + b)/2)2
for a x (a + b)/2 ,
140
K128
(v) Observe P
that once we know that f is Riemann integrable we
n1
ba
know that n
j=0 f (a + j(b a)/n) converges but the fact that
Pn1
ba
j=0 f (a + j(b a)/n) converges does not imply that f is Rien
mann integrable.
141
K129
For (A) note that, if Fj and Gj are positive bounded Riemann integrable functions with F1 G1 = F2 G2 then F1 + G2 = F2 + G1
and
Z b
Z b
Z b
F1 (t) dt +
G2 (t) dt =
F1 (t) + G2 (t) dt
a
a
a
Z b
=
F2 (t) + G1 (t) dt
a
Z b
Z b
G1 (t) dt
F2 (t) dt +
=
a
and so
b
a
F1 (t) dt
G1 (t) dt =
a
b
a
F2 (t) dt
G2 (t) dt.
a
The problems with (C) begin (I think) when we try to prove that if f
and g are Riemann integrable so is f + g. We also get problems (which,
I think have the same cause) when we try to show that anything which
is Riemann integrable under the old definition are Riemann integrable
under the new. One way forward is to show that anything which is
(A) integrable is (C) integrable. To do this we can first prove that a
bounded positive function f is Riemann integrable if and only if given
any > 0 we can find a dissection
D = {x0 , x1 , . . . , xN }
142
K130
(iii) Everything works well until we try to define the upper and lower
integrals but the set of s(D, f ) has no supremum in Q and S(D, f ) has
no infimum.
143
K131
(i) Since F is continuous on a closed bounded interval the infimum
and supremum are attained at and say. Now use the intermediate
value theorem.
(ii) Observe that
sup f (s)w(t) f (t)w(t)
s[a,b]
w(t) dt
f (t)w(t) dt
a
s[a,b]
144
K132
The graph of f splits the rectangle [a, b] [f (a), f (b)] into two parts
whose areas correspond to the given integrals.
Consider a dissection of [a, b]
D = {x0 , x1 , . . . , xN }
Similarly
so
g(s) ds
f (X)
XF (X) + (Y F (X))X = Y X
1 1
, g(y) = y q
1 1
145
K133
Since f is continuous at c, we can find a > 0 such that |f (c)
f (x)| < /4 when |x c| < . Thus if [, ] (c , c + )
I[,]
(f ) I[,] (f ) ( )/2.
146
K134
Suppose [a, b]. Then given 1 > > 0 we can find a > 0 such
that
|f (s) f ()|, |f (g(s)) f (g())|, |g 0 (s) g 0 ()| <
g()
g()
g() g()
sup |f (s) f (g())|
s[g(),g()]
t[,]
g() g()
sup |g 0 (t)|( )
t[,]
(|g 0 ()| + )( )
We also have, by a simpler argument
Z
Thus
Z
Z g()
f (s) ds
f (g(t)) dt (1 + |g 0 ()| + )( )
g()
(2 + |g 0 ()|)( ).
147
K135*
No comments.
148
K136
The formula
sup
x[xj1 ,xj ]
inf
x[xj1 ,xj ]
f 0 (t)(xj xj1 )
149
K137*
No comments.
150
K138
Let N 4. Take a dissection DN consisting of the points 1, 0, 1,
1 N 3 and rN 1 N 3 with 1 r N 1. Then
0 = s(DN , f ) S(DN , f ) N 1 + N 2N 3 = N 1 + 2N 2 0
R1
as N . Thus f is Riemann integrable and 1 f (t) dt = 0.
Similarly F (t) = 0 for all t. Thus F 0 (0) exists with value 0 = f (0).
Since f (1/n) = 1 9 0, f is not continuous at 0.
151
K139
(i)
n
X
r=1
n
X
(r + 1)r(r 1) r(r 1)(r 2)
r(r 1) =
r=1
n
X
r=
r=1
n
X
r =
r=1
n
X
(r + 1)r r(r 1)
r=1
n
X
r(r 1) +
r=1
n
X
r=1
r=
(n + 1)n(n 1)
.
3
(n + 1)n
2
n(n + 1)(2n + 1)
.
6
(ii)
n
X
r=1
n3 =
n2 (n + 1)2
.
4
(iii) If f (x) = xm
S(D, f ) =
and
s(D, f ) =
n
X
1
Pm (n)
1
r
+ m+1
n1 ( )m =
n
m+1
n
m+1
r=1
n1
X
r
1
Pm (n 1)
1
n1 ( )m =
(1 n1 )m+1 +
n
m+1
nm+1
m+1
r=0
so I f = I f = (m + 1)1 .
(iv) If f (x) = xm and F (x) = xm+1 /(m + 1) then F 0 = f and
Z b
f (x) dx = F (b) F (a).
a
152
K140
(v) We have |g(t) g(1)(t + 1)| K(t 1) so
Z 1
Z 1
g(t) dt g(1) K
(t 1) dt = 2K.
Scaling, we get
Z
arh
a(r1)h
K(b a)h.
f
(t)
dt
S
(f
)
h
153
K141
(ii) Let s = yt.
(viii) We have
tn+1
1
(1 + t + + tn ) =
.
1t
1t
So, integrating both sides from 0 to x, we have
Z x n+1
x2
t
xn+1
log(1 x) x +
=
+ +
dt 0
2
n+1
0 1t
X
xr+1
log(1 x) =
r+1
r=1
for |x| < 1. [The result also holds if x = 1 but the argument above
does not prove this without extra work.]
(ix) By (viii)
log
1+x
1x
= log(1 + x) log(1 x)
=
X
(1)r xr+1
r=1
=2
r=1
r+1
x2r+1
2r + 1
X
xr+1
r+1
r=1
1+x
1x
y1
y+1
so |x| < 1 and substitution in the formula of (ix) gives the desired
result.
x=
154
K142
Power series can be multiplied term by term within their radius of
convergence. We observe that
n1
n1
X
(1)r (1)nr
(1)n X 1
1
(1)n 2Sn1
=
+
.
=
r
nr
n r=1 r n r
n
r=1
The result is valid for |x| < 1.
155
K143
(i) Write
=
and observe that 1 > > 0, so
x3 x 2
x3 x 1
so that
and
.
y3 y 1
y2 y 1
y3 y 2
y3 y 1
(ii) By (i), if h > 0, then
f (c) f (d)
f (c + h) f (c)
h
cd
where d is fixed with c > d > a. Since a non-empty set bounded below
has an infimum, f (c+) exists.
By definition, given > 0 we can find a > 0 such that
f (c + ) f (c)
.
If h > > 0 then (looking at the last paragraph of our discussion of (i))
f (c+) +
so
f (c+) +
f (c + h) f (c)
f (c + ) f (c)
f (c+)
h
f (c + h) f (c)
f (c+)
h
as h 0+.
(iii) We can find a > 0 such that, if > h > 0, then
f (c+) + 1
f (c + h) f (c)
f (c+)
h
and so
|f (c + h) f (c)| (|f (c+)| + 1)|h| 0
156
We know that there exists a B such that f (x) f (c) B(x c) and
Now integrate.
whence
n
!
n
n
n
X
X
X
X
j f (xj ) f
j xj =
j (f (xj ) f (c))
j (xj c) = 0.
j=1
j=1
j=1
j=1
157
To extend the result to (a, b), observe that the result just proved
shows that that if a < a + n1 < c < b n1 < b we can find a Bn
such that
1
f (x) f (c) Bn (x c)
158
K144
(ii) Suppose that < . If 0 < h, k < ( )/2 then
f ( h) f ( + h
f ( + k) f ()
f () f ( h)
h
hk
k
so
f () f (+).
Thus
N
X
j=1
f (cj +) f (cj )
= f (cN +)
N
X
j=1
f (cN +) f (c1 )
f (+) f ().
Ej
j=1
is countable. If c
/ E, then f (c) = f (c+) = f (c), say, and
f (c + h) f (c)
f (c)
h
as h 0.
(iv) Enumerate the rational points in (1, 1) as q1 , q2 , . . . . If we set
fn (x) = n2 |x qn |
then fn is convex and
0 fn (x) 2n2
P
Thus, using the comparison test,
n=1 fn (x) converges everywhere to
f (x), say.
159
=
h
h
h
fn (qn + h) + f (qn h) 2f (qn )
h
2
= 2n 9 0
as h 0+, so f is not differentiable at qn
160
K145
(i) Given > 0, we can find a > 0 such that |f (x) f (c)| < for
|x c| < . If [an , bn ] (c , c + ), then
Z
Z
1
f (t) dt f (c) =
f (t) f (c) dt
|In |
|In | In
In
Z
1
|f (t) f (c)| dt
|In | In
Z
1
dt =
|In | In
(ii) The following is one generalisation (interpreting the notation
appropriately). Let B(c, r) be the ball centre c then (provided the
integrals exist)
Z
1
f (t) dV (t) f (c)
Vol B(c, r) B(c,r)
as r 0+ whenever f is continuous at c.
161
K146*
No comments.
162
K147
(i) The range of integration is not fixed.
(ii) If x is fixed the fundamental theorem of the calculus shows that
F,2 exists and
F,2 (x, y) = g(x, y)
so F,2 is continuous.
Theorem 8.57 tells us that, if x is fixed we may differentiate under
the integral to obtain
Z y
F,1 (x, y) =
g,1 (x, y) dx.
0
sup
(s,t),(s0 ,t)K,|ss0 ||xx
+ |y y0 |
0|
sup
(s,t),(s0 ,t)K
sup
|t|max(|y|,|y0 |)
163
K148
Observe that differentiating under the integral, using the symmetry
of partial derivatives and the fundamental theorem of the calculus
E 0 (t) =
=2
1
0
u
(x, t)
t
u
(x, t)
x
2 !
dx
2
2 !
Z 1
u
d
u
(x, t) +
(x, t)
=
dx
t
x
0 dx
"
2
2 #1
u
u
=
=0
(x, t) +
(x, t)
t
x
=2
so by the constant value theorem (and thus the mean value theorem)
E is constant.
164
K149
H is a smooth function with
H (t) = 0
H (t) = c
for t 0
for t 2
L (t) = 1
and 0 L (t) 1 for all t.
for t
/ [0, 1]
for t [2, 1 2]
Now let kn (x) = (1)[2nx] L/20 (2nx2[nx]) and follow Exercise 8.66.
165
K150
We seek to minimise
Z b
F (x, y(x), y 0 (x)) dx with F (u, v, w) = g(u, v)(1 + w 2 )1/2 .
a
g,1 (x, y)y 0 + g,2 (x, y)y 0 2 + g(x, y)y 00 g(x, y)y 0 2 y 00
+
.
= g,2 (x, y)(1 + y )
(1 + y 0 2 )1/2
(1 + y 0 2 )3/2
Multiplying through by 1 + y 0 2 gives the required result.
0 2 1/2
We seek to minimise
2
x ds.
xy 00
=0
1 + y02
or
xy 0
=0
(1 + y 0 2 )1/2
(observe that g(x, y) has no direct dependence on y and recall Exercise 8.63). Thus
xy 0
=c
(1 + y 0 2 )1/2
whence
c
y0 = 2
(x c2 )1/2
and
x
y + a = c cosh1
c
for appropriate constants a and c.
d
dx
166
K151
Precisely as in the standard case, we consider
Z b
G() =
f (x, y(x) + g(x), y 0 (x) + g 0 (x), y 00 (x) + g 00 (x)) dx
a
and require
0
0 = G (0) =
Z b
d
g(x) f,2 (x, y(x), y 0 (x), y 00 (x)) f,3 (x, y(x), y 0 (x), y 00 (x))
0=
dx
a
2
d
+ 2 f,4 (x, y(x), y 0 (x), y 00 (x)) dx
dx
for all g as above, so we get an Euler-Lagrange type equation
d
0 = f,2 (x, y(x), y 0 (x), y 00 (x)) f,3 (x, y(x), y 0 (x), y 00 (x))
dx
d2
+ 2 f,4 (x, y(x), y 0 (x), y 00 (x)).
dx
In the case given this becomes
0 = 24 +
d2 00
y (x)
dx2
that is to say
y 0000 (x) = 24
so y = A + Bx + Cx2 + Dx3 + x4 . The boundary conditions give us
A = B = 0, 0 = C + D + 1, 4 = 2C + 3D + 4 so C = 3, D = 2 and
y(x) = 3x2 + 2x3 + x4 .
167
K152*
No comments.
168
K153
(i) g(1/2) = 1, g(x) = 0 otherwise.
(ii) Could take
g(x) =
169
K154
Second sentence done exactly as the case h = 1 (Lemma 9.6).
R
If 1 f (t) dt converges then, since
Z Nh
N
1
N
1
X
X
f (nh),
f (t) dt h
f (nh)
h
n=0
we have (allowing N )
Z
X
f (nh)
h
n=0
and so
as h 0+.
n=1
f (t) dt h
f (nh),
n=1
f (t) dt hf (0) 0
f (nh)
h
n=1
0
G(x) =
170
K155
(i) Observe that
Z
Z x
0
f (t) dt f (n) +
f (x) f (n) +
n
x
0
|f (t)| dt f (n) +
for n x n + 1 and so
Z n+1
Z
f (x) dx f (n) +
n
n+1
n
|f 0 (t)| dt
n+1
|f 0 (t)| dt.
Similarly
n+1
n
f (x) dx f (n)
n+1
n
|f 0 (t)| dt.
It follows that,
Z
Z m+1
m
X
m+1
f (r)
f (x) dx +
|f 0 (x)| dx
n
n
r=n
R
Pn
and, if 1 f (x) dx converges, r=0 f (r) is a Cauchy sequence and converges by the general principle of convergence.
P
similar argument shows that
R nConversely, if r=0 f (r) converges, a P
f (x) dx tends to a limit. Further, if
r=0 f (r) converges we have
1
f (n) 0 and the arguments above show that
Z n+1
Z n+1
|f 0 (x)| dx 0
|f (x)| dx |f (n)| +
n
n
R
as n and so 1 f (x) dx converges.
(ii) If f is decreasing, positive and has a continuous derivative
Z X
Z X
0
f 0 (x) dx = f (1) f (X) f (1).
|f (x)| dx =
1
(2n)1 22n
(2n + 1)1 22n1
if 22n + 21 x 22n+1 21 , n 0,
if 22n+1 + 21 x 22n+2 21 , n 0,
171
K156
n+1/2
n1/2
log x dx log n =
n+1/2
(log x log n) dx +
n
n+1/2
log(x/n) dx +
n
1/2
n
n1/2
(log x log n) dx
log(x/n) dx
n1/2
t
t
+ log 1
dt
log 1 +
=
n
n
0
making substitutions like x = t + n and x = n t.
Z
t
t
1
1
2
log 1 +
sup
+ log 1
n
n
n t
t(0,1/2) t + n
2t
4
= 21 sup
2.
2
2
3n
t(0,1/2) n t
Thus by the comparison test
X
t
t
log 1 +
+ log 1
n
n
n=1
= (N + 12 ) log(N +
Thus
N +1/2
dx
1/2
1
) 21 ) log 21
2
(N + 1).
(N + 12 ) log(N + 12 ) N log N ! c0
for some constant C and the result follows on taking exponentials (and
noting that exp is continuous).
172
K157
With the notation of the previous question
2
n!
2n!
2n
n
e
e
C 1
(2n + 1/2)(2n+1/2)
(n + 1/2)(n+1/2)
where C is the constant of K156. Thus
1 2n
) 2n 1/2
2n (1 + 4n
C0
1 2n 2 n
n (1 + 2n )
and
2n
C 00
4 n
n
as n for appropriate constants C 0 and C 00 . In particular, we can
find constants K1 and K2 with K1 > K2 > 0 such that
2n
n 1/2
4n n1/2 .
K1 4 n
n
Thus
2n n
2n n
n
z 0 if |z| < 1/4 and n
z 0 if |z| > 1/4
n
n
P
2n n
z has radius of convergence 1/4.
so
n=0 n
n
n
z | K n1/2 for all |z| = 1/4 so,
If < 1/2 we have |n 2n
P n 2n n 1
z converges absolutely and so
by the comparison test,
n=0 n
n
converges everywhere on the circle of convergence.
n
P 2n n
If 1/2 we have |n 2n
z
z
|
K
for
all
|z|
=
1/4
so
2
n=0 n
n
n
diverges everywhere on the circle of convergence.
2n
If 1/2 then
n
K2 n1/2 and the comparison test tells
P 2n n n
us that n=0 n n z diverges when z = 1/4.
n 1/2
P (1 + x)
2n n
1 + x for 0 and x 0.) If 1/2 > 1/2 then n=0 n n z
converges everywhere on the circle of convergence except the point
z = 1/4.
173
K158
(i) Observe that
Z
g(X)
f (s) ds =
0
f (x)g 0 (x) dx
0
sin(x + n)
sin(x + (n + 1))
(1)n+1
x + (n + 1)
x + n
so
0 (1)
n+2
sin(x + (n + 1))
dx (1)n+1
x + (n + 1)
and, if we write
vn = (1)
n+1
(n+1)
n
sin(x + n)
dx
x + n
sin x
dx,
x
we have
as X .
(n+1)
n
sin x
1
x dx (n) 0,
sin x
dx L
x
174
Z n
n1 Z (r+1)
X
sin x
sin x
dx =
x dx
x
r=1 r
n1 Z (r+1)
X
x r
dx
r
r
r=1
Z
n1
X
1
g(x) dx
=
r
0
r=1
as n .
R x
(v) If > 0, then the arguments above show that 1 sin
dx conx
verges. If 0, then
Z
(n+1) sin x Z (n+1)
| sin x| dx 9 0
dx
n
x
n
R x
so 1 sin
dx does not converge.
x
However, if 0 x 1 we have x sin x 2x/ so
sin x
x1 2x1 /
x
and comparison shows that
Z 1
sin x
dx
x
175
K159
(i) Observe that
n
n
X
X
sin((n + 12 )x)
1 ei(2n+1)x
=
1+2
cos rx =
eirx = einx
1 eix
x
r=1
r=n
for |x| .
sin (n + 12 )x
dx
sin 12 x
Z
sin (n + 1 )x Z alpha sin (n + 1 )x
2
2
dx =
dx
0
1
1
(n + f rac12)| sin |
sin 2 x
sin 2 x
as n .
