What Is the Smallest Good Number for a Subset of Squares?

In summary, we observe that $n$ is a good number if there are two elements in any subset of $M$ of size $n$ whose sum is a prime number. The smallest good number is $11$, as shown by the table of all possible sums of an odd and an even number in the set of squares of the first $20$ natural numbers. It is proven that there is no bad subset of $11$ elements, showing that $n=11$ is indeed the smallest good number.
  • #1
lfdahl
Gold Member
MHB
749
0
$M$ is the set of squares of the fi rst $20$ natural numbers:\[M = \left\{1^2,2^2,3^2,...,19^2,20^2\right\}\]We say that $n$ is a good number, if in any subset of $M$ of size $n$ there are two
elements $a$ and $b$ such that $a + b$ is a prime number. Find the smallest good number.
 
Mathematics news on Phys.org
  • #2
lfdahl said:
$M$ is the set of squares of the first $20$ natural numbers:\[M = \left\{1^2,2^2,3^2,...,19^2,20^2\right\}\]We say that $n$ is a good number, if in any subset of $M$ of size $n$ there are two
elements $a$ and $b$ such that $a + b$ is a prime number. Find the smallest good number.
I'm not quite sure about this :confused:
[sp]

We observe first that, if we have a subset $S\subset M$ of $k$ elements such that all the sums of pairs of elements of $S$ are composite (we will call that a bad subset), then $n > k$, where $n$ is the smallest good number. Taking $S$ as the set of $10$ even squares, we see that $n\ge11$.

To prove that $n=11$, we must show that there is no bad subset of $11$ elements. As the sum of any two distinct elements of the same parity is composite, we need only consider the sums of an odd and an even number.

The following table contains all the possible such sums of squares.
$$
\begin{array}{r|rrrrrrrrrr|r}
& 1 & 3 & 5 & 7 & 9 & 11 & 13 & 15 & 17 & 19 & t_r\\
\hline
2 & 5 & 13 & 29 & 53 & \bf 85 & \bf 125 & 173 & 229 & 293 & \bf 365 & 3 \\
4 & 17 & \bf 25 & 41 & \bf 65 & 97 & 137 & \bf 185 & 241 & \bf 305 & \bf 377 & 5 \\
6 & 37 & \bf 45 & 61 & \bf 85 & \bf 117 & 157 & \bf 205 & \bf 261 & \bf 325 & 397 & 6 \\
8 & \bf 65 & 73 & 89 & 113 & \bf 145 & \bf 185 & 233 & \bf 289 & 353 & \bf 425 & 5 \\
10 & 101 & 109 & \bf 125 & 149 & 181 & \bf 221 & 269 & \bf 325 & 389 & 461 & 3 \\
12 & \bf 145 & \bf 153 & \bf 169 & 193 & \bf 225 & \bf 265 & 313 & \bf 369 & 433 & \bf 505 & 7 \\
14 & 197 & \bf 205 & \bf 221 & \bf 245 & 277 & 317 & \bf 365 & 421 & \bf 485 & 557 & 5 \\
16 & 257 & \bf 265 & 281 & \bf 305 & 337 & \bf 377 & \bf 425 & \bf 481 & \bf 545 & 617 & 6 \\
20 & 401 & 409 & \bf 425 & 449 & \bf 481 & 521 & 569 & \bf 625 & \bf 689 & 761 & 4 \\
\hline
t_c & 2 & 5 & 4 & 4 & 5 & 5 & 4 & 6 & 5 & 4
\end{array}
$$
The composite numbers are in bold; the last row ($t_c$) and the last column ($t_r$) contain the numbers of composite numbers in the corresponding columns and rows, respectively.

Assume that there exists a bad subset of $11$ elements; let $n_c$ and $n_r$ be the numbers of rows (even numbers) and columns (odd numbers) in that subset; $n_c + n_r = 11$. We must select $n_r$ rows and $n_c$ columns in such a way that all the remaining entries are composite.

As the row with the largest $t_r$ contains $7$ composite numbers, we must have $n_c \le 7$; this implies that we must have at least $4$ rows.

As the fourth larges value of $t_r$ is $5$, we must have $n_c\le5$, and therefore $n_r\ge6$. As the sixth largest $t_r$ is $5$ again, this gives nothing new.

As the largest column total $t_c$ is $6$, we must have $n_r\le6$; this means that we must have $n_r=6$ and $n_c=5$. Now, as the fifth largest $t_c$ is $5$, we must have $n_r\le5$, and this is a contradiction.

We conclude that there can be no bad subset of $11$ elements, and $n=11$.

[/sp]
 
Last edited:
  • #3
castor28 said:
I'm not quite sure about this :confused:
[sp]

We observe first that, if we have a subset $S\subset M$ of $k$ elements such that all the sums of pairs of elements of $S$ are composite (we will call that a bad subset), then $n > k$, where $n$ is the smallest good number. Taking $S$ as the set of $10$ even squares, we see that $n\ge11$.

To prove that $n=11$, we must show that there is no bad subset of $11$ elements. As the sum of any two distinct elements of the same parity is composite, we need only consider the sums of an odd and an even number.

