
Assignment 12
The unresolved problem
Consider 4 real numbers A, B, C, and D arranged
in a row and their successive differences in absolute terms as shown in the
array below:
![]()
![]()
If the procedure is continued for the fourth row,
fifth row and so on, we ultimately end with a row having 4 zeros. The aim is to
have a maximum number of non-zero zeros.
For example, if A = 1, B =
4, C = 5, and D = 9, we have the following structure.
Example 1:
|
Difference |
A |
B |
C |
D |
|
|
1 |
4 |
5 |
9 |
|
1st |
3 |
1 |
4 |
8 |
|
2nd
|
2 |
3 |
4 |
5 |
|
3rd
|
1 |
1 |
1 |
3 |
|
4th
|
0 |
0 |
2 |
2 |
|
5th
|
0 |
2 |
0 |
2 |
|
6th
|
2 |
2 |
2 |
2 |
|
|
0 |
0 |
0 |
0 |
We observe that after the 6th difference all
the terms are zero. The problem is to find a set of four numbers such that we
have a maximum of non-zero rows.
Example 2:
|
Difference |
A |
B |
C |
D |
|
1st |
900 |
4 |
500 |
260 |
|
2nd
|
896 |
496 |
240 |
640 |
|
3rd
|
400 |
256 |
400 |
256 |
|
4th
|
144 |
144 |
144 |
144 |
|
|
0 |
0 |
0 |
0 |
Table 12.2
Observe that the maximum difference in each row
decreases as we move down the rows. Further, the row just above the final zero
row has all even terms or are divisible by 2.
Example 3:
In this example, we select four real numbers.
|
Difference |
A |
B |
C |
D |
|
1st |
0.9 |
0.4 |
0.7 |
1.5 |
|
2nd
|
0.5 |
0.3 |
0.8 |
0.6 |
|
3rd
|
0.2 |
0.5 |
0.2 |
0.1 |
|
4th
|
0.3 |
0.3 |
0.1 |
0.1 |
|
5th
|
0 |
0.2 |
0 |
0.2 |
|
6th
|
0.2 |
0.2 |
0.2 |
0.2 |
|
|
0 |
0 |
0 |
0 |
Table 12.3
The problem has a ÔcyclicÕ nature since the difference
between the first and fourth numbers in each row are being considered. The four
numbers in example 1 have been represented on a circle to give another
representation of the problem.