(iv) Write
2(sin 12 x 21 x)
1
2
=
x sin 21 x
x sin 12 x
and use LHopital or Taylor expansions.
(v) Given any > 0 we can find an > 0 with > such that
1
x sin x <
2
2 sin((n + 21 )x)
sin((n + 12 )x)
dx
dx 2 2 2 .
sin x
x
sin
x
2 2
2 2 .
,
dx
sin x
, dx =
x
176
K160
(i) and (ii) (Here permits means permits but does not imply.)
f (n) = O(g(n)), permits f (n) = o(g(n)), permits f (n) = (g(n))
and permits f (n) g(n).
f (n) = o(g(n)), implies f (n) = O(g(n)), forbids f (n) = (g(n)) and
forbids f (n) g(n).
f (n) = (g(n)) permits f (n) = O(g(n)), permits f (n) g(n) and
forbids f (n) = o(g(n)).
f (n) g(n) implies f (n) = O(g(n)), implies f (n) = (g(n)) and
forbids f (n) = o(g(n)).
Suitable examples will be found amongst the pairs f (n) = g(n) = 1;
f (n) = n, g(n) = 1; f (n) = 1, g(n) = n.
(iii) f (n) = o(1) means f (n) 0 as n 0. f (n) = O(1) means f is
bounded.
(iv) Could take f (n) = 1 + n(1 + (1)n ), g(n) = 1 + n(1 + (1)n+1 ).
177
K161
(i) True. If 0 fj (n) Aj gj (n) with Aj 0 then
and
(sin x)2 x2 = O(x4 ).
(vi) False. Use LHopital or (I think better) observe that
cos x 1 + x2 /2 x4 /4! = O(x6 ), sin x = x x3 /6 + (x)x5 ,
where |1 (x)|, |2 (x)| 10 for |x| small (we can do better but we do
not need to) so
cos x 1 + sin2 x/2 sin4 x/4! + x4 /6 = O(x6 ).
178
K162
(i) If 1 < 0, then x is decreasing so
Z n+1
n
n Z r+1
n+1
X
X
X
r
x dx =
x dx
r .
r=1
r=1
r=2
Since
this gives
x dx =
1
n+1 1
,
+1
P
( + 1) nr=1 r 1 + 2n
n+1
n+1
as n .
If 0, then x is increasing so
Z
n Z r+1
n
X
X
x dx =
r
r=1
r=1
n+1
1
x dx
n+1
X
r=2
r=n
r1 diverges.
179
K163
Argument similar to that in K156. We have |g 00 (x)| Ax for
x R say. If |t| 12 and n R + 1, then applying the mean value
theorem twice gives
|(g(t + n + 12 ) g(n + 12 )) + (g(t + n + 12 ) g(n + 12 ))|
= |g 0 (s + n + 12 ) g 0 (s + n + 12 )|
= |2sg 0 (u + n + 12 )| A(n 12 )
1
2
n+1
n
(g(x) g(n + 21 ) dx
Z 1
2
1
1
1
1
so
1
2
Bn
Z
n+1
We have
Z
n+1
n
g(x) dx g(n +
1
)
2
Bn
n
n+1
X
Bn+1
g(x) dx
g(r) C +
+ 1
r=1
R n+1
for some constant C. Dividing through by 1 g(x) dx gives the required result.
180
K164
Since sin x > 2x/ for 0 x /2, we have
/2
log 2 dx +
0
/2
log(sin x) dx +
0
Z /2
= log 2 + 2
log(sin x) dx.
2
0
Combining our formulae gives the result.
/2
log(cos x) dx
0
181
K165
so
Ra
Suppose that 0 f (x) dx converges. Since f is decreasing
Z a
Z a
Z a
N
a X ar
f (x) dx
f
f (x) dx
f (x) dx
N r=1
N
0
a/N
0
Z a
N
ar
X
f
f (x) dx.
N
0
r=1
Ra
Suppose that 0 f (x) dx diverges. Then given any K we can find an
> 0 such that
Z a
f (x) dx K + 1.
f (x) dx
N
N rN
f (x) dx 1
f
N
N rN
and so
N
ar
X
K
f
N
r=1
N
ar
X
f
N
r=1
as N .
(ii) Consider a = 1,
(
2n (1 24n |x 2n |) if |x 2n | 24n ,
f (x)
0
otherwise.
(iii) Observe that
(1 z)
so, dividing by z 1,
N
1
X
r=0
N
1
X
r=0
z =1z
z =1z
N
1
Y
r=0
N
1
Y
r=1
(z r )
(z r )
182
for all z.
z =1z
N
1
Y
r=1
(z r )
N
1
Y
r=1
(1 ) =
1+2++(N 1)
r=1
where = /N . Thus
N =
N (N 1)/2
(2i)
N 1
N
1
Y
N
1
Y
r=1
( r r )
N
1
Y
r
r
N 1
sin
=2
sin .
N
N
r=1
X
log sin
sin
log
=
2N r=1
N
2N
N
r=1
log N 1
2N
2
log N
N 1
log 2 +
log 2.
=
2N
2N
2
=
183
K166
Observe that
Ek = {x : f (x) = k}
is the union of a finite set of disjoint intervals (of the form [a, b)) of
total length (9/10)k1 /10. Thus (writing IA for the indicator function
of A)
!
X
X
kIEk + X 1
I Ek
fX =
1kX
1kX
Z 1
X
X
(9/10)k1 .
k(9/10)k1 + X 1 91
fX (x) dx = 101
0
1kX
1kX
!
X
(9/10)k1 = X(9/10)[X] ([X] + 1)(9/10)[X] 0
0 X 1 101
1kX
(j + 1)tj
j=0
gX (x) fX (x) = 0
for all x k=1 Ek so, given any > 0, we know that gX (x)fX (x) = 0
on the union of a finite set of disjoint intervals of total length 1 /2
and so
IfX X = I fX X I gX I gX I fX + X = IfX + X.
184
K167
(ii) Observe that if > 0 is fixed, we can find an R0 () such that
Z
Z b
fR,S (x) dx ab f (x) dx
a
for all R, S R0 (), and if R and S are fixed, we can find an 0 (R, S) >
0 such that
Z
Z b
b
fR,S (x) dx
a f (x) dx
a
(iii) If (A) is true, the argument of (i) shows that, if we write f + (x) =
max(f (x), 0) and f (x) = min(f (x), 0), then (A) remains true for both
f+ and f . There is this no loss of generality in assuming that f is
positive. The argument of (ii) now applies.
(iv) The substitution t = x1 suggests
f (t) =
sin t1
.
t
185
K168
(Note that, since (x, t) 7 etx is continuous on [0, X] [a, b], the
apparent singularity at x = 0 must be a trick of the notation.)
Z X ax
Z XZ b
e
ebx
etx dt dx
dx =
x
0
a
0
Z bZ X
=
etx dx dt
a
0
Z b
1 etX
=
dt.
t
a
Now
etX
eaX
0
t
t
for t [a, b] so
Z b tX
e
b
aX
0
0
dt e
log
t
a
a
as X 0 so
Z b
Z b tX
Z X ax
1
e
e
ebx
b
dx =
dt
dt log
x
t
a
a t
a
0
as x .
186
K169
By the fundamental theorem of the calculus
Z t Z d
Z d
d
f (u, v) dv du =
f (t, v) dv.
dt a
c
c
By differentiating under the integral sign and using the fundamental
theorem of the calculus
Z t
Z d
Z d Z t
Z d
d
f (u, v) du dv =
f (u, v) du dv =
f (t, v) dv.
dt c
a
a
c
c t
Thus
Z t Z d
Z d Z t
d
f (u, v) dv du
f (u, v) du dv = 0,
dt
a
c
c
a
and by the constant value theorem
Z t Z d
Z d Z t
f (u, v) dv) du
f (u, v) du dv
a
c
c
a
Z a Z d
Z d Z a
=
f (u, v) dv du
f (u, v) du dv = 0.
a
187
K170
(i) We have
Z
so
Z
1
0
f (x, y) dx = 0
0
f (x, y) dy =
0
2 if 21 < y < 1,
0 otherwise,
f (x, y) dx dy = 0 6= 1 =
1
0
f (x, y) dy dx.
0
(ii) We have
Z
1
g(x, y) dy =
0
g(x, y) dx = 0
0
so
1
0
g(x, y) dx dy = 0 6= 1 =
1
0
g(x, y) dy dx.
0
Theorem 9.20 demands continuity everywhere. (The underlying problem is that g is not bounded.)
(iii) We have
Z
1
0
1+y 2
y(w 2y 2 )
dw
w3
y2
1+y2
1
y
y3
=
+ 2
2
w w y2
y
=
2(1 + y 2 )2
1
h(x, y) dx =
2
so that
Z
1
0
1
0
1
h(x, y) dx dy =
4(1 + y 2 )
1
0
= 1/8.
188
K171
(ii) Observe that, if R 1, then
Z
Z
Z R
f (x, y) dx
f (x, y) dx
F
(1 +
+ y2)
tan1 R
=A
1 + y2
A( 2 tan1 R)
x2 )(1
dx
Z R
/4
f
(x,
y)
dx
f
(x,
y)
dx
0
0
RR
for all y. Since 0 f (x, y) dx is a continuous function of y it follows
that given any y0 we can find a > 0 such that |y0 y| < implies
Z R
Z R
/2
f
(x,
y
)
dx
f
(x,
y)
dx
and so
Z
Z
f (x, y0 ) dx < .
f (x, y) dx
0
0
R
If follows that 0 f (x, y) dx is a continuous function of y. (This can
be done better using uniform convergence.)
R
Since 0 f (x, y) dx is continuous in y and
Z
A/2 ,
f
(x,
y)
dx
1 + y2
0
RR
comparison shows that 0 0 f (x, y) dx dy exists. A similar but simRRR
pler argument shows that 0 0 f (x, y) dx exists Equation F shows
that
Z Z
Z Z R
( tan1 R) A .
f
(x,
y)
dx
dy
f
(x,
y)
dx
dy
2
2
0
0
0
0
We also know that
Z Z R
Z
f (x, y) dx dy
so that
Z Z
f (x, y) dx dy
Similarly
Z Z
Z
f (x, y) dy dx
R
0
R
0
Z
Z
R
0
R
0
A
f (x, y) dx dy ( 2 tan1 R)
2
189
Z Z
2( tan1 R)A.
f
(x,
y)
dy
dx
f
(x,
y)
dx
dy
2
190
K172
Just use the rectangles [0, 1] [0, N 1 ] and [0, 1] [N 1 , 1].
F2 (0) is not defined because x 7 f (x, 0) is not Riemann integrable.
191
K173*
No comments
192
K174
(iv) Observe that U is open, [2, 2] U and U does not have Riemann length so [3, 3] \ U is closed, bounded and does not have Riemann length.
193
K175*
No comments.
194
K176
(i) Observe that f (tj ) f (b). Recall that an increasing sequence
bounded above tends to limit.
(ii) Chose n(j) and xj so that n(1) = 1, x1 = tn(1) , n(2j) > n(2j 1),
x2j = sn(2j) > x2j1 , n(2j + 1) > n(2j), x2j+1 = tn(2j)+1 > x2j [j 1].
Consider the limit of f (xj ).
Let a = 0, b = 2, f (t) = 0 for t < 1, f (t) = 1 for t 1 tj = 1 j 1 ,
sj = 1.
(vii) If x
/ E {a}, then f (x) = f (x+). But f (x) f (x+) so
f (x) = f (x+). Thus f = f.
If x < y then, choosing x < xn < y < yn < b with xn , yn strictly
decreasing with xn x, yn y, we have f (xn ) f (yn ) so f(x)
f(y). A similar but simpler argument applies when y = b. Thus
f is increasing. If x [a, b) we can find xn x with xn strictly
decreasing and xn
/ E {a}. Thus f(xn ) = f (xn ) f(x) and f is
right continuous.
195
K177
(iii) The function f is Riemann integrable with respect to H if and
only if f (t) f (0) as t 0.
Observe that, if D is a dissection, then
x(t,0]
196
K178
(i) By adding a constant, we may suppose that u(t) 0 as t
We know that u is continuous except at a countable
P set of points xj and
and that, writing j = u(xj ) supt<xj u(t), j=1 j converges (indeed
X
j=1
j lim f (t).
t
and
0 un+1 (t) un+1 (s) un (t) un (s)
s<x<t
and U is continuous.
A very much simpler argument shows that
for all x.
Pn
j=1
H(xxj ) converges
X
j=1
xj )
j H(t
where H(t)
= 0 for t 0, H(t)
= 1 for t > 0, j > 0 and
j=1 j
converges. By the first paragraph we need only show that the function
fN given by
fN (t) = v(t) +
N
X
j=1
xj )
j H(t
197
s<x0 <t
198
K179
(ii) The condition is G strictly increasing.
If G is not strictly increasing, choose a < b such that G(a) = G(b).
Take f (x) = max(0, 14(ba)1 |x(a+b)/2| to see that the theorem
fails.
Suppose G is strictly increasing. If f is bounded continuous positive
function with f (x0 ) 6= 0, observe that there exists an > 0 such that
|f (t) f (x0 )| f (x0 )/2 so f (t) f (x0 /2 for |x0 t| 2. Now
Z
(G(x0 + ) G(x0 ))f (x0 )
f (x) dG(x)
> 0.
2
R
(iii) Same condition as (ii).
Suppose G is strictly increasing. If f (x0 ) 6= 0 there exists an > 0
such that |f (t)f (x0 )| |f (x0 )|/2 for |x0 t| . Choose g continuous
so that g vanishes outside (x0 , x0 + ), g(t)f (x0 ) 0 for all t and
g(X0 )f (x0 ) = 1. By part (ii), g(t)f (t) = 0 for all t which is impossible.
Thus f is identically zero.
199
K180
(i) Observe that
n
n
X
X
|g(xj ) g(xj1 )| =
g(xj ) g(xj1 ) = g(b) g(a).
j=1
j=1
j=1
n
X
j=1
as N .
|f (xj ) f (xj1 )|
N
+5
X
j=5
j 1
200
K181
(ii) Given > 0, we can find a = x0 y0 x1 . . . xn yn = t
such that
n
X
F (yj ) F (xj ) F+ (t)
j=1
and a =
x00
y00
0
= t such that
x01 . . . xm ym
n
X
j=1
Let a = z0 z1 . . . zN = t with
0
}
{x00 , x01 , x02 . . . x0m } {y00 , y10 , y20 . . . ym
Then
F (t) F (a) =
N
X
j=1
F (zj ) F (zj1 )
X
F+ (t)
F (zj ) F (zj1 ) +
F (zj ) F (zj1 ),
and
F (t)
n
X
G+ (yj ) G+ (xj ) +
j=1
n
X
j=1
n
X
j=1
G+ (xj ) G+ (yj1 )
G+ (yj ) G+ (xj )
j=1
n
X
j=1
n
X
f (yj ) f (xj )
201
202
K182
(ii) If F is right continuous at t then, since F = F (a) + F+ F ,
either both F+ and F are right continuous or neither is (which is
impossible by (i)).
(iii) Write
G+ (t) =
t
0
j good
so that
VF (t)
xj1
|F 0 (x)| dx
a
t
a
j bad
(t a)/N
|F 0 (x)| dx (t a)
t
a
203
K183
Observe that f, is well behaved away from 0, so this exercise is
about behaviour near 0. The behaviour is different according as > 0,
= 0 or < 0.
(A) If > 0, we have
|x| sin(|x| ) = |x|+ + |x|+2 (|x|)
with (|x|) 0 as x 0.
204
K184*
No comments.
205
K185
(i) Observe that, in a self explanatory notation,
S(f, D, G + F ) = S(f, D, F ) + S(f, D, G)
and
so
and
206
K186*
No comments.
207
K187
The main problem is the behaviour of (y) for y close to 1.
Observe that writing t = 1 s we have, using appropriate Taylor
theorems,
(1 (1 s)2 )1/2 = (2s s)1/2
208
K188
The points at issue are very similar to those of K187.
209
K189
(iv) Observe that
Z R
Z
2
f (x) dx =
1
x2 dx = (1 R1 )
but
R
0
2 1/2
f (x)(1 + f (x) )
1
dx 2
as R .
R
1
x1 dx = 2 log R
(v) We have
Vol K =
but
0
Vol(K \ K) =
Z0
x1 dx =
(x1/2 + x1 )2 (x1/2 )2 dx
(x2 + 2x3/2 ) dx
0
210
K190
We have
Z
f (x) ds =
=
=
=
f (x) ds
2
f (x) ds
3
f (x) ds
4
Z 1
= (0, )
Also
and
f (x) ds =
f (x) ds = 2
f (x) ds = (0, )
f (x) ds = (0, 2).
211
K191
(ii) We have
p(n k)
un (k + 1)
=
un k
(1 p)k
p(n k)
p(1 p /2)
un (k + 1)
=
un k
(1 p)k
(1 p)(p + /2)
Pn
Now 1 =
j=0 un (j) un (k) so, writing k1 = k1 (n) for the least
integer greater than n(p + /2), we see that
kk1
p(1 p /2)
un (k)
(1 p)(p + /2)
for all k k1
p(1 p /2)
p(1 p /2)
1
0
(1 p)(p + /2)
(1 p)(p + /2)
as n .
Similarly
X
knpn
so
un (k) 0
Pr{|Nn np| n} =
as n .
|knp|n
un (k) 0
(iii) We have
EXj = p1 + (1 p)0 = p
and
EXj2 = p1 + (1 p)0 = p
so var Xj = p p2 = p(1 p).
212
j=1
var Nn = np(1 p)
Pr{|Nn np| n}
as n .
var Nn
p(1 p)
=
0
2
(n)
2 n
213
K192
Recall that (A), (B) and (C) give
d(x, y) = (d(x, y) + d(y, x))/2 d(x, x)/2 = 0.
(i) Set x = y in (C)0 to obtain (B). (C) now follows from (B) and
(C)0 .
(ii) Setting d0 (x, y) = d(x, y), we see that d0 is a metric space. If
d(x, y) 0 for x, y X then X has exactly one point.
214
K193
By (A)0 , x x.
By (B), x y implies y x.
By (C) if x y and y z then
so d(x, z) = 0 and x z.
