AW: a sequence from analysis
Pfoertner, Hugo
Hugo.Pfoertner at muc.mtu.de
Tue Feb 17 11:56:14 CET 2004
-----Ursprüngliche Nachricht-----
Von: N. J. A. Sloane [mailto:njas at research.att.com]
Gesendet am: 16 February, 2004 21:17
An: seqfan at ext.jussieu.fr
Betreff: a sequence from analysis
The following are a very nice pair of sequences
for which i could use more terms!
Thanks
Neil
%I A088403
%S A088403 0,227
%N A088403 Values m_0 = 0, m_1, m_2, ... associated with series T shown
below.
%C A088403 T = 1
%C A088403 - (1/2 + 1/4 + 1/6 + ... + 1/(2m_1))
%C A088403 + (1/3 + 1/5 + 1/7 + ... + 1/(2m_2+1))
%C A088403 - (1/(2m_1+2) + 1/(2m_1+4) + ... + 1/(2m_3)
%C A088403 + (1/(2m_2+3) + 1/(2m_2+5) + ... + 1/(2m_4+1))
%C A088403 - (1/(2m_3+2) + 1/(2m_3+4) + ... + 1/(2m_5)
%C A088403 + (1/(2m_4+3) + 1/(2m_4+5) + ... + 1/(2m_6+1))
%C A088403 - ...
%C A088403 so that the partial sums of the terms through the ends of the
rows are respectively 1, just < -2, just > 3, just < -4, just > 5, etc.
%C A088403 Every positive number appears exactly once as a denominator in T.
%C A088403 The series T is a divergent rearrangement of the conditionally
convergent series Sum_{j>=1} (-1)^j/j which has the entire real number
system as its set of limit points.
%D A088403 B. R. Gelbaum and J. M. H. Olmsted, Counterexamples in Analysis,
Holden-Day, San Francisco, 1964; see p. 55.
%e A088403 1 - (1/2 + 1/4 + 1/6 + ... + 1/454) = -2.002183354..., which is
just less than -2; so a(1) = m_1 = 227.
%O A088403 0,2
%K A088403 nonn,nice,more,bref
%Y A088403 Cf. A088507.
%A A088403 njas, Feb 16 2004
Neil, SeqFans,
I just wrote a little Fortran program using 64-bit floats to calculate the
next 2 terms in A088403:
C Neil Sloane's A088403
implicit real*16 (p-z)
integer*8 m1, m2, m3, m4
s = 1.0Q0
m1 = 0
C...+....1....+....2....+....3....+....4....+....5....+....6....+....7..
10 continue
m1 = m1 + 1
t = 1.0Q0/qext(m1+m1)
s = s - t
if ( s .lt. -2.0Q0 ) then
write (*,*) m1, s
m3 = m1
goto 15
endif
goto 10
15 continue
C...+....1....+....2....+....3....+....4....+....5....+....6....+....7..
m2 = 0
20 continue
m2 = m2 + 1
t = 1.0Q0/qext(m2+m2+1)
s = s + t
if ( s .gt. 3.0Q0 ) then
write (*,*) m2, s
m4 = m2
goto 25
endif
goto 20
25 continue
C...+....1....+....2....+....3....+....4....+....5....+....6....+....7..
C Diagnosis to see progress
target = 2.0Q0
write (*,*) ' m3_start=', m3+1
30 continue
m3 = m3 + 1
t = 1.0Q0/qext(m3+m3)
s = s - t
if ( s .lt. target ) then
write (*,*) m3, s
target = target - 1.0Q0
endif
if ( s .lt. -4.0Q0 ) then
write (*,*) m3, s
goto 35
endif
goto 30
35 continue
C...+....1....+....2....+....3....+....4....+....5....+....6....+....7..
write (*,*) ' m4_start=', m4+1
C Diagnosis to see progress
target = -3.0Q0
40 continue
m4 = m4 + 1
t = 1.0Q0/qext(m4+m4+1)
s = s + t
if ( s .gt. target ) then
write (*,*) m4, s
target = target + 1.0Q0
endif
if ( s .gt. 5.0Q0 ) then
write (*,*) m4, s
goto 45
endif
goto 40
45 continue
end
The results are:
227 -2.00218335417278301168821810857870
22945 3.00002111305704094022709303989168
m3_start= 228
1681 1.99987586182063452263072644346311
12422 0.999964322905817529888522274884975
91784 -2.723741766763051976129695675715836E-0006
678197 -1.00000027644030271358908625231417
5011237 -2.00000008949500077609454932713499
37028308 -3.00000000155465546011825940052659
273604248 -4.00000000108863447148035343522068
273604248 -4.00000000108863447148035343522068
m4_start= 22946
169549 -2.99999788153012963960202669811616
1252809 -1.99999966215429848913139095062578
9257077 -0.999999952272842961021019165421510
68401062 6.396099474124162027484980949915391E-0009
505419285 1.00000000072408638564662275136320
3734571452 2.00000000009008129850949083036703
....working, but without a chance to finish in a reasonable time.
From an extrapolation the next term will be ~1.5066336*10^12.
If I made no mistakes, the first terms of A088403 will be
1, 227, 22945, 273604248, ~1.5066336*10^12
Hugo
More information about the SeqFan
mailing list