Figure 12.1
A number of possibilities for A, B,
C, and D have
been tried out in attempt to maximize the number of non-zero rows. If we choose
A, B, C, and D such that
,
, and
, then this increases the number of non-zero rows
substantially for certain values of A.
For instance, if we choose A =
1.83928675521416, we have 61 non-zero rows as shown in Table 12.4.
Table 12.4
If
we choose A = 0.543689012692077,
we equally have 61 rows as shown in Table 12.5.
|
|
A |
B |
C |
D |
|
1 |
0.543689013 |
0.295597743 |
0.160713245 |
0.087378025 |
|
2 |
0.24809127 |
0.134884498 |
0.073335219 |
0.456310987 |
|
3 |
0.113206772 |
0.061549278 |
0.382975768 |
0.208219717 |
|
4 |
0.051657494 |
0.32142649 |
0.174756051 |
0.095012945 |
|
5 |
0.269768995 |
0.146670439 |
0.079743106 |
0.043355451 |
|
6 |
0.123098557 |
0.066927333 |
0.036387655 |
0.226413545 |
|
7 |
0.056171224 |
0.030539677 |
0.190025889 |
0.103314988 |
|
8 |
0.025631547 |
0.159486212 |
0.086710901 |
0.047143764 |
|
9 |
0.133854665 |
0.072775311 |
0.039567137 |
0.021512218 |
|
10 |
0.061079355 |
0.033208174 |
0.018054919 |
0.112342448 |
|
11 |
0.027871181 |
0.015153255 |
0.094287529 |
0.051263093 |
|
12 |
0.012717926 |
0.079134274 |
0.043024435 |
0.023391913 |
|
13 |
0.066416348 |
0.036109839 |
0.019632523 |
0.010673987 |
|
14 |
0.030306509 |
0.016477316 |
0.008958536 |
0.055742361 |
|
15 |
0.013829193 |
0.00751878 |
0.046783825 |
0.025435852 |
|
16 |
0.006310413 |
0.039265045 |
0.021347974 |
0.011606659 |
|
17 |
0.032954632 |
0.017917071 |
0.009741315 |
0.005296246 |
|
18 |
0.015037561 |
0.008175757 |
0.004445069 |
0.027658386 |
|
19 |
0.006861804 |
0.003730688 |
0.023213317 |
0.012620826 |
|
20 |
0.003131117 |
0.01948263 |
0.010592492 |
0.005759021 |
|
21 |
0.016351513 |
0.008890138 |
0.00483347 |
0.002627905 |
|
22 |
0.007461375 |
0.004056668 |
0.002205566 |
0.013723608 |
|
23 |
0.003404707 |
0.001851102 |
0.011518043 |
0.006262233 |
|
24 |
0.001553605 |
0.009666941 |
0.005255809 |
0.002857526 |
|
25 |
0.008113335 |
0.004411131 |
0.002398284 |
0.00130392 |
|
26 |
0.003702204 |
0.002012848 |
0.001094363 |
0.006809415 |
|
27 |
0.001689356 |
0.000918485 |
0.005715052 |
0.003107211 |
|
28 |
0.000770872 |
0.004796567 |
0.002607841 |
0.001417854 |
|
29 |
0.004025695 |
0.002188726 |
0.001189986 |
0.000646983 |
|
30 |
0.001836969 |
0.00099874 |
0.000543004 |
0.003378713 |
|
31 |
0.000838229 |
0.000455736 |
0.002835709 |
0.001541744 |
|
32 |
0.000382493 |
0.002379973 |
0.001293965 |
0.000703515 |
|
33 |
0.00199748 |
0.001086008 |
0.00059045 |
0.000321021 |
|
34 |
0.000911472 |
0.000495557 |
0.000269429 |
0.001676458 |
|
35 |
0.000415914 |
0.000226128 |
0.001407029 |
0.000764986 |
|
36 |
0.000189786 |
0.001180901 |
0.000642043 |
0.000349072 |
|
37 |
0.000991115 |
0.000538858 |
0.000292971 |
0.000159286 |
|
38 |
0.000452257 |
0.000245887 |
0.000133685 |
0.000831829 |
|
39 |
0.000206369 |
0.000112202 |
0.000698144 |
0.000379572 |
|
40 |
9.4167E-05 |
0.000585942 |
0.000318571 |
0.000173203 |
|
41 |
0.000491775 |
0.00026737 |
0.000145368 |
7.90361E-05 |
|
42 |
0.000224404 |
0.000122002 |
6.63322E-05 |
0.000412739 |
|
43 |
0.000102403 |
5.56697E-05 |
0.000346406 |
0.000188334 |
|
44 |
4.67329E-05 |
0.000290737 |
0.000158072 |
8.59315E-05 |
|
45 |
0.000244004 |
0.000132664 |
7.21408E-05 |
3.91986E-05 |
|
46 |
0.000111339 |
6.05236E-05 |
3.29421E-05 |
0.000204805 |
|
47 |
5.08159E-05 |
2.75814E-05 |
0.000171863 |
9.34657E-05 |
|
48 |
2.32345E-05 |
0.000144282 |
7.83973E-05 |
4.26498E-05 |
|
49 |
0.000121047 |
6.58843E-05 |
3.57474E-05 |
1.94154E-05 |
|
50 |
5.51628E-05 |
3.01369E-05 |
1.63321E-05 |
0.000101632 |
|
51 |
2.5026E-05 |
1.38048E-05 |
8.52997E-05 |
4.64689E-05 |
|
52 |
1.12212E-05 |
7.14949E-05 |
3.88308E-05 |
2.14429E-05 |
|
53 |
6.02737E-05 |
3.26641E-05 |
1.73878E-05 |
1.02218E-05 |
|
54 |
2.76096E-05 |
1.52763E-05 |
7.16604E-06 |
5.00519E-05 |
|
55 |
1.23332E-05 |
8.11028E-06 |
4.28859E-05 |
2.24424E-05 |
|
56 |
4.22296E-06 |
3.47756E-05 |
2.04435E-05 |
1.01091E-05 |
|
57 |
3.05527E-05 |
1.43321E-05 |
1.03344E-05 |
5.88618E-06 |
|
58 |
1.62206E-05 |
3.9977E-06 |
4.44821E-06 |
2.46665E-05 |
|
59 |
1.22229E-05 |
4.50514E-07 |
2.02183E-05 |
8.44591E-06 |
|
60 |
1.17724E-05 |
1.97678E-05 |
1.17724E-05 |
3.77695E-06 |
|
61 |
7.9954E-06 |
7.9954E-06 |
7.9954E-06 |
7.9954E-06 |
|
|
0 |
0 |
0 |
0 |
Table 12.5
Interestingly,
the two numbers 0.543689012692077 and
1.83928675521416 are related,
![]()
i.e.,
they are reciprocals. The large number of decimal places for the above two
values of A shows that accuracy is
also a determining factor in the number of non-zero rows.
Table 12.6 shows the number of non-zero rows as the
accuracy of A increases from 1.839 (3 d.p) to 1.83928675521416 (14
d.p).
|
Value of A |
No. of non-zero rows |
|
1.839 |
19 |
|
1.8392 |
21 |
|
1.83928 |
26 |
|
1.839286 |
30 |
|
1.8392867 |
33 |
|
1.83928675 |
37 |
|
1.839286755 |
42 |
|
1.8392867552 |
46 |
|
1.83928675521 |
49 |
|
1.839286755214 |
54 |
|
1.8392867552141 |
54 |
|
1.83928675521416 |
61 |
The values in Table 12.4 and 12.5 are plotted in
Figure 12.1 and Figure 12.2, respectively to show how the numbers decrease
along the four columns.

Figure 12.1

We observe that the values decrease monotically in
Figure 12.1 in contrast to Figure 12.2. The values in Table 12.4 have also been
plotted along the rows. Once again, the values display some pattern as shown in
Figure 12.3.

Figure 12.3
The
topmost curve is for row 1. The second curve is for row two and so on. The
decreasing pattern in Figures 12.1 and 12.2 may suggest a clue to the problem
of finding the maximum number of non-zero rows.
Ajay Ramful
3 December 2006