Observe x x0 , y y 0 gives
215
K194
(i) The result is true with A1 = 1. We use induction on n to prove
it. The result is trivially true for n = 1. Suppose it is true for n = m.
If Sm+1 then writing = (m + 1, (m + 1)) and 0 = 1 we see
that 0 fixes m + 1 so is generated by at most m elements of the form
(ij) (transpositions). Thus = 0 is is generated by at most m + 1
transpositions. Any shuffle of m cards needs at most m swaps.
(ii) Observe that (1i)(1j) = (ij) and use (i).
(iii) Observe that
(123 . . . n)k (12)(123 . . . n)k = (k k + 1)
so any transposition of the form (k k + 1) is a distance at most 2n + 1
from e. Next observe that
(k k + 1)(1 k)(k k + 1) = (1 k + 1)
so any transposition of the form (1 k) is a distance at most n(2n + 1)
from e. Thus, using (ii) and (i), the diameter of Sn is at most 2n2 (2n+1)
with respect to X3 .
(iv) Observe that ar b = bar . Let N be the integer part of (n 1)/2.
If N r then, by induction, the product of r elements of the form a, b,
a1 , must have one of the forms au with |u| r or bav with |v| r 1.
Looking at aN we see that the diameter of Dn is at least N .
(v) Sn has n! elements and Dn has 2n. There appears to be little
relationship between size and diameter.
216
K195
(i) We wish to prove the statement P (N ). If |x| 4k for some
integer k. and
x=
N
X
j=1
PN
2n(j) 2k .
P
If x = 1j=1 xj then x = x1 so P (1) is true.
then
j=1
x=
j=1
then if n(j) k for any j we are done. Thus we need only consider
the case n(j) k + 1 for all 1 j N . Let M be the smallest m such
that
m
X
|xj | 4k1 .
j=1
We know that
M
X
j=1
so
|xj |
N
X
M +1
M
1
X
j=1
|xj | |x|
M
X
j=1
|xj | 2 4k1
x =
M
X
j=1
00
|xj | and x =
N
X
M +1
|xj |
we see that
N
X
j=1
2n(j) =
M
X
j=1
2n(j) +
N
X
j=M +1
217
P
(iii) If 4k+1 > |x| 4k then taking x = 1j=1 xj with x1 = x, we
see that 2k+1 d(x, 0). We also know, by (i), that d(x, 0) 2k .
Thus
4|x|2 d(x, 0) 41 |x|2 .
(iv) I think the graph is quite complex with f constant on intervals
of the form [4k (1 ), 4k ] (for some > 0) and on other intervals
which form a dense set (cf the devils staircase in K251).
218
K196
(i) and (iv) are false. Let X = R with usual metric, X1 = {x : x
0}, X2 = {x : x > 0}, f (x) = 0 for x X1 and f (x) = 1 for x X2 .
Parts (ii) and (iii) have one line solutions for the sufficiently sophisticated. Here are longer solutions.
(ii) True. If x X1 X2 then, without loss of generality, we may
suppose x X1 . Thus there exists a 1 > 0 such that, if z X and
d(x, z) < 1 , we have z X1 . Since f |X1 is continuous, we know that,
given > 0 there exists a 2 () > 0 such that d(x, z) < 2 () implies
(f (x), f (z)) < for all z X1 . Now d(x, z) < min(1 , 2 ()) implies
(f (x), f (z)) < for all z X.
(iii) True. If x X \ Xj , the argument of (ii) shows that f is
continuous at x. If x X1 X2 , then given > 0 there exists a
j () > 0 such that d(x, z) < j () implies (f (x), f (z)) < for all
z Xj and d(x, z) < minj=1,2 j () implies (f (x), f (z)) < for all
z X.
Set f (x) = g(x) when kg(x)k < 1, f (x) = kg(x)k1 g(x) otherwise.
Let
X1 = {x X : kg(x)k 1} and X2 = {x X : kg(x)k 1}
219
K197
(i) Consider X = {1, 2} with the discrete metric and A = {1}.
(iii) By (ii), f is continuous. If A, B 6= choose a A, b B and
apply the intermediate value theorem to show that there exists a c with
f (c) = 1/2.
(iv) If A is closed, then {t [0, 1) : (1 t)a + tb A} is closed. If
A is open, then {t [0, 1) : (1 t)a + tb
/ A} is closed.
220
K198
Suppose A satisfies (i). Then, if ym A and ym y, we can find
Thus A F.
xm A with d(xm , ym ) < 1/m so xm y and y A.
Suppose x A,
then (writing B(x, 1/n) for the
A is closed x A.
open ball centre x radius 1/n) if B(x, 1/n) A = A \ B(x, 1/n) is a
strictly smaller closed set containing A contradicting our assumption.
Thus B(x, 1/n) A 6= and we can find xn B(x, 1/n) A. We have
xn A and xn x.
(iii0 B is an open set with B B such that, if U is open and
B U , we have B U .
Int Q = , Int Z = , Int{1/n : n 1} = , Int[a, b] = Int(a, b) =
Int[a, b) = (a, b).
221
K199
Observe that Cl U Int(Cl U )) and Cl U is closed so that
Cl U Cl(Int(Cl U )).
and Cl(Int(Cl U )) = Cl U .
222
K200
(i) (C) Take = (max(2, kxk))1 , for example.
(iii) Since is absorbing we can define
n(x) = inf{t 0 : t1 x }.
We note that n(x) 0. Our object is to show that kxk = n(x) defines
the required norm.
If 0 the definition gives n(x) = n(x). The fact that is
symmetric now gives
for all R.
n(x) = ||n(x)
(1 + )1 n(x)1 x, (1 + )1 n(y)1 y
223
K201
Observe that
but that
xn = (1, 21 , 31 , . . . , n1 , 0, 0, . . . ) E
N
X
xj ej =
|xj |
j=1
j=1
is a norm on E. But all norms on a finite dimensional space are Lipschitz equivalent so there exists K 1 with Kkxk kxk K 1 kxk.
If yn E, y V and kyn yk 0, then yn is a Cauchy sequence in
(V, k k) so a Cauchy sequence in (E, k k) so, by Lipschitz equivalence, a
Cauchy sequence in (E, k k ) so, since (E, k k ) is complete, converges
in (E, k k ) to some z E. Since kyn zk 0, it follows, by
Lipschitz equivalence, that kyn zk 0. By the uniqueness of limits,
y = z E. Thus E is closed.
224
K202
(i) If
0 y + e0 = y + e
with , 0 R and e, e0 E then
(0 )y = e e0
(0 )T y = 0
so = 0 and e = e0 .
Also
Tx
y E.
Ty
If T = 0 then N = E.
(ii) Observe that the statement that
B(y, ) N 6=
ky + ek <
x = y + e
kxk || whilst T x = T y.
kT xk 1 |T x|kxk
j=1
aj = 0}.
225
if j = 1
1
1
ej (n) = n
if 2 j n + 1
0
otherwise.
n
Set A(n) =
PN (n)
X
j=2
j 1 n + 1.
j 1 . Take
1
ej (n) = A(n)1 j 1
j=2
if j = 1
if 2 j N (n)
otherwise.
X X
X
j=1
j=1
j=1
P
so j=1 ej = 0.
(d) If we use the norm k2 , N not closed. Consider the same example
as (a).
(e) If we use the norm k ku then N is closed. We can use much the
same proof as (c).
[To see T not continuous in (a), (b) and (d) we can consider the
norms of e e(n) and T (e e(n)). In (c) and (e) simple inequalities
give kT k 1, indeed kT k = 1.]
226
K203
The key observation is that, since 21/2 is irrational, x + y21/2 =
x + y 0 21/2 for x, y, x0 , y 0 Q if and only if x = x0 and y = y 0 .
0
for (x, y) Q2 .
227
K204*
No comments.
228
K205
Pick xn Fn . If m n then
d(xn , xm ) diam(Fn ) 0
229
K206
(i) If is a metric, then (x, y) > 0 for x 6= y so f is strictly
increasing.
If f is strictly increasing, then it is easy to check that is a metric.
(ii) If f is not continuous at x, then without loss of generality, we may
suppose that f is not left continuous and there exist x1 < x2 < . . . with
xj < x and xj x together with a > 0 such that f (x) f (xj )
for all j. Since
N
X
j=1
we have
j=1
for all t, s 0,
g(0) = d(0, 0) = 0.
n+1
2X
j=1
g(2n1 ) g
2n+1
X
j=1
2n1
= g(1)
230
231
K207
(i) Write ej for the vector with 1 in the jth place and 0 elsewhere.
Then
and so
kaj ej k kak
|aj | kakkej k1 .
232
K208
We have
N
X
j=1
N
X
(aj + bj )2
a2j
j=1
a2j
j=1
!1/2
!1/2
N
X
a2j
j=1
j=1
!1/2 2
!1/2 2
,
a2j
so,
sequence bounded above converges we have
Psince an increasing
2
j=1 (aj + bj ) convergent. Moreover
X
j=1
(aj + bj )2
a2j
j=1
!1/2
X
j=1
a2j
!1/2 2
(aj aj (n))2
+
aj (n)2
j=1
j=1
M
X
M
X
j=1
j=1
j=1
(aj aj (n))2
!1/2
+ ka(n)k2
(aj aj (n))2
!1/2
+ sup ka(r)k2
r1
sup ka(r)k2
r1
M
X
j=1
a2j
!1/2
sup ka(r)k2
r1
233
Finally, if n, m N
!1/2
!1/2 M
!1/2 M
M
X
X
X
(a2j (n) aj (m))2
(a2j aj (m))2
+
(a2j aj (n))2
j=1
j=1
j=1
M
X
M
X
j=1
j=1
(a2j aj (m))2
!1/2
+ ka(n) a(m)k2
(a2j aj (m))2
!1/2
so, allowing m ,
M
!1/2
X
(a2j aj (n))2
sup ka(s) a(r)k2
r,sN
j=1
and, allowing M ,
ka a(n)k2 0
j=1
j=1
so,
Psince an increasing sequence bounded above converges, we have
j=1 aj bj absolutely convergent and so convergent.
234
K209
(ii) we have
log
x y
+
p q
1
1
log x log y
p
q
235
K210
(i) Observe that (p 1)q = p, so, taking g(t) = |f (t)|p1 , we have
1/q
Z b
1/q
Z b
Z b
p
(p1)q
p
,
|f (t)| dt
=A
|f (t)|
|f (t)| dt A
a
b
a
|f (t)| dt
b
a
|f (t)g(t)| dt +
Z
b
p
|h(t)| dt
|h(t)g(t)| dt
a
1/p ! Z b
a
|g(t)| dt
1/q
236
K211
It may be helpful to look at K209.
(iv) If k k is derived from an inner product, then
Taking
1
p
f (x) dx =
0
1/4(4t)p dt.
0
kf gk1 Akgk
237
K212
(i) p = 1. A square with sides at /4 to the axes.
p = 2. A disc.
p = A square with sides parallel to the axes.
Other values of p give smooth boundary (as does p = 2).
(ii) If 1 < p < , then the chain rule shows that fp differentiable
except at (0, 0). Since
fp (x, 0)
= sgn x
x
does not tend to a limit as x 0, fp is not differentiable at (0, 0).
If x, y > 0 then f1 (x, y) = x + y so f1 is differentiable on {(x, y) :
x, y > 0}. On the other hand
f1 (h, y) f1 (0, y)
= sgn h
h
does not tend to a limit as h 0 so f1 is not differentiable at (0, y).
Similar arguments show that f1 is differentiable at the points (x, y)
with both x 6= 0 and y 6= 0 and only there.
If x > y 0 then f (x, y) = x so f1 is differentiable on {(x, y) :
x > y 0}. However
(
f (x + h, x) f (x, x)
1
if h > 0,
=
h
1 if h < 0,
so f is not differentiable at (x, x) if x > 0 and a similar argument
shows that f is not differentiable at (0, 0). Similar arguments show
that f is differentiable at the points (x, y) with x 6= y and only there.
(iii) If kxks = 1 then |xj | 1 and |xj |s |xj |r so
1 = kxkss kxkrr
so kxkr 1.
x = (1, 0, 0, . . . , 0).
(iv) We have
n
X
j=1
|xj |r
n
X
j=1
|xj |rp
!1/p
n1/q .
238
If bj = j , then b lr but b
/ ls .
(vii) Argument similar to (iii).
Consider gn = max(1 nx, 0).
Essentially same but there is scale change and we have the formula
(b a)1/r kf ks (b a)1/s kf kr .
239
K213
(ii) First look at maps from (Rn , k k ) to (Rm , k k ) Observe that
n
n
X
kT xk = max
aij xj max
|aij |kxk
1im
1im
j=1
j=1
Pn
n
X
j=1
j=1
|aij | =
n
X
j=1
|aij |ky(i)k .
|aij |.
n
X
kT xk = max
aij xj max
|aij ||xj | max max |aij |kxk1
1im
1im 1jn
1im
j=1
j=1
1im
m X
kT xk1 =
aij xj
i=1
=
=
j=1
n
m X
X
i=1 j=1
n X
m
X
j=1 i=1
n
X
j=1
|xj |
max |
1jn
|aij ||xj |
|aij ||xj |
m
X
i=1
m
X
i=1
|aij |
|aij |kxk1
Pm
m
X
i=1
i=1
|aij | =
|aij |.
m
X
i=1
|aij ky(j)k1 .
240
1
1
1 1
1
and
1 1
1
we see that, in the first case the bound is attained but in the second
case it is not since
1 1
x
x+y
=
1 1
y
xy
and |x + y| + |x y| 2 max(|x|, |y|). However, if we really need a
to find kT k we can obtain it as a solution of a linear programming
problem.
241
K214
Suppose a, b X. If > 0, we can find a y E such that
d(a, E) d(a, y) .
Thus
Since is arbitrary
and, similarly, d(a, E) d(a, b) + d(b, E). Thus |f (a) f (b)| d(a, b)
and f is continuous.
Suppose k K, l L. In the previous paragraph we saw that
so, by definition,
242
K215
The intersection of closed bounded sets is closed and bounded. Choose
xn Kn . Since K1 is closed and bounded, the theorem of Bolzano
Weierstrass says that we can pick n(j) such that xn(j) x for
some x K1 . Since xn(j) Km for j sufficiently large and Km is closed
x Km for all m so x K.
If (K, Kn ) 9 0 then we can find a > 0 such that there exists xn
Kn with d(xn , K) > (recall KN KN +1 ). By the argument of the
first paragraph we can find n(j) and x K with d(xn(j) , x) 0
which is absurd.
243
K216
(i) and (iii). It is helpful to observe that d(e, f ) = ({e}, {f }).
(ii) Take y Y if and only if d(y, E) 2/N . Observe that the set
of possible Y is finite.
244
K217
Since E is closed, r = inf{kz yk : y E} > 0. If 1 > 0 > 0 we
can find y0 E such that
kz y0 k > r(1 0 )
kx ek kx0 k1 r = (1 + 0 )1 > 1
245
K218
(i) If n 9 0 we can find a > 0 and n(j) with n(j) . If
un(j) (j) = , uk (j) = 0 if k 6= n(j), then u(j) E but the sequence
has no convergent subsequence.
Suppose n 0 and xn E. Set n(0, j) = n(j). Then since
[r , r ] is closed and bounded we can use BolzanoWeierstrass to
find sequences n(r, j) and yr [r , r ] such that
(a) n(r, j) is a subsequence of n(r, j 1) and n(r, j) > n(r 1, j).
(b) xr n(r, j) xr as j 0.
jr+1
max(2 , sup |j |) 0
jr+1
so x(n(r, r)) x as r .
P
PN (j)
(ii) If
diverges,
we
can
find
N
(j)
such
that
n
n=1
n=j n > 1.
PN (j)
If uk (r) = k /( n=j n ) for j k N (j), uk (r) = 0 otherwise, then
u(k) E but the sequence has no convergent subsequence.
The proof of the positive result resembles that in (i).
246
K219
(i) Take complements.
(ii) If not we can find xn such that B(xn , 1/n) does not lie in any U .
By the BolzanoWeierstrass property we can find n(j) 0 and x X
such that xn(j) x. Now there must exist a U U with x U . Since
U is open there exists a > 0 such that B(x, ) U . But we can find
a J such that n(J) > 2 1 and kxn(J ) xk < /2 so
B(xn(J) , 1/n(J)) B(x, ) U
S
(iii) By Lemma 11.22 we can find y1 , y2 , . . . yM such that M
m=1 B(ym , ) =
X and by (ii) we can find Um U with B(ym , ) Um . Thus
SM
m=1 Um = X.
247
K220
(i) Take contrapositives.
(ii) If the space has property (B) then, taking U to be the collection of
S
B(y, y ), we can find y1 , y2 , . . . , yM such that X = M
m=1 B(ym , ym ).
Now set N = max1mM Nm . We have xN
/ B(ym , ym ) for each
1 m M and this gives a contradiction.
(iii) Combine the results of this question with that of K219.
248
K221
k kB is not a norm since k1kA = 0 but 1 6= 0. The rest can be
checked to be norms.
Recall that Lipschitz equivalent metrics are either both complete or
neither complete.
kf k kf kA 2kf k
249
K222
(i) By the inverse function rule
d
1
sin1 x =
dx
(1 x2 )1/2
for |x| < 1. By the binomial expansion (or by some rigorous Taylor
expansion)
n
Y
X
(2r 1) n
1
=
1
+
t
(1 t)1/2
2r
n=1 r=1
Y
n
X
d
(2r 1) 2n
1
sin x = 1 +
x
dx
2r
n=1 r=1
for |x| < 1. But we may integrate term by term within the radius of
convergence so (since sin1 0 = 0)
Y
n
X
X
22n 2n 2n+1
(2r 1) x2n+1
1
sin x = x +
x
=
2r 2n + 1 n=0 2n + 1 n
n=1 r=1
(ii) The power series has radius of convergence 1 so the only other
points to consider are x = 1. Taking logarithms (cf K96) shows that
n
n
Y
1
(2r 1) Y
1
=
n1/2
2r
2r
r=1
r=1
so
n
X
1 Y
(2r 1)2r
2n
+
1
n=1
r=1
converges by comparison with n3/2 so we actually have (by the Weierstrass M-test) uniform convergence on [1, 1] so the power series is
continuous on [1, 1] so, since sin1 is also continuous on [1, 1], we
have equality at x = 1 and x = 1 as well.