The following table contains all the possible such sums of squares.
$$
\begin{array}{r|rrrrrrrrrr|r}
& 1 & 3 & 5 & 7 & 9 & 11 & 13 & 15 & 17 & 19 & t_r\\
\hline
2 & 5 & 13 & 29 & 53 & \bf 85 & \bf 125 & 173 & 229 & 293 & \bf 365 & 3 \\
4 & 17 & \bf 25 & 41 & \bf 65 & 97 & 137 & \bf 185 & 241 & \bf 305 & \bf 377 & 5 \\
6 & 37 & \bf 45 & 61 & \bf 85 & \bf 117 & 157 & \bf 205 & \bf 261 & \bf 325 & 397 & 6 \\
8 & \bf 65 & 73 & 89 & 113 & \bf 145 & \bf 185 & 233 & \bf 289 & 353 & \bf 425 & 5 \\
10 & 101 & 109 & \bf 125 & 149 & 181 & \bf 221 & 269 & \bf 325 & 389 & 461 & 3 \\
12 & \bf 145 & \bf 153 & \bf 169 & 193 & \bf 225 & \bf 265 & 313 & \bf 369 & 433 & \bf 505 & 7 \\
14 & 197 & \bf 205 & \bf 221 & \bf 245 & 277 & 317 & \bf 365 & 421 & \bf 485 & 557 & 5 \\
16 & 257 & \bf 265 & 281 & \bf 305 & 337 & \bf 377 & \bf 425 & \bf 481 & \bf 545 & 617 & 6 \\
20 & 401 & 409 & \bf 425 & 449 & \bf 481 & 521 & 569 & \bf 625 & \bf 689 & 761 & 4 \\
\hline
t_c & 2 & 5 & 4 & 4 & 5 & 5 & 4 & 6 & 5 & 4
\end{array}
$$
The composite numbers are in bold; the last row ($t_c$) and the last column ($t_r$) contain the numbers of composite numbers in the corresponding columns and rows, respectively.

Assume that there exists a bad subset of $11$ elements; let $n_c$ and $n_r$ be the numbers of rows (even numbers) and columns (odd numbers) in that subset; $n_c + n_r = 11$. We must select $n_r$ rows and $n_c$ columns in such a way that all the remaining entries are composite.

As the row with the largest $t_r$ contains $7$ composite numbers, we must have $n_c \le 7$; this implies that we must have at least $4$ rows.

As the fourth larges value of $t_r$ is $5$, we must have $n_c\le5$, and therefore $n_r\ge6$. As the sixth largest $t_r$ is $5$ again, this gives nothing new.

As the largest column total $t_c$ is $6$, we must have $n_r\le6$; this means that we must have $n_r=6$ and $n_c=5$. Now, as the fifth largest $t_c$ is $5$, we must have $n_r\le5$, and this is a contradiction.

We conclude that there can be no bad subset of $11$ elements, and $n=11$.

[/sp]

Thankyou, castor28, for your careful analysis and correct solution. Your approach works in my view. I´m glad, that you participated in this challenge. (Yes)

The suggested solution demonstrates a shortcut:

The answer is $n = 11$.
Let $K = \left \{ 1^2,3^2,5^2,...,19^2 \right \}$. Since the sum of any two elements of is not prime, $n \geq 11$.
Now let us show, that $n \leq 11$. We partition $M$ into $10$ subsets of order $2$ such that the sum of two elements in any subset is prime:

\[M = \left \{ 1^2,4^2 \right \}\cup \left \{2^2,3^2 \right \}\cup \left \{5^2,8^2 \right \}\cup \left \{ 6^2,11^2\right \}\cup \left \{7^2,10^2 \right \}\cup \left \{9^2,16^2 \right \}\cup \left \{12^2,13^2 \right \}\cup \left \{14^2,15^2 \right \}\cup \left \{17^2,18^2 \right \}\cup \left \{19^2,20^2 \right \}.\]

Any subset of $M$ having $11$ elements contains both elements of at least one of these $10$ subsets. Done.
 

FAQ: What Is the Smallest Good Number for a Subset of Squares?

What does the statement "N is a good number" mean?

The term "N is a good number" is used to refer to a positive integer that satisfies the following condition: in any subset of M of size n, there exist two elements a and b such that their sum, a+b, is a prime number.

How is the concept of "good number" related to prime numbers?

The concept of "good number" is closely related to prime numbers because it involves the sum of two numbers being prime. In other words, a good number is an integer that can be partitioned into two smaller integers whose sum is a prime number.

Are there any specific values for N that are considered good numbers?

Yes, there are certain values for N that are commonly known as "good numbers". These include 2, 4, 8, 16, 28, and 64, among others. These numbers have the property that any subset of their size will contain two elements whose sum is a prime number.

How is the concept of "good number" used in mathematics and science?

The concept of "good number" has applications in various areas of mathematics and science, such as number theory, graph theory, and cryptography. It is also used in the study of prime numbers and their properties.

Can the statement "N is a good number" be proven for all values of N?

No, the statement "N is a good number" cannot be proven for all values of N. This is because there are infinitely many positive integers, and it is impossible to test them all. However, there is strong evidence to suggest that the statement holds true for a large number of values of N.

Similar threads

Replies
5
Views
1K
Replies
22
Views
3K
Replies
1
Views
737
Replies
13
Views
4K
Replies
1
Views
1K
Replies
4
Views
2K
Back
Top