250
K223
(i) fn (x) = (n2 + x2 )1/2 (with positive square root) will do.
(ii) Use (i).
(iii) If fn is Cauchy in k k then both fn and fn0 are Cauchy in k k
converge uniformly to continuous f and F but then we know that f is
differentiable with f 0 = F . Thus f C 1 ([0, 1]) and kfn f kC 0.
251
K224
Let hn : [1, 1] R be given
2n+3 (x + 1)
1
hn (x) =
2n+3 x
h(x)
by
for
for
for
for
x [1, 1 + 2n3 ],
x [1 + 2n3 , 2n3 ]
x [2n3 , 0]
x [0, 1].
1
hn (t) dt
1
By the fundamental theorem of analysis, gn is differentiable with continuous derivative hn so |gn0 (x)| 1 for all x and |gn (x)| = 1 outside
three intervals of total length 2n1 . We observe that gn (x) 1 |x|
uniformly on [1, 1]
We define kn : R R to be the 1-periodic function with kn (x) =
21 gn (2(x 1)) for x [1, 1] and define fn : [0, 1] R by fn =
n1 kn (nx) for x [0, 1].
Observe that fn A, kfn k 0 as n 0 but I(fn ) 0 and
I(0) = 1. Thus I is not continuous with respect to the uniform norm.
However if gn , g A and kgn gk 0 then gn0 (x) g(x) uniformly,
so (1 (gn0 (x))4 )2 (1 (g 0 (x))4 )2 uniformly, so
(1 (gn0 (x))4 )2 + gn (x)2 (1 (g 0 (x))4 )2 + (x)2
252
K225
(i) If fN, (x) = sin 2N x then
Z 1
I(fN, ) =
1 2(2N )4 (cos 2N x)4
0
with
= 1 AN 4 + B8 N 8 + C2
A = 2(2)
B = (2)
C = (2)
Also
0
1
kfN, k = (1 + 2N ).
J(f ) I(0) =
1
0
h(x) dx 0
with equality only if h(x) = 0 for all x [0, 1] i.e. only if f = 0. Thus
if kf k < 1/2 we have J(f ) 0 with equality if and only if f = 0.
Use the fn of K224.
(iv) This is really no more complicated than finding a smooth function u : R R with 0 < u(x) < 1 for all t and inf xR u(x) = 0,
supxR u(x) = 1.
Set v(x) =
1
2
253
K226
(iv) It remains true that d is a metric and that the continuous
functions form a closed subspace of B(E) and that the uniform limit
of continuous functions is continuous since the proofs do not use the
completeness of (Y, ).
Suppose that Y is not complete, so we can find a Cauchy sequence
y(n) in Y which does not converge. Let E = e and fy (e) = y. (Note
B(E) = C = C(E) and the members of C(E) are precisely the fz .)
Then d (fy , fz ) = (y, z) so the fy(n) do not converge. B(E), C(E)
and C(E) are not complete and the general principle of convergence
fails.
254
K227
Observe that, if 0 < v < u, then evx eux x1 0 as x 0 (exponentials beat powers) so we can find an A such that
0 eux x1 Aevx
R
for x 1 so, by comparison, 1 eux x1 dx converges. Also
so, by comparison,
converges.
R1
0
0 eux x1 x1
R
0
eux x1 dx
0 (u) = u1 (u).
Thus
d
u (u) = 0
du
and, by the constant value theorem,
(u) = A u .
(ii) We have
n (u) =
1
u .
(n 1)!
(iii) The arguments including the proofs of convergence for the appropriate integrals are similar to but simpler than those of (i).
Z
h
i
2
2
0
(u) =
xeux ex /2 dx = eux ex /2
u(u) = u(u)
so
2 /2
d
2
(u)eu /2 = 0
du
for some constant A.
255
(iv) To obtain (i) make the substitution s = ux. To obtain (iii) make
the substitution y = x u.
256
K228
Since f (n+1) g uniformly on the closed interval with end points a
and x, we have
Z x
g(t) dt = g(x) g(a).
a
d x
e g(x) = 0
dx
so, by the constant value theorem, g(x) = Cex for some C.
No, f (x) = 1 + ex has the property.
257
K229
(i) Since exponentials beat polynomials, fn (x) 0 for x 6= 0. But
fn (0) = 0 0, so fn 0 pointwise for all , > 0.
(ii) We observe that (if x (0, 1))
so, examining the sign of fn0 , we see that there is a unique maximum
at x = /( + n). Thus
=n
1
0 f (x) f
+n
1 + /n
+ nr
fn (x) dx
2/n
fn (x) dx
1/n
2/n
n (n )
1/n
=n
so
R1
0
2
1
n
2
1
n
dx
fn (x) dx 9 0 if + 1.
.
1>>
+1
We have
Z 1
Z
fn (x) dx =
0
as n .
fn (x) dx +
0
n x dx +
n
Z 1
n
fn (x) dx
n (1 x)n dx
n(+1)
n
=
+
(1 n )n+1
+1
n+1
n
n(+1)
n(1)
+
(1 n ) (1 n )n
0,
=
+1
n+1
258
259
K230
(i) Suppose that gn g uniformly and xn x. Let > 0. By the
continuity of g at x, we can find a > 0 such that |g(t)gn (t)| < /2 for
|x t| < and t [a, b]. Now there exists an N such that |xn x| <
for n N and an M such that kgn gk < /2 for n M . Thus if
n max(N, M )
|gn (xn ) g(x)| |gn (xn ) g(xn )| + |g(xn ) g(x)| < .
|g(xn(j) ) g(y)|
Now we know that gn (xn(j) ) g(xn(j) ) so we can find m(j) > n(j)
such that
|gm(j) (xn(j) ) g(xn(j) )| < /2,
(ii) The proof only if in (i) still works. However if part fails (even
if gn is continuous). Let (a, b) = (0, 1) and gn (x) = max(1 nx, 0). If
260
x (0, 1), then we can find a > 0 such that [x , x + ] (0, 1).
Since gn 0 uniformly on [x , x + ] we apply part (i) to show
that xn x implies gn (xn ) g(x). However gn does not converge
uniformly on [0, 1].
261
K231
(i) Since f is continuous on [0, 1 ], it attains its bounds so there
exists a y [0, 1 ] with f (y) f (x) 0. Since 0 f (y) < 1, we
have
Z 1
f (x)n dx nf (y)n 0.
n
0
Observe that (if f 0 (1) 6= 0) f 0 (1) > 0 since f (1) > f (1 x) for
x (0, 1). Given any with f 0 (1) > > 0 we can find a 1 > > 0
such that
|f (x) 1 f 0 (0)(1 x)| = |f (x) f (1) f 0 (1)(1 x)| (1 x)
and so
1 (f 0 (1) )(1 x) f (x) 1 (f 0 (1) )(1 x)
for x [1 , 1]. Thus
Z 1
Z
0
n
(1 (f (1) )(1 x)) dx
1
f (x)n
1
1
1
Now
n
1
n
(1 A(1 x)) dx = n
=
(1 At)n dt
n
(1 (1 A)n+1 ) A1
A(n + 1)
1
.
f 0 (1)
2n
(1 (1 A)n+1 ) 1 .
(n + 1)
262
1
0
g(x)n dx
263
K232
Observe that, if is fixed with 1/3 > > 0,
Z 1
Z 1/2 /2
Sm, (x) dx
Sm, (x) dx
1/2 /2
1/2 /2
1/2 /2
3
1+
4
dx
264
K233
Define F : [1, 1] R by
(
F (x) =
f (x)
if x [0, 1]
f (x) if x [1, 0].
uniformly on [1, 1]. Also, since Rn (t) = Rn (t), we can find a polynomial Pn with Pn (t2 ) = Qn (t).
Let g : [1, 1] R be given by g(x) = x. Then
265
K234
Write Xn (t) = Y1 + Y2 + + Yn where Yj = 1 if the jth trial is
a success and Yj = 0 otherwise. Observe that (since the expectation
of the sum of random variable is the sum of the expectations, and the
variance of the sum of independent random variables is the sum of the
variances)
EXn (t) =
var Xn (t) =
n
X
j=1
n
X
j=1
EYj = nt
var Yj = nt(1 t).
Xn (t)
Xn (t)
Xn (t)
Pr
t = Pr
E
n
n
n
var(Xn (t)/n)
2
t(1 t)
=
n 2
1
2
n
so, writing IA for the indicator function of A,
Xn (t)
Xn (t)
Xn (t)
Ef
f (t)E
f
f (t) I[t,t+][0,1]
n
n
n
Xn (t)
Xn (t)
+E
f
f (t) I[0,1]\[t,t+]
n
n
Xn (t)
Xn (t)
t + 2M Pr
t >
Pr
n
n
2M
+ 2.
n
Allowing n and observing that is arbitrary gives the required
result.
=
We have
n
n
r
r n
X
X
pn (t) =
Pr(Xn (t) = r) =
f
f
tr (1 t)nr ,
n
n
r
r=0
r=0
so pn is polynomial of degree n.
266
Since
and
n
X
n r
n r
nr
t (1 t)
0,
t (1 t)nr = 1
r
r
r=0
n
X
n r
r=0
r n
tr (1 t)nr = t
we can use Jensens inequality (or we could just quote K144 (iv)).
267
K235
(i) Suppose that fn does not converge uniformly to f . Since the fn (x)
are decreasing this means that there exists a > 0 and xn I such
that fn (xn ) > f (x) + . By the theorem of BolzanoWeierstrass we can
find a subsequence n(j) and an x I such that xn(j) x. Since
fn (x) f (x) we can find an N such that f (x) + /2 > fN (x) Since
f fN is continuous we can find an > 0 such that > fN (y) f (y)
for all y I with |y x| < . Choosing a J such that |xn(J) x| <
we obtain a contradiction.
(ii) False without (a). Let I = [0, 1], fn (1/r) = 1 for all r n,
fn (x) = 0 otherwise.
False without (b). Let I = [0, 1], fn (x) = xn .
False without (c). Witchs hat.
False without (d). Let I = (0, 1), fn = 1/(nx).
(iii) Easy to check n = 0.
If t [1, 1] and 0 pn1 (t) pn (t) |t| then
1
pn+1 (t) pn (t) = pn (t) pn1 (t) + (pn (t)2 pn1 (t)2 )
2
pn (t) + pn (t)
(pn (t) pn1 (t))
= 1
2
(1 |t|)(pn (t) pn1 (t) 0
and
t2 + |t|
|t|(|t| + 1)
=
|t|.
2
2
The result follows by induction.
pn+1 (t)
268
K236
(ii) Proceed inductively. Write pn (x) = xn . Set s0 = 1. When s0 , s1 ,
. . . , sn1 have been defined set
sn = p n
n1
X
hpn , sj i
j=0
hsj , sj i
bj xj =
n
X
aj x j
j=0
j=0
(iv) Write Pn for the set of polynomials of degree or less. Since the
polynomials are uniformly dense, we have, using (iii),
n
X
infn+1
aj sj f = inf kP f k2
P Pn
aR
j=0
(b a)1/2 inf kP f k 0
P Pn
as n .
(v) We have
k
n
X
j=0
bj s j
f k22
=
=
kf k22
kf k22
+
+
n
X
j=0
n
X
j=0
2
hf, sj iksj k1
2 )
n
X
(hf, sj iksj k1
2 )
j=0
2
2
n
n
X
X
hf, sj i
hf, sj i
2
= f
sj +
ksj k2 bj
ksj k22
ksj k22
j=0
j=0
2
2
n
n
X
X
= f
f (j)sj +
ksj k22 (bj f(j))2
j=0
j=0
269
(vi) Observe that sn (t)q(t)w(t) is single signed (that is always positive or always negative) and only zero at finitely many points. Thus
Z b
sn (t)q(t)w(t) dt 6= 0
a
270
K237
Observe (by Leibniz rule or induction or otherwise) that, if n > r
dr
0, then
(x2 1)n has (x2 1) as a factor and so vanishes when
dxr
x = 1. Thus, if n m 0, integrating by parts n times gives
Z 1 n
m
d
2
nd
(x 1)
(x2 1)m dx
n
m
dx
1 dx
1
n1
Z 1 n1
n
m+1
d
d
2
nd
2
n
2
nd
(x 1)
(x 1)
(x 1)
(x2 1)m dx
=
n1
n
n1
m+1
dx
dx
dx
1 dx
1
Z 1 n1
m+1
d
2
nd
(x
1)
(x2 1)m dx
=
n1
m+1
dx
dx
1
= ...
Z 1
dm+n
n
= (1)
(x2 1)n m+n (x2 1)m dx.
dx
1
Thus
Z
1
1
m
dn
2
nd
(x
1)
(x2 1)m dx =
dxn
dxm
0
if n 6= m
R1
2 n
(2n)! 1 (1 x ) dx if n = m.
If In =
R /2
/2
In =
/2
/2
cosn1 cos d
/2
= cos
n1
= (n 1)
sin
Z
/2
/2
/2
+ (n 1)
/2
/2
cosn2 sin2 d
/2
cosn2 (1 cos2 ) d
= (n 1)In2 (n 1)In
so nIn = (n 1)In2 . Since I1 = 2 we have
n = (2n)!I2n+1 = (2n)! 2
n1
Y
j=0
2n 2j
2n+1 (n!)2n
=
.
2n + 1 2j
(2n + 1)!
271
272
K238
P
(i) Observe that P nj=1 P (xj )ej is a polynomial of degree at most
n 1 which vanishes at the n points xk and so must be the zero polynomial.
aj =
ej (t) dt.
1
j=1
273
K239
(i) The argument of K236 shows that sn + sn1 must change sign
at least n 1 times in (a, b). It changes sign only at zeros of odd order
and has exactly n zeros if multiple roots are counted multiply so, by
counting, it must have n simple zeros in (a, b).
(ii) Let the common root be y. We have P (x) = (x y)p(x) and
Q(x) = (x y)q(x). If p(y) = 0, then P has a multiple root at 0 and
we set = 1, = 0. If not then (q(y)/p(y))p q has a root at y and
(q(y)/p(y))P Q has a multiple root at y.
If sn and sn1 have a common root at y then there exist and
both non-zero (since sn and sn1 do not have multiple roots) such that
sn +sn1 and so sn +1 sn1 has multiple roots which is impossible.
(iii) Without loss of generality we may suppose that P has no real
roots in the open interval (x, y). Thus both P and Q are single signed
and, without loss of generality, we may suppose that P and Q are
positive in the open interval (x, y). Since a continuous function on a
closed bounded interval is bounded and attains its bounds we can find
s1 , s2 , s3 [x, y] such that
Q(s1 ) Q(s) Q(s2 ) > 0 and P (s3 ) P (s) 0
P (t) 0 Q(t) 0
274
K240
Fix M 1 temporarily. Observe that (if |x| < M/2) we have
4
1
(x n)2 n2
P
2
so by Weierstrass M-test N
n=M (x n) converges uniformly. We write
FM,N (x) =
N
X
n=M
1
1
+
.
(x n)2 (x + n)2
2 2
8
(x n)3 + (x + n)3 n3
0
so by the Weierstrass M test fM,N
converges uniformly on [M/2, M/2]
to limit gM say. Thus fM is differentiable with derivative gM on
(M, M ). Since M was arbitrary we are done.
If y
/ Z we can find a > 0 such that B(y, 2)Z = and an integer
M such that |y| + 2 < M/4. Thus since the sum of two differentiable
functions is differentiable
M
1
X
1
1
+
F (x) = FM (x) +
(x n)2 (x + n)2
n=0
N
X
n=N
N
+1
X
1
1
2 F (x) F (x + 1)
2
(x n)
(x
+
1)
n
n=N +1
as N 0, so F (x) = F (x + 1).
275
K241
(i) Since g is continuous on [0, 1] it is bounded, with |g(x)| K say
for all x [0, 1]. Now use periodicity.
x + 1
K
x
g
+ g
2
2
2
2(x)2
1
3
= 2 1+
+ 3 (x)x
x
6
1
2
+
+ 3 (x)x
x2
3
where j (x) 0 as x 0.
Observe that g(x) g(0) as x 0 so g is continuous at 0. Now use
periodicity.
(iv) Observe that, if 2x
/ Z, then
x
x+1
+g
g
2
2
X
1
1
=
+ x
2 (cosec2 (x/2) cosec2 ((x + 1)/2)
x
1 2
2
( 2 n)
(2 n + 2)
n=
2
2
X
4
2
2 cos (x/2) sin (x/2)
+
(x
2n
+
1)
=
(x 2n)2
cos2 (x/2) sin2 (x/2)
n=
= 4g(x).
1
= 2 cosec2 (x)
2
(x
n)
n=
276
and looking at x = 0,
X
1
2
= f1 (0) = 2
.
3
n2
n=1
277
K242
(i) fn0 (x) = x(x2 + n2 )1/2 . Thus
if x (0, 1],
1
0
fn (x) 0
if x = 0,
1 if x [1, 0).
We also have
fn |x| =
1
n2
0
2
2
1/2
|x| + (x + n )
n
1
R x(ii) We could take gn (x) = n max(1 |x (2n) |, 0) and fn (x) =
g (t) dt. fn tends pointwise to a (continuous from below at 0) Heav0 n
iside function.
278
K243*
No comments.
279
K244
Observe that
|t|1/2 v(t) = |t|1/2
v(t) v(0)
0 v 0 (0) = 0.
t
(ii) Standard results (chain rule, product rule etc) show that g is
continuous except possibly at points (x0 , 0). Observe that since u is
non-zero only a bounded closed set we can find a K such that |u(t)| K
for all t. Thus, if y 6= 0
|g(x, y) g(x0 , 0)| = |g(x, y)| K|y 1/2 v(y)| 0
(iii) Standard results (chain rule, product rule etc) show that g has
continuous partial derivatives on the open set
A = {(x, y) R2 : x, y 6= 0}.
and so, trivially, has continuous partial derivatives there. The only
point not in A B is (0, 0). But g(x, 0) = 0 for all x and g(0, y) = 0
for all y so partial derivatives exist at (0, 0)
(iv) If y 6= 0, and |y| 1/9
Z 1
G(y) =
g(x, y) dx
0
Z 1
1/2
= |y|
v(y)
u(x|y|1/2 ) dx
0
= v(y)
y 1/2
u(s) ds = v(y),
280
K245
R
By comparison, 0 |fn (t)| dtR converges and so, since absolute con
fn (t) dt converges. Given > 0, we
vergence implies convergence,
0
R
can find an X such that X g(t) dt /2 and then
Z
Z
Z
|fn (t)| dt
g(t) dt /2.
fn (t) dt
X
X
X
R
Similarly 0 f (t) dt converges and
Z
/2.
f
(t)
dt
Thus
Z
Z
f
(t)
dt
f
(t)
dt
n
n
0
0
Z
Z
Z
Z
f (t) dt
fn (t) dt +
fn (t) dt
fn (t) dt +
X
X
X
ZX
fn (t) f (t) dt +
X
The proof above works under the conditions of the last sentence.
281
K246
(ii) If follows by comparison and (using Dinis theorem) dominated
convergence (K245).
R
To prove only if, Robserve that ifR any of the 0 fn (t) dt fails to con
and so
Xn
f (t) dt an 2n
0
R
R
. Thus if 0 f (t) dt converges so do all the 0 fn (t) dt and the values an of the integrals form an increasing sequence bounded above so
converging to a limit A. We have
Z
Z X
Z X
A
fn (t) dt
fn (t) dt
f (t) dt
0
0
0
R
by
part
(i).
Since
X
is
arbitrary
f (t) dt converges. The fact that
0
R
R
fn (t) dt 0 f (t) dt now follows from the if part.
0
282
K247
(i) Observe that
0
2j dj (x, y)
2j
1 + dj (x, y)
= A d(x, y)(1 + d(y, z))(1 + d(x, z)) + d(y, z)(1 + d(x, y))(1 + d(x, z))
Since
0 d(xn , z)
=
N
X
j=1
N
X
2 dj (xn , z) +
j=1
2j
j=N +1
2j dj (xn , z) + 2N 2N
283
X
j=1
284
K248*
No comments.
285
K249*
No comments
286
K250
(i) Suppose that the result is false for some j. Then we can find
(j)
(j)
gk D such that kgk k 1/k but kgk k 1. Since kgk k 0 but gk
does not converge uniformly to 0 as k we have a contradiction.
(ii) Use the fact that we have a norm.
287
K251*
No comments.
288
K252*
No comments.
289
K253
(v) Observe that, given any > 0, we can find
0 = x0 < y0 < x1 < y1 < < xn1 < yn1 < xn < yn = 1
P
such that f (yj1 ) = f (xj ) for 1 j n and nj=0 (yj xj ) . Thus,
if |f (x)f (y)| K|xy| for all x, y [0, 1], we have 1 = f (1)f (0)
K which is impossible if is chosen sufficiently small.
290
K254
(i) Observe that rn (x) = xn is a strictly increasing, differentiable
function. It thus has a differentiable inverse r1/n whose derivative is
given by the rule for differentiation of inverses
rn1 (x)
rn1 (x)
1
=
=
.
(rn1 )0 (x) = 0 1
rn (rn (x))
rn (nrn1 (x))
nx
(ii) Take nn0 th powers of both sides.
(iii) (f) is trivial but must be checked. rn (x) = (r1 (x))n = xn .
(iv) r (x) = (r (x))1 .
291
K255
The key here is bounding |r (x) r (x)| with , Q. Since
|r (x) r (x)| = r (x)|1 r (x)|.
0 r (x) rN (x) = xN RN
F
|r (x) r (x)| RN max r|| (R) 1, 1 r|| (R1 )
whenever N and R x R1 .
292
K256
(v) Since
f 0 (y) = y 11/n y 11/(n+1) 0
x
x
n
P (1/(n + 1)) P (0)
n+1
1/(n + 1)
as n .
1L=L
(viii) To deal with 1 > a > 0 repeat argument or use chain rule on
the maps t 7 t and t 7 at .
If La = 0 then Pa is constant, if La > 0 then Pa is strictly increasing
and if La < 0 then Pa is strictly decreasing. It follows that L1 = 0,
La > 0 for a > 1 and La < 0 for a > 0.
(ix) By the chain rule,
d t d t
(a ) = a = La at = La (a )t
dt
dt
so La = La .
1
L2
Choose = L1
.
2 and set e = 2
(x) We step outside the context of the question and observe that
et = exp t so
La = Lelog a = log aLe = log a.
293
K257
(i) Observe that
ex + ex eix + eix
2
2
1 21/2 x
1/2 3
1/2 5
1/2 7
+ e2 x + e2 x + e2 x )
= e
4
cosh x cos x =
with = ei/4 .
n
n
X
1 X 2n
(1)r
=
(1)r
(2r)!(2n
2r)!
(2n)!
2r
r=0
r=0
1 (1 + i)2n + (1 i)2n
=
=
(2n)!
2
0
22n
(2n)!
if n is odd
if n is even.
(iii) Set f (x) = cosh x cos x. Our main problem is to calculate f (n) (x)
and then to bound it. Using Leibnizs rule gives a calculation for f (r) (0)
which resembles the calculation for (ii). Also
n
X
n
(n)
cosh x 2n cosh R
|f (x)|
r
r=0
for |x| R. This gives an estimate for the remainder
(2R)n cosh R
|Rn (f, x)|
(n 1)!
for |x| R (or something similar) so Rn (f, x) 0.
294
K258
P
n
n=0 an z has radius of convergence at least R, we know that
PSince
n
n=0 an z converges absolutely. Thus if |h| < R |z0 | we know that
X
n
|an |(|z0 | + |h|)n .
|z0 |nr |h|r =
|an |
r
n=0
r=0
n
X
X
n=0
Write cn,r = |an | nr |z0 |nr |h|r if r n, cn,r = 0 otherwise. By Fubinis
theorem for sums,
X
cn,r =
n=0 r=0
cn,r
r=0 n=0
so
n
X
X
n nr r
z h
an
an (z0 + h) =
r 0
n=0 r=0
n=0
X
X
n nr r
an
z h
=
r 0
r=0 n=r
X
X
n nr
r
z0
h
=
r
n=r
r=0
n
br h r
r=0
with
br =
X
n+m
m=0
Thus
P
r=0 br z
r
r=0 br h
z0m .
n=0
an (z0 + h)n =
295
P
r
Let 2 = R |z0 |. Since
r=0 br converges we can find an M such
r
that |br | M . It follows that, if |w| /2,
g(z0 + w) g(z0 )
r1
=
b
w
r
1
w
r=2
|w|
X
k=0
|w|M
bk2 |w|k
|w|M 2
2
as |w| 0.
k=0
(w| 1 )k
2k
k=0
2M |w| 0
296
K259
P
(r)
for all n > r and all x so, by the Weierstrass M-test,
n=r+1 fn (x)
converges uniformly and
fn(r) (x) 2r .
n=r+1
P
P
(r)
Thus
f
(x)
converges
uniformly
for
each
r
and
f
=
n
n=1
n=1 fn is
infinitely differentiable with
X
(r)
f (x) =
fn(r) (x).
n=1
Also
(r)
(0) f
(r)
(0)
r1
X
n=1
|fn(r) (0)|
n=r+1
297
K260
(i) Just observe that
X
n=0
|an z |
X
n=0
|an |(R )n .
if |z| R .
(ii) Since |nan z n | |an z n |, the radius of convergence does not increase.
observe that, if |z| < R and we choose r = (2R+|z0 |)/3, we
PNow
n
n=0 nan z converges absolutely and so converges.
P
P
n
n1
Thus
nz
n=0 nan z has radius of convergence
PR and so n=1 na
n2
folhas radius of convergence R. The result for n=2 n(n 1)an z
lows.
(iii) and (iv). Observe that
n2
n(n 1)
j
=
j!(n 2 j)!
j!(n j)!
n
.
j
so
n2
X n
n
n
n1
j nj
|(z + h) z nz h| =
z h
j
j=0
n2
X
n
|z|j |h|nj
j
j=0
n2
X
n2
2
n(n 1)|h|
|z|j |h|nj2
j
j=0
n(n 1)(|z| + |h|)n2 .
Suppose > 0 and |z| + |h| < r . By parts (ii) and (i), there exists
an A() independent of h and z such that
X
n=2
298
Thus
N
!
N
N
X
X
X
n
n
n1
a (z + h)
an z
h
nan z
n=0 n
n=0
n=1
=
an (z + h)n z n nz n1 h
n=0
|h|2
N
X
n=2
A()|h|2
and, allowing N ,
X
X
X
n
n1
n
nan z A()|h|2
an z
an (z + h)
h
n=0
n=1
n=0
(v) Divide by h and allow |h| 0.
299
K261
P
j
We try, formally, y =
j=0 aj x obtaining, on substituting in the
Legendre equation,
X
X
X
2
j2
j1
(1 x )
j(j 1)aj x 2x
jaj x + l(l + 1)
aj xj = 0.
j=2
j=1
j=0
Rearranging gives
so we must have
that is
aj+2 =
(j + l + 2)(j l)
aj .
(j + 2)(j + 1)
This gives
y(x) = a0
Qr
X
2k)(l + 2k + 1) 2r
x
(2r)!
Qr
X
k=0 (l 2k 1)(l + 2k + 2) 2r+1
x
+ a1
(2r + 1)!
r=0
k=0 (l
r=0
dn+1
(1 x2 )v 0 (x)
dx
= (1 x2 )v (n+2) (x) 2(n + 1)xv (n+1) (x) + n(n + 1)v (n) (x)
300
and
dn+1
2nxv 0 (x) = 2nxv (n+1) (x) + 2n(n + 1)v (n) (x).
dx
Thus differentiating the final equation of the first sentence of this paragraph n + 1 times gives
(1 x2 )vn00 (x) 2xvn0 (x) + n(n + 1)v(x) = 0
301
K262
(this is a special case of Bessels equation.)
P
j
We try, formally, w =
j=0 aj z obtaining, on substituting in the
given equation,
z
X
j=2
j(j 1)aj z
j2
+z
X
j=1
jaj z
j1
+ (z 1)
aj z j = 0
j=0
that is to say
a0 +
j=2
Thus a0 = 0 and
aj2
(j 1)(j + 1)
j
(1)
for j 2. It follows that a2j = 0 for all j 0 and a2j+1 = a1 22j (j!)
2 (j+1) .
Thus
X
(1)j
w(z) = a1
z 2j+1
2j
2
2 (j!) (j + 1)
j=0
Since
1
0
(j 1)(j + 1)
the ratio test shows that our power series has infinite radius of convergence so our solution is valid everywhere.
We note that we can only choose one constant a1 freely. Observe
that setting z = 0 in the original equation shows that automatically
w(0) = 0.
P
j
Let us try and find a solution w =
j=0 bj z for the equation
z2
d2 w
dw
+
z
+ (z 2 1)w = z 2 .
dz 2
dz
We obtain
b0 + (3b2 b0 1)z
j=3
bj2
(j 1)(j + 1)
302
for j 3. Thus
w(z) = b1 W (z) +
X
j=1
with
W (z) =
X
j=0
(1)j
z 2j
Qj1
2
(2j + 1) k=1 (2k + 1)
(1)j
z 2j+1 .
22j (j!)2 (j + 1)
(We can talk about a particular integral plus a complementary function.) As before the solution is valid everywhere.
P
j
If we try and find a solution w =
j=0 cj z for
d2 w
dw
+z
+ (z 2 1)w = z,
2
dz
dz
looking at the coefficient of z gives 0 = 1 which is impossible.
z2
303
K263
By Lemma 11.81 we have (provided |x| < 1)
(1 + x)1/2 =
an x n
n=0
with
n1
1 Y 1
an = (1)
( j).
n! j=0 2
n
an z
n=0
!2
cn z n
n=0
c n xn
n=0
for all |x| < 1 and x real. By the uniqueness of power series (or we
could use K268 below) we have c0 = c1 = 1 and cj = 0 otherwise so
an z
n=0
!2
=1+z
304
P
n
1/2
(1 +
is not continuous at 1, the power series
n=0 cn z of (1 z)
1 1/2
K z) has radius of convergence exactly 1. Now
X
X
n
cn z n = (1 + z)1/2 (1 + z)1/2 (1 + K 1 z)1/2
an z
|
n=0
n=0
= (1 + K 1 z)1/2
X
an K n z n
=
n=0
X
X
X
j
j
uj z
vj z =
wj z j
j=0
j=0
j=0
n
X
unj vj = wn
j=0
Thus
u2n = (2n)! and v2n+1 = (2n + 1)! (so
P we j can ensure
P that
j
u
z
and
v
z
certainly
have radius of convergence 0 and
j
j=0 j
Pj=0
j
j=0 wj z is any series we please (so with any radius of convergence
we please).]
Repeating much the same arguments, if L > K > 1, then (1 +
z) (1+K 1 z)1/2 has power series with radius of convergence 1 and (1+
z)(1 + L1 z)1/2 has power series with radius of convergence L and their
product has power series with radius of convergence K. On the other
hand, if K > L > 1 then (1+z)1 (1+L1 z)1/2 (1+K 1 z)1/2 has power
series with radius of convergence 1 and (1 + z)(1 + L1 z)1/2 has power
series with radius of convergence L and their product has power series
with radius of convergence K. If K > 1 then (1 + z)1 (1 + K 1 z)1/2
has power series with radius of convergence 1 and (1 + K 1 z)1/2 has
has power series with radius of convergence K and their product has
power series with radius of convergence K. Remaining cases (infinite
radius of convergence) dealt with similarly.
1
305
K264
If we write
Sn (t) =
n
X
aj (t),
j=1
then
q
X
j aj (t) =
j=p
q
X
j=p
= q+1 Sq (t) +
q
X
j=p
and so
q
q
X
X
(j j+1 )K + p K = 2p K
j aj (t) q+1 K +
j=p
j=p
for all P
q p 1 and so, by the general principle of uniform convergence,
j=1 j aj (t) converges uniformly.
q
q
q
X
X
X
j
p
j
j
bj x 2x sup
bj x 2 sup
bj x 0
qp
qp
j=p
j=p
j=p
P
j
as p , so by the general principle of uniform convergence,
j=1 bj x
converges uniformly.
We may thus integrate term by term to get
!
Z 1 X
Z 1
X
X
bn
n
n
bn x
dx =
bn x dx =
.
n+1
0
n=0
n=0 0
n=0
P
P
bn xn /(n+1)
If
n=0
n=0 bn /(n+1) converges, then the power series
P
n
has radius of convergence at least 1 so the
Ppowernseries n=0 bn x has
radius of convergence at least 1 and so n=0 bn x converges uniformly
on [0, 1 ] and
!
Z 1 X
1 Z 1
X
X
bn (1 )n+1
n
n
bn x dx =
dx =
bn x
n+1
0
n=0
n=0 0
n=0
for all 1 > > 0. By the result of the last paragraph but one,
P
j+1
/(j + 1) converges uniformly on [0, 1]. The function
j=1 bj x
xj+1
bj
j+1
j=1
306
X
X
bj
bj xj
j+1
j+1
j=1
j=1
as x 1. Thus
as 0+.
1
0
X
n=0
bn x
X
bj
dx
j+1
j=1
j
1
for 0 x < 1.
j=0 bj /(j + 1) converges and
j=0 bj x = (1 + x)
Also
Z 1
1
dx = [log(1 + x)]10 = log 2.
0 1+x
j
Taking bP
series
2j = (1) , b2j+1 = 0 we know,
P by thej alternating
2 1
test, that j=0 bj /(j + 1) converges and
b
x
=
(1
+
x
)
for
j=0 j
0 x < 1. Also
Z 1
1 1
1
dx
=
tan x 0 = .
2
4
0 1+x
X
X
j+1
(1) log2j+1 e =
j=0
j=0
log 2
(1)j+1
=
= 1.
(j + 1) log 2
log 2
307
K265
Can be done in several ways. Here is one. Write pn = Pr(X = n).
Observe that, since pn 0 and
1 tn
= 1 + t + + tn1 n
0
1t
for 1 > t > 0, we have
PM
PN
N
M
j
X
X
1 tj
j=0 pj
j=0 pj t
pj
jpj EX
1t
1t
j=0
j=0
for M N and so
N
X
j=0
pj
1 tj
X (1) X (t)
EX
1t
1t
X (1) X (t)
X (1) X (t)
lim sup
EX.
1t
1t
t1
308
K266
Just as in K82, if m n,
n
m
Y
(1 + aj (z)) (1 + aj (z))
j=1
j=1
!
!
!
m
n
X
X
|aj (z)| 1
|aj (z)|
exp
exp
j=n+1
j=1
exp
X
j=1
Mj
exp
j=n+1
Mj
as n
QN so, by the general principle of uniform convergence, it follows
that j=1 (1 + aj (z)) converges uniformly.
309
K267
(i) We have
|z 2 n2 | R2 n2
for all |z| R so, by the previous question, SN (z) converges uniformly
on D(R) = {z : |z| < R}. Since the uniform limit of continuous
functions is continuous, S is continuous on D(R) for each R > 0 and
so on all of C.
(ii) Observe that (if z is not an integer)
Q
(N !)2 N
SN (z + 1)
r=N (r (1 + z))
=
Q
SN (z)
(N !)2 N
r=N (r z)
N +1z
1 (z 1)/N
=
1.
N z
1 z/N
Thus S(z + 1) = S(z) for z not an integer. Extend to all z by
continuity.
P
RecallQfrom K82 that, if
1 for all
n=1 |an | converges, and an 6= Q
n then
(1
+
a
)
=
6
0.
Thus,
if
N
>
|z|
we
know
that
n
n=1
n=N (1
Q
N 1
2 2
2 2
z n ) 6= 0 and so S(Z) = 0 if and only if z n=1 (1 z n ) = 0 i.e.
if and only if z is an integer.
=
[We might guess that S(z) = A sin z. Near 0, S(z) behaves like z
so, if the guess is correct, A = 1 .]
310
K268
P
n
(i) Take R > r > 0 As usual we note that
n=0 an r converges,
n
n
so an r 0 and there exists an M such that |an | M r . Thus, if
|z| < r/2,
q
q
X
X
n
n
a
z
=
a
z
n
n
n=0
n=N
|aN z N |
= |z|
|z|N
N
q
X
n=N +1
|aN | |z|
|an z n |
q
X
n=N +1
|an z
|aN | M |z|r N 1
N 1
nN 1
q
X
(|z|r)nN 1
n=N +1
).
q
|a ||z|N
X
N
n
a
z
n=0
2
for all q N and so
X
|a ||z|N
N
n
.
an z
n=0
!
!
311
K269
(i) Proof of K268 works.
(ii) Exercise 5.91.
(iii) sin x =
X
(1)n x2n+1
n=0
(2n + 1)!
312
K270
(i) Observe that
n!
1
n
1
n
=
+
+
r
(r 1)!(n r)! n r + 1 r
r1
n!
n+1
=
(n + 1) =
.
(r!(n + 1 r)!
r
(ii) Result true for n = 0. If the result is true for n then
(x + y)n+1 = x(x + y)n + (x + y)n y
n
X
n
n
n+1
xr y n+1r + y n+1
+
=x
+
r
r
1
r=1
r
n+1
X
n+1
=
y n+1r
x
r=0
n
n
nr1
nr
r
r
X n Y
X n Y
Y
Y
(y k) +
=
(x j) (y k)
(x j)
r j=0
r j=0
r=0
r=0
k=0
k=0
n
n1
r
nr
n
Y
X
Y
Y
Y
n
n
=
(x j) +
+
(x j) (y k) +
(y k)
r
r
1
r=0
j=0
j=0
k=0
k=0
nr
r1
n+1
Y
X n+1 Y
(x j) (y k)
=
r
r=0
j=0
k=0
313
K271
(i) Since |w| < 1, we can find a > 1 such that |w| < 1. We can
now find an N 1 such that
x j M + j
= 1 + M j 1 <
j
j
X 1 n1
(x j) (N + M )N = K, say
n=0 n!
j=0
for all |x| M and all n N . Thus
X 1 n1
Y
(x j)w n K(w)n
n=0 n!
j=0
n1
X
1 Y
(x j)w n
n! j=0
n=0
X
cn
f (x)f (y) =
j=0
with
cn =
n Y
r1
X
n
r=0
j=0
(x j)
nr1
Y
k=0
(y k).
314
K272
(i) We prove that AN |aN | inductively. The result is trivial for
N = 0. If the result is true for N = k 1, we observe that
kak =
k1
k1 X
X
n=0 m=0
and
kAk =
k1
k1 X
X
n=0 m=0
where the Pn.m are multinomials with all coefficients positive. Thus
k|ak |
k1 X
k1
X
n=0 m=0
k1 X
k1
X
n=0 m=0
X
dw
=K
(t)n (w)m
dt
n=0 m=0
!
!
X
X
=K
(t)n
(w)m
n=0
m=0
K
.
(1 t)(1 w)
K
dt
1 t
K
log(1 t) + C.
K
log(1 t)
and
w(t) =
315
X
Aj tj
w(t) =
j=0
n
F. By the arguments of (i), |an | An , so, by comparison,
n=0 an t
P
has radius of convergence at least . If we set u(t) = n=0 an tn , then,
since a0 = 0, we have u(0) = 0. By continuity, we can find with
0 < and |u(t)| < for |t| < . The required results can now be
read off.
316
K273
(i) Observe that, since f is continuous on [0, 4], it is uniformly continuous on [0, 4]. Since f is 2 periodic it is thus uniformly continuous
on R. Similarly g is bounded. Thus
Z
|f g(t + u) f g(t)| =
(f (t + u s) f (t s))g(s) ds
2
kgk
|xy||u|
as u 0.
(iii) Observe, for example, that making the substitution x = t s
we have
Z
Z t+
1
1
f (t s)g(s) ds =
f (x)g(t x) dx = g f (t).
f g(t) =
2
2 t
Fubini gives
f (g h)(t) =
=
=
=
=
Z
Z
1
g(s u)h(u) du ds
f (t s)
2
Z Z
1
f (t s)g(s u)h(u) du ds
(2)2
Z Z
1
f (t s)g(s u)h(u) ds du
(2)2
Z Z u
1
f (t u v)g(v)h(u) dv du
(2)2 u
Z
1
f g(t u)h(u) du = (f g) h(t).
2
/n
=
un (s)(f (t s) f (t)) ds
2 /n
Z /n
1
un (s) ds
sup |f (u) f (v)|
2 /n
|uv|/n
=
sup
|uv|/n
|f (u) f (v)| 0
317
as n .
(vii) Take un to be a three times continuously differentiable function
in (v).
318
K274
(i) Fubini gives
Z
Z
1
1
int
e
f (t s)g(s) ds dt
f[
g(n) =
2
2
Z Z
1
eint f (t s)g(s) ds dt
2
(2)
Z Z
1
eint f (t s)g(s) dt ds
=
(2)2
Z
Z
1
1
ins
in(ts)
=
e
g(s)
f (t s)e
dt ds
2
2
g(n)f(n).
(iii) We have
|f(n) g(n)| = |f[
g(n)| kf gk .
(v) If e f = f for all f then
f(n) = e[
f (n) = e(n)f(n)
for all f and all n. Take f (t) = eimt to see that e(m) = 1 for all m
contradicting the Riemann-Lebesgue lemma.
319
K275
(i) We have
1
f1 (n) =
2
g1 (t)
g1 (n + 1) + g1 (n 1) g1 (n + 1) g1 (n 1)
0.
2i
(ii) If we set
f2 (t)
if t/ is not an integer,
sin t
0
g2 (t) = f2 (0)
if t = 2n,
f 0 () if t = (2n + 1),
2
f20 () = g2 ()
g2 (t) =
t
sin(t )
f4 (n) =
f3 (t/2)eint dt
2
Z
Z
1
1 2
i2ns
f3 (s)e
ds
f3 (s)ei2ns dt = f4 (2n).
=
2
2
320
K276
(i) Let S(z) = z f (z). We have S 0 (z) = 1 f 0 (z) so |S 0 (z)| 1/2
for |z| and
|T z| |w| + |S(z) S(0)| |w| + |z| sup |S 0 (u)| /2 + /2 =
|u|
321
K277
(i) True. w is the solution.
(ii) False. f2 (z) is real so f2 (z) = i has no solution for > 0.
(iii) and (v) False. For (iii), observe <f3 (z) 0 so f3 (z) = has
no solution for > 0.
(iv) and (vi) True. For (vi), observe that we can find a with 1/6 >
> 0 such that |F 0 (z) 1| 1/3 for |z| 0 . Let X = {z : |z| }.
Set = 0 /3. If |w| write Sz = F (z) 1 and T z = Sz + |z|1/2 w.
Using the mean value inequality, we have
0
322
K278
(i) Observe that
f (x)f 00 (x)
f 0 (x) f (x)f 00 (x)
=
+
f 0 (x)
f 0 (x)2
f 0 (x)2
so, by the mean value inequality,
T 0 (x) = 1
|T a T b| |||a b|.
F (x)F 00 (x)
F 0 (x)2
and
F 00 (x) F (x)F 000 (x) 2F (x)(F 00 (x))2
+
,
F 0 (x)
F 0 (x)2
F 0 (x)3
so T (0) = T 0 (0) = 0 and T 00 (0) = F 00 (0)/F 0 (0). Thus
00
F
(0)
2
= |T x T (0) T 0 (0)x 1 T 00 (0)| (x)x2
T x
x
2
0
2F (0)
T 00 (x) =
(iii) We have quadratic convergence (roughly doubling of correct decimal places) with Newton rather than linear convergence (increasing the
number of correct decimal places by the same amount) as is implied by
the contraction mapping theorem. There may be problems if F 0 (0) is
small.
323
K279
Suppose f has a unique fixed point x. Then f (g(x)) = g(f (x)) =
g(x) and, since x is the unique fixed point of f , g(x) = x.
Let X = {1, 0, 1}, f (t) = t, g(t) = t. Then f g = gf = f , f has
only one fixed point but g has three.
Let X = {1, 1}, f (t) = t, g(t) = t. Then f g = gf = f , f has two
fixed points but g has none.
324
K280
We first prove uniqueness. If f (x) = x and f (y) = y then d(x, y) =
d(f (x), f (y)) Kd(x, y) so d(x, y) = 0 and x = y.
For existence choose x0 X and then choose inductively xn+1 X
such that that T xn+1 = xn (this uses surjectivity). We observe that
so
and precisely the same argument as in the contraction mapping theorem shows that xn z for some z X. Now z = T w for some w
and
d(z, w) d(xn , z) + d(xn , w) d(xn , z) + Kd(T xn , T w)
= d(xn , z) + Kd(xn1 , z) 0
325
K281
(i) If T (a) = a and T (b) = b, then kT (a) T (b)k = ka bk so
a = b.
(ii) Uniqueness much as in (i). Existence not implied see e.g. second
paragraph of K281.
(iii) Let E = {1, 1} be a subset of R. The mapping T : E E
given by T x = x preserves length but has no fixed point. The map
T : E E given by T x = x preserve length and has two fixed points.
kx0 T x0 k kT x0 T 2 x0 k
326
K282
Observe that, if X = [0, ) with the usual metric, the map f given
by x 7 x + 1 is an isometry but we cannot find an isometry g : X 0
such that f g(x) = x.
We now restrict ourselves to bijective isometries. Since the bijective
maps on X form a group under composition we need only show we
have a subgroup.
If f and g are isometries then
d(f g(x), f g(y)) = d(g(x), g(y)) = d(x, y)
so f 1 is a bijective isometry.
327
K283
(i) Observe that, since (X, d) has the BolzanoWeierstrass property,
we can find a sequence m(j) of strictly positive integers and an
X such that d(am(j) , ) 0. Take a sequence j(r) of strictly
positive integers such that m(j(r + 1)) > 3m(j(r)) + 1. Then setting
n0 (r) = m(j(r + 1)) m(j(r)) we have
d(an0 (r) , a) d(f m(j(r) an(r) , f m(j(r) a) = d(am(j(r+1)) , am(j(r)) )
as r .
d(am(j(r+1)) , ) + d(am(j(r)) , ) 0
d(a, f (b)) d(a, an(r(j)) ) + d(f (br(j) , f (b)) = d(a, an(r(j)) ) + d(br(j) , b) 0
as j .
328
K284
Observe that g f : X X is expansive so, by K283, g f is bijective and an isometry. Thus g is surjective and (since it is expansive)
injective. Thus f is bijective. Since
d(x, y) (f (x), f (y)) d(g f (x), g f (y)) = d(x, y),
f is an isometric bijection so (Y, ) has the BolzanoWeierstrass property and g is an isometric bijection.
If X = Y = {x R2 : kxk = 1} with the usual metric, then, taking
f = g to be rotation through /4, we see that f and g need not be
inverses.
If X = Y = [0, ) with the usual metric and we set f (x) = 2x,
g(x) = x + 1 then f is a distance increasing bijection but not distance
preserving and g is distance preserving but not a bijection.
329
K285
We note that x 7 f (x) and x 7 kf (x)k are continuous maps on a
closed bounded subset of Rn . Thus there exists an M with kf (x)k M
for all x A.
By convexity, g maps A into A. We have
kg(a) g(b)k = (1 )kf (a) f (b)k (1 )ka bk
so g is a contraction mapping and so, since A is closed and we are
dealing with a complete metric space, g has a fixed point z, say. We
observe that
kf (z) zk = kf (z) gzk
kf (x)k + ka0 k
(M + ka0 k).
Thus
inf{kf (a) ak : a A} = 0.
But x 7 kf (x) xk is a continuous map on a closed bounded subset of
Rn so attains its infimum. Thus kf (y) yk = 0 for some y A which
is thus a fixed point.
Let V = R, A = [2, 1] [1, 2] and T x = x. Then A is closed
and bounded and T is distance non-increasing with no fixed point.
Let V = R, A = (0, 1) and T x = x/2. Then A is convex and bounded
and T is distance decreasing with no fixed point.
Let V = A = R and T x = x + 1. Then A is closed and convex and
T is distance non-increasing with no fixed point.
In final paragraph take
A = {q R :
n
n
X
i=1
qi = 1 and qi 0 for 1 i n}
with
and f (q) = q
qj =
n
X
i=1
qi pij .
330
n X
n
n X
X
X
kf (a) f (b)k1 =
|ai bi |pij
(ai bi )pij
j=1 i=1
n
n X
X
i=1 j=1
n
X
i=1
|ai bi |pij
|ai bi | = ka bk1 .
j=1 i=1
331
K286
(i) Note that
|d(x, T x) d(y, T y) d(x, y) + d(T x, T y) 2d(x, y)
= d T N (z,T z) z, T N (z,T z) (T z)
< d(z, T z) = S(z)
332
K287
Observe that if q X then
m
X
i=1
qi pij 0 and
m X
m
X
qi pij =
j=1 i=1
m X
m
X
qi pij = 1.
i=1 j=1
m X
kT u T vk1 =
(ui vi )pij
j=1
i=1
m
m X
X
j=1 i=1
m X
m
X
i=1 j=1
|ui vi |pij
m
X
pnik pkj
k=1
m
m
X
X
N
|ui vi |pN
(u
v
)p
<
i
i ij
ij
i=1
i=1
for each j.
333
334
K288
We have
X
X
i=1 j=1
|xi pij | =
X
i=1 j=1
|xi |pij =
X
i=1
|xi | = kxk1
so Fubinis
P theorem tells us that i=1 xi pij converges for each j. Further i=1 |xi |pij converges and
X
X
X
X
|xi |pij
xi pij
j=1
i=1
j=1 i=1
X
X
i=1 j=1
X
X
qi pij =
qi pij = 1
j=1 i=1
and, trivially,
j=1 qi pij
i=1 j=1
0 so T q X.
X
i =
r pN (k) pri I pN (k) pIi > 0.
r=1
i = 21 i+1 + 14 i + 41 i1 .
so
2i+1 3i + i1 = 0
335
336
K289
(i) Observe that J > 0 and set h = J1 .
(ii) If e(j)k = 1 if k = j and e(j)k = 0 otherwise, then the e(k)
form a sequence in X with no convergent subsequence. (Observe that
ke(k) e(l)k1 = 2 for k 6= l.) Thus X does not have the Bolzano
Weierstrass property. However (see Exercise K218)
Y = {y : hi |yi | for i 1}
X
(T q)j =
qi pij
i=1
hi pij = hj
i=1
P
Now if q X and > 0 we can find M such that M
r=1 qr > 1 .
PM
Write u = q r=1 qr e(r). Then
!
M
M
M
X
X
qr + kT n uk1
kT n q k1 T n
qr e(r)
qr + 1
r=1
r=1
r=1
1
M
X
r=1
M
X
r=1
qr kT n e(r) k1 + + kuk1
qr kT n e(r) k1 + 2
337
K290
For uniqueness. Suppose T x = T y. Then T n x = T n1 T x =
T
T y = T n y so, by the contraction mapping theorem, x = y.
n1
Z t
n+1
n+1
|(T
h)(t) (T
k)(t)| (h(s), s) (k(s), s) ds
a
Z t
|M (h(s) k(s))| ds
a
Z t
M n+1
kh kk
(s a)n ds
n!
a
n+1
M
kh kk (t a)n+1 .
(n + 1)!
The induction can proceed.
We thus have
kT n h T n kk
1 n
M (b a)n kh kk .
n!
Since
1 n
M (b a)n 0
n!
(observe that un+1 /un 0) we can find an N such that
1
kT N h T N kk kh kk
2
for all h, k C([a, b]). Thus there exists a unique f C([a, b]) with
Z t
f (t) = g(t) +
(f (s), s) ds
un =
f 0 (t) = (f (t), t)
with initial condition f (a) = c.
338
K291
Consider Rn with its usual Euclidean norm and recall that it is
complete. Set
n
X
T x = y, with yi =
aij f (xj ) + bi
j=1
!2 1/2
n
n
X
X
aij (f (uj ) f (vj )
kT u T vk =
j=1
i=1
n
n
X
X
n
n
X
X
a2ij
j=1
= M ku vk
with K 0 and
j=1
i=1
i=1
a2ij
= Kku vk
2
n
X
j=1
K =M
n
X
(f (uj ) f (vj ))2
j=1
n
X
(M |uj vj |2 )
j=1
a2ij
!1/2
n
X
a2ij .
j=1
!!1/2
!!1/2
339
K292
A little fiddling around (we could write y dy = dx obtaining, in a
purely exploratory, non-rigorous manner (1 )y 1 = x + A) shows
that, if 0 < < 1, we have y(t) = 0 for all t, and y(t) = 0 for t < s,
y(t) = (1 )1/(1) (t s)1/(1) for t s are solutions. (We need
s 0 to satisfy y(0) = 0).
If 1 then using the mean value theorem
and this Lipschitz condition shows that the obvious solution y(t) = 0
for all t is unique.
340
K293
(i) If u > 0, then f (t, u) 0 for t > 0 so, by continuity, f (u, 0) 0.
Similarly f (t, u) 0 so f (u, 0) 0 Thus f (u, 0) = 0 for u > 0.
Similarly f (0, u) = 0 for u < 0 so f (0, 0) = 0 by continuity.
Observe, just as in the proof of Rolles theorem, that, if y is constant
on [0, c], then y(t) = 0 for t [0, c]. Otherwise, since a continuous function on a closed bounded interval is bounded and attains its bounds, y
must have a maximum or minimum at some (0, c) with y() 6= 0.
Without loss of generality, let be a maximum. Then y() > 0 so,
using the fact that at an interior maximum the derivative is zero,
0 = y 0 () = f (, y()) > 0
which is impossible. Thus y(t) = 0 for t [0, c] for all c > 0 so y(t) = 0
for t 0 and, similarly, for all t. Thus F has at most one solution.
Since y = 0 is a solution it is the unique solution.
(ii) If E = {(x, t) : k(x, t)k } with > 0 then E is closed and the
restriction of f is continuous on each of the three closed sets,
and
S
and so the restriction f |E is continuous on E = 3i=1 Ei . Thus f is
continuous at every (x, t) with k(x, t)k > with > 0 and so at every
(x, t) 6= (0, 0) Since |f (t, u)| 2 max(|x|, |t|), f is continuous at (0, 0)
(Lots of other ways to do this, many probably better.)
Now
y1 (t) =
t
2
f (s, s ) ds =
0
t
0
(2s) ds = y0 (t)
341
K294
Suppose, if possible, that y( ) 0 for some > 0. Let
= inf{s 0 : y(s) 0}.
which is absurd. Thus y(t) > 0 for all t, so y 0 (t) > 0 for all t, so y is
strictly increasing, so y 0 (t) y(0) = 1 so y(t) 1 + t for all t 0.
On (tj , tj+1 ), we have
y 0 (t) y(t) y((tj ) = 2j
and so
tj+1 tj 2j(1) .
Thus
tr =
r1
X
j=0
tj+1 tj
r1
X
j=0
2j(1)
1
1 21
1
.
1 21
342
K295
Observe that
y((r + 1)h) y(rh)
y 0 (rh).
h
Observe that
|f (t, u) f (t, v)| sup |f,2 (t, s)||u v| K|u v|
s
Ls2
2
whence
|y((n + 1)h) y(nh) f (nh, y(nh))h|
Lh2
2
and
|y((n + 1)h) yn+1 | |y(nh) yn | + |y((n + 1)h) yn+1 y(nh) + yn |
|y(nh) yn | + |y((n + 1)h) y(nh) f (nh, y(nh))h| + |f (nh, y(nh))h f (nh, y n )h|
Lh2
+ Kh|y(nh) yn |
2
Lh2
.
= (1 + Kh)|y(nh) yn | +
2
|y(nh) yn | +
Thus if
|y(nh) yn |
it follows that
Lh
(1 + Kh)n+1 1 ,
2K
Lh2
LhN
n+1
|y((n + 1)h) yn+1 | (1 + Kh)
(1 + Kh)
1 +
2K
2
Lh
(1 + Kh)n+2 1 .
2K
Since the result is true for n = 0 it is true inductively.
In particular
|y(N h) yN |
Lh
(1 + Kh)N +1 1 .
2K
343
K296*
No comments.
344
K297
As in K295,
|y((n + 1)h) y(nh) y 0 (nh)h|
Lh2
,
2
Lh2
+
2
so, by induction,
|y(nh) yn |
Thus
Lh
+
2K Kh
|y(a) yN |
(1 + Kh)n+1 1 .
Lh
+
2K Kh
(eaK 1).
The error due to machine inaccuracy increases directly with the number of computations.
(ii) and (iii) Recall that
1/2
2
A
A
1/2 1/2
B h
+ 2A1/2 B 1/2
+ Bh =
h
h1/2
345
K298
(i) Set f (t, y(t)) = g(t), M = L. Either redo calculations with K = 0
or observe that
Z
N
1
a
X
LhN eaK 1
[hN ]
g(rh)
=
|y(a)
y
|
g(t)
dt
0
2
K
n=0
aM h
aLhN
=
2
2
n=0
1 + sin 2N t
dt = = kGk .
2
2
2
346
K299
(i) The result is trivially true for n = 0. If it is true for n then
(uv)(n+1) (x) = ((uv)0 )(n) (x) = (u0 v + uv 0 )(n) (x)
n
n
X
X
n (n+1r)
n (nr)
(r)
=
u
(x)v (x) +
u
(x)v (r+1) (x)
r
r
r=0
r=0
n
X
n
n
(n+1)
u(n+1r) (x)v (r) (x) + u(x)v (n+1) (x)
+
=u
(x)v(x) +
r
1
r
r=1
n+1
X n+1
=
u(n+1r) (x)v (r) (x)
r
r=0
1 + x(1 + x2 )1/2
x + (1 + x2 )1/2
Since y(0) = 0 and y 0 (0) = 1 (by setting x = 0 in the first two displayed
equations in the statement of (ii)) we have y (2n) (0) = 0 and
y (2n+1) (0) = 2n n!.
Set un = y (n) (0)/n!. Then
|u2n+1 x2 | x2
|u2n1 | 2n + 1
which tends to zero if |x| 1 but diverges otherwise. Thus the Taylor
series for y has radius 1.
P
n
Now
term by term with in its circle
j=0 un x may be differentiated
P
n
of convergence so u(x) =
j=0 un x is a solution for the given differential equation with the same value as y at 0. Since the Lipschitz
conditions apply, y(t) = u(t) for |t| < 1 so
X
y (n) (0) n
y(x) =
x
n!
j=0
for |x| < 1.
347
K300
Since K is continuous on the closed bounded subset [0, 1]2 of R2 , it
is bounded so there exists an M with M K(s, t) for (s, t) [0, 1]2 .
If f C([0, 1]), set
(T f )(t) = g(t) +
We observe that T f C([0, 1]) since K is uniformly continuous (because K is continuous on the closed bounded subset [0, 1]2 ) and so
Z 1
kf k
as u v.
sup
kxyk|uv|
|K(x) K(y)| 0
Z 1
||M kf hk
348
K301
(ii) Observe that
u (t)
d 01
u1 (t)
dt 00
u1 (t)
0
u (t)
1
= u01 (t)
u00 (t)
1
u2 (t)
u02 (t)
u00
2 (t)
u02 (t)
u02 (t)
u00
2 (t)
u3 (t)
u03 (t)
u00
3 (t)
u03 (t) u1 (t)
0
u3 (t) + u00
1 (t)
u00 (t)
u00
3 (t)
1
u2 (t)
u00
2 (t)
u00
2 (t)
u1 (t)
u2 (t)
u3 (t)
0
0
0
u2 (t)
u3 (t)
= 0 + 0 + u1 (t)
u000 (t) u000 (t) u000 (t)
1
2
3
u
(t)
1
u01 (t)
=
a(t)u00 (t) b(t)u0 (t) c(t)u1 (t)
1
1
0
0
0
u3 (t) u1 (t)
00
u3 (t) + u01 (t)
u000 (t)
u00
3 (t)
1
u2 (t)
u02 (t)
u000
2 (t)
u3 (t)
0
u3 (t)
u000
3 (t)
u2 (t)
u02 (t)
0
a(t)u00
2 (t) b(t)u2 (t) c(t)u2 (t)
u3 (t)
u03 (t)
00
0
a(t)u3 (t) b(t)u3 (t) c(t)u3 (t)
349
K302
Observe that, since W (a) 6= 0, we have g(a) 6= 0. Similarly g(b) 6= 0.
Thus, if g has no zeros in (a, b), g has no zeros in [a, b]. Thus f /g
is well defined continuous function on [a, b] which is differentiable on
(a, b). By Rolles theorem there exists a c (a, b) with
0
f
f 0 (c)g(c) g 0 (c)f (c)
W (c)
0=
(c) =
=
.
g
(g 0 (c))2
(g 0 (c))2
350
K303
(i) Without loss of generality suppose that 1 6= 0. Then without
loss of generality we may suppose 1 = 1 and set 2 = . Then
f1 f2
W (f1 , f2 )(x) = det
f1 f20
f1 + f2 f2
= det 0
f1 + f20 f20
0 f2
= 0.
= det
0 f20
However, if g1 (x) = (x 1)2 for x 1 and g2 (x) = g1 (x) (we could
play the same trick with infinitely differentiable functions) then g1 and
g2 are continuously differentiable and their Wronskian vanishes everywhere but they are not linearly dependent.
(ii) By Exercise 12.24 applied to the vectorial equation
d y(t)
v1 (t)
,
=
a1 (t)v(t) a2 (t)y1 (t)
dt v(t)
it follows that if a1 and a2 are continuous (so bounded on [a, b]), the
equation
y 00 (x) + a1 (x)y 0 (x) + a2 y(x) = 0
has exactly one solution with y(a) = A and y 0 (a) = B.
If W (f1 , f2 )(x) never vanishes then setting
1
f1 (x) f2 (x)
a1 (x) = W (f1 , f2 )(x)
det 00
f1 (x) f200 (x)
and
1
f1 (x) f20 (x)
a2 (x) = W (f1 , f2 )(x)
det 00
f1 (x) f200 (x)
351
352
K304
Observe that
0 = (u00 v + 2v 0 u0 + v 00 u) + p(u0 v + v 0 u) + qvu
= v(u00 + pu0 + qu) + v 00 u + v 0 (2u0 + pu) = v 00 u + v 0 (2u0 + pu)
so, writing w = v 0 we have
Z x
y(x) = u(x)v(x) = u(x) A + B
w(t) dt
0
t
0
exp(x2 x) dx.
353
K305
(i) We have
f = (u001 y1 + 2u01 y10 + u1 y100 + p(u01 y1 + u1 y10 ) + qu1 y1 )
+ (u002 y2 + 2u02 y20 + u2 y200 ) + p(u02 y2 + u2 y20 ) + qu2 y2
= 2u01 y10 + u001 y1 + pu01 y1 + 2u02 y20 + u002 y2 + pu02 y2 .
(ii) If we differentiate , then we get
so becomes
ex
ex
ex
x
0
=
+
e
1
and
u
(x)
=
.
2
1 + ex
1 + ex
1 + ex
Thus
u1 (x) = A1 + log(1 + ex ) ex + x and u2 (x) = A2 + log(1 + ex )
354
We get
f = y100 u1 + y200 u2 + y300 u3 .
Since the Wronskian never vanishes, we can solve the last three equations to find u01 , u02 and u03 . Integrating now gives u1 , u2 , u3 and thus
the solution.
355
K306
Observe that since K is continuous on the closed bounded set S it
is uniformly continuous. Thus
Z 1
kf k
sup
kzwk|st|
|K(z) K(w)| 0
so kTKn k n
0.
1
1/n
2
356
K307
(ii) We apply Exercise 12.24 to the vectorial equation
d y1 (t)
v1 (t)
=
a(t)v1 (t) b(t)y1 (t)
dt v1 (t)
to obtain the existence and uniqueness of y1 .
R1
Rt
(iii) We have (noting that t u(x) dx = 1 u(x) dx)
Z 1
Z t
0
,2 (s, t)f (s) ds
t)f (t) +
H
H,2 (s, t)f (s) ds H(t,
y (t) = H(t, t)f (t) +
t
0
Z t
Z 1
=
H,2 (s, t)f (s) ds +
H,2 (s, t)f (s) ds
0
and
Summing gives
= f (t).
Further
y(0) =
1
0
1)f (s) ds =
H(s,
0 ds = 0
0
357
K308
We seek a twice continuously differentiable function y(x) = G(x, t)
which is four times differentiable (using left and right derivatives at
end points) on [0, t) and (t, 1] and satisfies y(0) = y 0 (0) = 0,
y (4) (x) k 4 y(x) = 0 for x [0, t)
whilst
This gives
0 = c + d, 0 = b + d
so
y(x) = A(sinh kx sin kx) + B(cosh kx cos kx) for x [t, 1]
y(x) = C(sinh k(1 x) sin k(1 x)) + D(cosh k(1 x) cos k(1 x)) for x (t, 1].
The continuity of the zeroth, first
condition on the discontinuity for the
in the next paragraph)
M A B C D = 0 0 0 1 so
where
sinh kt sin kt
k cosh kt cos kt
M =
k2 sinh kt sin kt
3
k cosh kt + cos kt
cosh kt cos kt
k sinh kt + sin kt
k2 cosh kt cos kt
k3 sinh kt + sin kt
A B C D
T
= M 1 0 0 0 1
cosh k(1 t) cos k(1 t)
k sinh k(1 t) + sin k(1 t)
.
3
k sinh k(1 t) sin k(1 t)
and
then writing
358
But
sinh
k
sin
k
cosh
k
cos
k
det
cosh k cos k
sinh k + sin k
359
K309
We seek a continuous function y(x) = G(x, t) which is twice differentiable (using left and right derivatives at end points) on [0, t) and
(t, 1]) and satisfies y(0) = y 0 (0) = 0,
y 00 (x) + 2y 0 (x) + 2 y(x) = 0 for x [0, t)
and
whilst
y 0 (t+) y 0 (t) = 1.
This gives
and []
for x [t, ).
0
if x t
1 (xt)
e
sinh((x t))
is
y(t) =
0
2
as
y(t) =
t
0
360
as
y(t) =
If is large the pointer takes a long time to return to close to equilibrium. If is small the pointer overshoots and we have violent oscillation so it also takes a long time before the pointer remains close to
equilibrium. As a rule of thumb close to is preferred.
361
K310
(i) Observe that, if m n
m
m
X
X
kT kn+1
r
0
kSn Sm k
kT k
kT kr
1
kT
k
r=n+1
r=n+1
n
X
r=0
kT kr
(1 kT k)1
kT kn+1 + kS Sn kkI T k 0
so A is invertible. Thus
but
so
1 > kB CkkB 1 k kA Ik 0
kC 1 k kB 1 k(1 kB CkkB 1 k)1 .
= {C L(U, U ) : kB Ck < kB 1 k1 } E
362
so E is open. If C then
as kC Bk 0. Thus is continuous.
(iv) Calculations as in (ii).
(v) Use the chain rule or (essentially the same) observe that
(B + H) = (B + H)1 = (B(I + B 1 H))1
= (I + B 1 H)1 B 1
= (I B 1 H + (B 1 H)kHk)B 1
= B 1 B 1 HB 1 + (H)kHk
as kHk 0.
363
K311
(ii) If kAk = 0, then A = 0 and there is nothing to prove so we
assume kAk 6= 0. exists since any non-empty set of real numbers
bounded below has an infimum. Now choose > 0. We can find a j
such that kAj k1/j + . Thus
1/(jk+r)
n 1/n
(iii) and (iv) If (A) < 1, set = ((A) + 1)/2. There exists an
n
N such that kAn k1/n and so kAkP
n for n N . The kind
of arguments used in K310 show that j=0 Aj converges and that the
limit is the inverse of I A.
n 1/n
If (A) > 1 then there exists an N such
1 for all
P thatj kA k
n N so, by the easy part of the GPC, j=0 A fails to converge.
364
K312
(i) k()n k1/n = ||kn k1/n
(ii) Take a basis ej and let be the unique linear map with ej =
ej+1 for 1 j m 1, em = 0. We have () = 0.
(iii) The
Pm eigenvectors ej with
Q eigenvalues j are linearly independent. (If m
e
=
0
apply
j=1 j j
j6=k ( j ) to both sides.) They thus
form a basis. If the standard basis is uj , let be the unique linear map
with ej = uj .
Observe that
k(1 )n k1/n = kn 1 k1/n k1 k1/n kn k1/n kk1/n ()
so (1 ) (). Similarly
() = ( 1 (1 )) ( 1 )
365
K313
(i) This is just the formula for change of basis.
(ii) Here and elsewhere we use the fact that
m2 sup |aij | kk sup |aij |
1i,jm
1i,jm
when (aij ) is the matrix of with respect to the matrix. (In fact we
merely use
K sup |aij | kk K 1 sup |aij |
1i,jm
1i,jm
for some K and this follows from the fact that all norms on a finite
dimensional space are Lipschitz equivalent.)
Given any > 0, we can certainly find cij with cij = bij for i 6= j,
|cii bii | < and the cii all distinct.
(iii) Use the notation of (i). Use (ii) to find n with m distinct
eigenvalues and kn k 0. Let n = 1 n . We know that n
has the same eigenvalues as n and
k n k = k1 ( n )k k 1 kk n kkk 0
as n 0.
366
K314
(i) We know that, given any > 0, we can find a() 1 and b() 1
such that
kr k a()(() + )r and k r k b()(() + )r
X n
k( + )n k =
j nj
j
j=0
n
X
n
kj kk nj k
j
j=0
n
X
n
a()b()
(() + )j (() + )nj
j
j=0
= a()b()(() + () + 2)n
( + ) (() + () + 2).
m
m
X
X
j
n
n
xj ej
kej k 0
j=1
j=1
so () . Thus () = |1 |.
( + r ) () + (r ) () + kr k = () + 1/r
() = ( + r r ) ( + r ) + 1/r
367
368
K315
(i) Observe that
n
xn = x0 +
n1
X
j b
j=0
so xn ( ) b.
369
K316
We require (I A) < 1. If kI Ak is small, convergence will be
rapid with the error roughly multiplied by kI Ak at each step.
Now the determinant of a matrix is the product of its eigenvalues (multiple roots counted multiply) (consider the coefficient of t0 in det(tI
A)) so, if | det H| > 1, we have (H) > 1. Thus the scheme fails for
< 0 or > 2.
370
K317
From K310 we know that the map 7 1 is continuous on the open
subset of L(U, U ) where the inverse is defined. Thus thus composition
of the maps
is continuous.
x 7 Df (x), 7 1
371
K318
(i) Observe that, if m n,
m
m
n
X r X
r
X
kkr
r=0 r!
r! r=n+1 r!
r=0
P
Pn
r
r
(since m
r=0 (x /r!) converges for all x) so, by completeness,
r=0 r!
converges.
n
n
X r X
r s
r
r
r s
X
X
X
( + )
r=0 r! r=0 r!
r!
r! s! 0r+sn r,s0 r! s!
r=0
0r,sn
r s
X
r!
s!
r+s>n, 0r,sn
kkr kks
r! s!
0r,sn
r+s>n,
kkr kks
r! s!
r,s0
2nr+s>n,
2n
X
1
=
(kk + |k)k 0
k!
k=n+1
n
n
X
X
r
r
h
h
h =
r!
r!
r=2
r=0
n
X
|h|kkr
r=2
r!
|h| kk
X
r=0
Thus
exp(h) = + h + h2 (h)h2
where k (h)k kk2 for 0 < khk < kk/2. Similarly
exp(h) = + h + h2 (h)h2
where k (h)k kk2 for 0 < khk < kk/2.
372
We thus have
kh2 (exp(h) exp(h) exp(h) exp(h)) ( )k
( + h + h2 (h)h2 )( + h + h2 (h)h2 )
( )k
n
X
n1
kkj kknj
n
j
1
j=1
n1
X
n1
= kk
n
kkr kkn1r
r
r=0
= nkk(kk + kk)n1
Thus
N
N
N
X ( + )n X
()n
X 1
k( + )n n k
n=0
n!
n!
n!
n=0
n=1
kk
= kk
so that
N
X
n
(kk + kk)n1
n!
n=1
N
1
X
n=0
1
(kk + kk)n
n!
= kk exp(kk + kk)
k exp( + ) exp()k kk exp(kk + kk) 0
373
K319
(ii) If L(U, U ) and L(U, U ) write
n ()() = n1 + n2 + 2 n3 + + n1 .
n
X
n2
kkj kknj
n(n 1)
j2
j=2
n2
X
n2
2
kkr kkn2r
= n(n 1)kk
r
r=0
= n(n 1)kk2 (kk + kk)n2
PN
n ()
j=1 n!
n!
N
X
nkkn1
j=1
n!
ekk
so
converges in the space of linear maps from L(U, U ) to
L(U, U ) in the appropriate operator norm. Call the limit ().
We have
N
N
N
N
X ( + )n X
()n X n ()
X 1
k( + )n n n ()()k
n=0
n!
n!
n!
n!
n=0
n=1
n=1
N
X
1
= kk2
2
so
N
2
X
n=0
1
(kk + kk)n
n!
kk exp(kk + kk)
k exp( + ) exp() ()k kk2 exp(kk + kk)
374
K320
Use the fact that
m2 sup |aij | kk sup |aij |
1i,jm
1i,jm
when (aij ) is the matrix of with respect to the matrix. (Or merely
use
K sup |aij | kk K 1 sup |aij |
1i,jm
1i,jm
for some K and this follows from the fact that all norms on a finite
dimensional space are Lipschitz equivalent.)
If A is upper triangular then Ar is upper triangular with (j, j) th
entry the rth power of ajj the (j, j) th entry of A. Thus exp A is upper
triangular with (j, j) th entry exp ajj . Thus
Y
X
det(exp A) =
exp ajj = exp
ajj = exp Trace A
and, since any has an upper triangular representation and det and
Trace depend on and not the representation chosen,
det(exp ) = exp Trace .
375
K321
(ii) Note, for example, that, writing r(B) for the rank of B, we have
2 r(A) r(A) r(A2 ) r(A2 ) r(A3 ) r(A3 ) r(A4 ).
Since A4 = 0 and A2 6= 0 we have r(A2 ) r(A3 ) 1 so 2 3 which is
absurd.
(iii) We have
a2 + bc = 1
b(a + d) = 0
c(a + d) = 0
d2 + bc = 1
so either a + d 6= 0, in which case b = c = 0 and a = d = 1 so A = I
or a + d = 0 and, either a2 = 1 in which case bc = 0 and
1 0
1
or A =
A=
1
0 1
for some or a2 6= 1 and
A=
a
b
b1 (1 a2 ) a
1 0
.
A=
0 1
(iv) Observe that S(I) = I and that S has a continuous derivative
and that DS(I)(C) = 2C so DS(I) is invertible. The result now follows
from the inverse function theorem.
(v) The result fails for Y = 0 since
1 0
1
is a solution for all .
376
u v
H=
w x
then
DS(Y )H = Y H + Y H
u v
u
v
+
=
w x
w x
u 0
=2
0 x
377
K322
We have
hx, xi = hx, T xi = hx, xi = hx, xi.
So hx, xi = 0.
If 2 x = x, then
hx, xi = hx, xi = h2 x, xi = hx, xi 0.
Since a real cubic has at least one real root, the characteristic equation det(t ) = 0 must have at least one real root. Thus has an
eigenvector e3 of norm 1. By the first paragraph e3 = 0. We observe
that e3 is an eigenvector of 2 and so since 2 is symmetric we can find
e1 and e2 such that the ej form an orthonormal basis. We know that
e1 is perpendicular to e1 . Also
he1 , e3 i = he1 , e3 i = 0
he2 , e1 i = he2 , e1 i =
With respect to the basis ej the linear map exp has the
representation
P (1)r 2r+1
n P (1)r 2r
0 0
r=0
r=0 (2r+1)!
(2r)!
X 1
P
P (1)
r 2r
(1)r 2r+1
0 0 =
r=0 (2r+1)!
r=0
(2r)!
n!
n=0
0 0 0
0
0
cos sin 0
= sin cos 0
0
0
1
matrix
0
0
1
378
K323
We have
= ( + T )/2 + ( T )/2.
The first term is symmetric the second antisymmetric. Note that the
decomposition is unique (if = + with symmetric and antisymmetric then ( + T )/2 = ).
If is as stated
k( T )/2k k( )/2k + k( )T /2k = k k <
so
But
so
2 = 2 + ( ) + 2 2 .
k( + T )/2k k( )/2k + k( )T /2k = k k <
2kk kk2 + 2kkk + 2 2 2 + 22 + 2 = 42
379
K324
We call matrices U with U U T orthogonal. If, in addition, det U = 1
we say it is special orthogonal.
(i) Observe that
1 = det U U T = det U det U T = (det U )2 .
(ii) Observe that
!T
N
X Bj
j=0
j!
N
X
(B T )j
j=0
j!
N
X
(B)j
j=0
j!
so
so
i.e. AB = BA.
s=1
s=1
380
a b
A=
b a
cos sin
,
A=k
sin cos
1 0
cos sin
1 0
cos sin
=
0 1
sin cos
0 1
sin cos
so under the new definition only matrices of the form I can be isotropic
(and it is easily checked that they are).
381
K325
The area of the inscribed n-gon is given by
n
n
X
k
r2 X
k
2
=
A=
sin k
r sin cos
2
2
2 k=1
k=1
subject to
n
X
k = 2.
k=1
n
n
X
r2 X
k = 2
sin k
L=
2 k=1
k=1
subject to
n
X
k = 2.
k=1
k
2
382
K326
Observe that the set
E = {x Rn : xj 0 for all j and
xpj = c}
j=1
P
p
is closed (observe that the map x 7
j=1 xj is continuous) and
bounded (if x E then c1/p xP
j 0.) Thus the continuous function f : E R given by f (x) = j=1 xj yj attains a maximum. This
maximum does not occur if any of the xj = 0 (for, if xJ = 0, then
if delta is very small and positive and we take x0J = and take so
that taking x0r = xr if xr 6= 0, x0r = 0 if xr = 0 and r 6= J we
have x0 E, we see that = O( p ) and f (x0 ) > f (x)). Thus, if the
Lagrange method gives us a unique stationary point with xj 6= 0 for
all j, it will indeed be the maximum.
Form the Lagrangian
L=
n
X
j=1
xj yj
n
X
xpj .
j=1
At a stationary point,
0=
L
= yj pxjp1 ,
xj
n
n
X
X
X
1+q/p
xj yj = c1/p
yj
yjq
j=1
j=1
=c
c1/p
1/p
n
X
j=1
yjq
yjq
!1/q
j=1
n
X
j=1
n
X
j=1
yjq
!1/p
383
Thus
c1/p
Pn
n
X
Thus
n
X
j=1
tpj
tpj
!1/q
n
X
t j yj
j=1
yjq
j=1
p
j=1 tj
whenever tj 0 and
n
X
j=1
yjq
!1/q
kyjq .
n
X
t j yj
j=1
tj
t j yj
yj
j=1
j=1
j=1
tpj
kyjq .
Thus if t, y 6= 0 then
n
! n
!1/q
n
X
X
X
p
q
|tj |
|yj |
|tj yj |
j=1
j=1
p
j=1
384
K327
(iii) The sum of the squares of the four sides of a parallelogram equals
the sum of the squares of the two diagonals.
(iv) Observe that
kxk kx yk + kyk
so
= 4kxkkyk
(v) Observe, for example, that the parallelogram law fails for the
uniform norm when we consider f (x) = max(1 4x, 0) and g(x) =
f (1 x).
385
K328
(ii) We have
ku + v + wk2 + ku + v wk2 = k(u + v) + wk2 + k(u + v) wk2
= 2ku + vk2 + 2kwk2 .
Thus
4(h(u + w, vi + h(u w, vi)
= 8hu, vi
so (1) holds.
= h2u, vi
= hx + y, vi.
(iii) A simple induction now gives hnx, yi = nhx, yi for n a positive
integer. The remark that
hx, yi = 41 (kx yk2 kx + yk2 = hx, yi
Now observe that the Cauchy-Schwarz inequality holds (see K327 (iv))
so
|hx, yi hk x, yi| = |h( k )x, yi| k( k )xkkyk 0
386
K329*
No comments.
387
K330
(i) Since |d(e, t) d(e, s)| d(s, t), the function fe is continuous.
Since |d(e, t) d(e0 , t)| d(e0 , e) it is bounded.
(ii) We have
and
388
K331
(ii) We have xn1 = ME (n1 , x) = M (n1 , x) M (0, x) as n 0
by the continuity of M so M (0, x) = 0 for all x 6= 0. In particular
M (0, 1) = M (0, 2), which is incompatible with group laws (ab = cb
implies a = b).
(iii) As for (ii).
(iv) If xn is Cauchy for d, then d(xn , 1) is bounded so there exist
A > 1 such that A1 xn A. Observe that (by the mean value
theorem) there exists a K > 1 such that K 1 |xy| d(x, y) K|xy|
for all |x|, |y| [A1 , A]. Take M (x, y) = xy.
389
K332
(i) If E = {0} then we are done. If not, then since x E implies
x E, we know that {x E : x > 0} is non-empty.
Let = inf{x E : x > 0}. If = 0 then we can find xn E with
0 < xn < 1/n. We have
E
n=1
{mxn : m Z}
390
391
K333
Second paragraph. Let X = [1, 1] with the usual metric d. Let
E = X \ {0} and let Y = E with the restriction of d to Y . Let
f (x) = x so f is certainly uniformly continuous. Suppose f : (X, d)
(Y, ) is continuous and f(x) = f (x) for all x E. If f(x) = y then
|x y| = (f(x), f(0)) 0
392
K334
Observe that
f (x) f (y)
0
xy
as y x. Thus f is everywhere differentiable with derivative zero and
so is constant.
If we replace R by Q, then f is uniformly continuous on Q and so
can be extended to a continuous function f on R. If x, y R and
x 6= y choose xn , yn Q with xn x and yn y. Since
we have
so, allowing n ,
The function in Example 1.3 satisfies the condition:Given x Q there exists a (x) > 0 with
for all x, Q with |x y| < (x) but the (x) is not uniform.
393
K335
This is the statement that a complete metric space is totally bounded
(see e.g. Section 11.2).
S
S
If Fn is finite set such that xFn B(x, 1/n) = X, then E =
n=1 Fn
is a countable dense subset.
394
K336
(i) and (ii). Let xn be a sequence of points forming a dense subset.
If U is open, consider xn . If xn
/ U , we do nothing, If xn U , then
put n . Since U is open we can find a n0 > 0 such that d(xn , y) > n0
for all y
/ U . If U = X set n = 1. Otherwise inf yU
/ d(xn , y) is a well
defined strictly positive number. Set n = inf yU
/ d(xn , y)/2.
S
We claim that Sn B(xn , n ) = U . Observe that if n then
B(xn , n ) U so n B(xn , n ) U . Now suppose x U . Since U
is open we can find a > 0 with 1 > such that B(x, ) U . Since
the sequence xn is dense we can find an N such that d(xN , x) < /4.
Automatically
inf yU
/ d(xN , y) > 3/4 so N 3/8 and x B(xN , N ).
S
Thus n B(xn , n ) = U .
(iii) The space R with the discrete metric d(x, y) = 1 if x 6= y is
complete. Any open ball of the form B(x, r) with r > 1 is R itself. Any
open ball of the form B(x, r) with r < 1 is {x}. Thus the countable
unions of open balls are countable subsets of R and R itself. But every
set is open so any uncountable set (e.g. {x : x > 0}) which is not the
whole space is a counterexample.
S
n
395
K337
(i) If x U , we can find a > 0 such that (x , x + ) U . Thus
x x.
If x y, then x, y (a, b) U so y x
If x y and y z, then x, y (a, b) U and y, z (c, d) U .
We then have (a, b) (c, d) an open interval (since both (a, b) and (c, d)
contain y) lying within U so x z.
(ii) Since x [x], [x] is non-empty. If [x] is bounded above look at
= sup[x]. If U then we can find > 0 such that (2, +2)
U . We can find [x] such that so x , and
+ . Thus x + contradicting the definition of . By
reductio ad absurdum,
{y [x] : y x} [x, ).
[
[
{y [x] : y x}
[x, cn )
[x, n ) = [x, ).
n=1
n=1
Thus
{y [x] : y x} = [x, ).
If [x] is unbounded above we can find n n with n x so we can
find (an , cn ) U with x, n (an , cn ). Thus
[
[
{y [x] : y x}
[x, cn )
[x, )
n=1
n=1
so
{y [x] : y x} = [x, ).
(iii) Take C to be the set of equivalence classes.
(v) Every open interval contains a rational. Thus each element of C
contains a rational which belongs to no other element so C is countable.
(Or say that the map f Q C given by f : (q) = [q] is surjective.)
If D is an open disc {x : kx x0 k = r} and kx x0 k = r then any
open disc containing y intersects D but y
/ D.
396
K338
(i) Observe that
n
X
j=1
n
X
j=1
|f (xj ) f (xj1 )| +
n
X
j=1
|g(xj ) g(xj1 )|
|fp (t) fq (t)| |fp (0) fq (0)| + V (fp V (fq ) = kfp fq kBV
so fn (t) is Cauchy in R for each t. Thus fn f (t) for some f (t). We
now show that f BV . Since any Cauchy sequence is bounded we
can find a K with K kfn kBV for all n. Now suppose 0 = x0 x1
xN = 1, Then
N
X
j=1
|f (xj ) f (xj1 )|
N
X
j=1
N
X
j=1
N
X
j=1
397
N
X
j=1
|(f fm )(0)| +
+
N
X
j=1
|(f fm )(0)| +
|(f fm )(0)| +
N
X
j=1
N
X
j=1
sup
t[xj1 ,xj ]
|g(t)||xj xj1 |.
398
Thus if we take
if 0 x 21 21 n1 ,
1
gn (x) = 4n(x 1/2) if 21 21 n1 < x < 21 + 21 n1 ,
1
if 21 + 21 n1 x 1,
Rt
and set Gn = 1/2 gn (x) dx, G(t) = |t 21 | we have Gn C 1 , G BV
and kG Gn kBV 0 as n but G
/ C 1.
|g(yr ) g(yr1 )|
n
X
j=1
rA
rA
rA
(f (yr ) f (yr1 ))
=1
1
1
X
rA
/
(f (yr ) f (yr1 ))
X
rA
X
rA
rA
|g(yr ) g(yr1 )|
X
rA
|g(yr ) g(yr1 )|
|g(yr ) g(yr1 )|
M |yr yr1 |
1 M (2/3)m = V (f ) M (2/3)m
399
and
X
rA
/
Thus
X
r
|g(yr ) g(yr1 )|
V (g) M (2/3)m .
V (f + g)
X
rA
X
rA
X
rA
/
|g(yr ) g(yr1 )|
|g(yr ) g(yr1 )|
V (f ) + V (g) 2M (2/3)m .
X
rA
/
X
rA
/
X
rA
/
|f (yr ) f (yr1 )|
(f (yr ) f (yr1 ))
kg f kBV V (f g) V (f ) + V (g) V (f ) = 1
so C is not dense in BV .
400
K339
(iii) f : R (1, 1) given by f (x) = (2/) tan1 x will do.
(iv) f : I J given by f (x) = exp(ix) will do. Observe that
(1, 1) is a dense subset of [1, 1] which is complete. Observe that J
is a dense subset of {z C : |z| = 1} which is complete.
The Cauchy sequences xn = 1 1/n and yn = 1 + 1/n have no
limits in I. But |xn yn | 9 0 so they can not have the same limit in
the completion.
401
K340
(i) Suppose, if possible, that f 1 is not uniformly continuous. Then
we can find an > 0 and un , vn Y such that (un , vn ) 0 but
d(f 1 (un ), f 1 (vn )) > . By the BolzanoWeierstrass property of X,
applied twice to obtain a subsequence and then a subsequence of that
subsequence, we can find n(j) 0 and , X such that
d(f 1 (un(j) ), ) 0 and d(f 1 (vn(j) ), ) 0.
which is absurd.
(iii) If f (x) = x1/3 , the mean value theorem shows that |f (x)
f (y)| |x y|/3 if |x|, |y| 1. Since f is continuous, it is uniformly
continuous on [2, 2] so f is uniformly continuous on R. However
f 1 (x) = x3 so f (x + ) f (x) 3x 2 as x for all > 0,
so f 1 is not uniformly continuous.
(iv) Observe that (x, y) < 1/2 implies x = y implies d(f (x), f (y)) =
0 but that d(1/n, 0) 0 and (f 1 (1/n), f 1 (0)) = (1/n, 0) = 1 9 0.
402
K341
Since (X, d) has the BolzanoWeierstrass property, there exists a K
such that d(x, y) K for all x, y X (direct
or use total boundP proof
j
edness), thus, by the Weierstrass M-test, j=1 (2 d(x, xj ))2 converges
and f (x) l2 .
Observe that
kf (x) f (y)k22 =
so f is continuous.
j=1
X
j=1
403
K342
(i) We can have (X, d) complete but (Y, ) not. Let X = R with the
usual metric d, Y = (/2, /2) with the usual metric , and f (x) =
tan1 (x). Using the mean value inequality |f (x) f (y)| |x y|.
We can have (Y, ) complete but (X, d) not. Let X = (4, 4) with
the usual metric d and y = [1, 1] with the usual metric . Take
f (x) = sin x and observe that by the mean value inequality |f (u)
f (v)| |u v|.
(ii) If (X, d) is complete (Y, ) must be. Let yn be Cauchy in Y .
Since f is surjective we can find xn X with f (xn ) = yn . Since
d(xn , xm ) K 1 (f (xn ), f (xm ) = (yn , ym ),
We can have (Y, ) complete but (X, d)) not. Let Y = R with
the usual metric , and X = (/2, /2) with the usual metric d
f (x) = tan(x). Using the mean value theorem, |f (x) f (y)| |x y|.
404
K343
Observe that
4 4 4
405
K344
(ii) Observe that the open unit ball E centre 1 consists of the odd
numbers. If r E and s
/ E then d(r, s) = 1 so E is closed.
406
K345
Recall that we can find E : R R infinitely differentiable with
E(x) = 0 for x < 0 and E strictly increasing on [0, ). Thus, if we
consider F (x, y) = E(x + 1)E(x + 1)E(y + 1)E(y + 1), we have F
infinitely differentiable, F (0, 0) 6= 0 and F (x, y) = 0 if |x| 1 and/or
|y| 1.
Set (xn , yn ) = (n2 , n4 ) and n = (10n)8 . Note that, if we write
An = [xn 2n , xn + 2n ] [yn 2n , yn + 2n , ]
m=1
1
1
(y ym )).
n1 F (4m
(x xm ), 4m