Prove that the integers ## c, c+a, c+2a, c+3a, ...., c+(n-1)a ## ....

  • Thread starter Math100
  • Start date
  • Tags
    Integers
In summary: Thanks.In summary, to prove that the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##, we suppose for the sake of contradiction that ## r, s\in \{0, ..., n-1\} ## for ## r<s ## where ## c+ra\equiv c+sa\pmod {n} ##. By using the fact that ## gcd(a, n)=1 ##, we can show that this leads to the contradiction that ## n<s-r ##, which is not possible since ## s-r<n ##. Therefore, any two of the integers are incon
  • #1
Math100
794
221
Homework Statement
Prove the following statement:
If ## gcd(a, n)=1 ##, then the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##.
Relevant Equations
None.
Proof:

Suppose for the sake of contradiction that ## r, s\in {0, ..., n-1} ## for ## r<s ## where ## c+ra\equiv c+sa\pmod {n} ##.
Then ## c+ra\equiv c+sa\pmod {n}\implies ra\equiv sa\pmod {n}\implies r\equiv s\pmod {n} ##.
Thus ## n\mid (r-s)\implies n<r-s ##, which is a contradiction because ## r\leq n-1 ## and ## s\leq n-1 ## together implies ## s-r<n ##.
Therefore, if ## gcd(a, n)=1 ##, then the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##.
 
Physics news on Phys.org
  • #2
I find this proof mostly correct. Two minor mistakes or to be more accurate, incomplete statements:
  • You should 've stated that because ##gcd(a,n)=1## then it follows that it holds $$ra\equiv sa \pmod n\Rightarrow r\equiv s \pmod n$$
  • We end up with ##n<r-s## which it should be actually ##n<s-r## (maybe that's a typo ) because you initially took ##r<s## hence ##r-s## is negative. And the fact that ##s-r<n## comes from ##s\leq n-1## and ##r\geq 0##.
 
Last edited:
  • #3
Math100 said:
Homework Statement:: Prove the following statement:
If ## gcd(a, n)=1 ##, then the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##.
Relevant Equations:: None.

Proof:

Suppose for the sake of contradiction that ## r, s\in {0, ..., n-1} ## for ## r<s ## where ## c+ra\equiv c+sa\pmod {n} ##.
Then ## c+ra\equiv c+sa\pmod {n}\implies ra\equiv sa\pmod {n}\implies r\equiv s\pmod {n} ##.
This is already the contradiction since ##s-r\in \{0,\ldots,n-1\}## is a complete set of remainders.
Math100 said:
Thus ## n\mid (r-s)\implies n<r-s ##, which is a contradiction because ## r\leq n-1 ## and ## s\leq n-1 ## together implies ## s-r<n ##.
You should have used ##s-r## since ##r-s <0## which makes things complicated, e.g. ##n<r-s ## is wrong.<
Math100 said:
Therefore, if ## gcd(a, n)=1 ##, then the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##.
Where did you use ## gcd(a, n)=1 ##?

You have shown that no two remainders ##c+ra,c+sa## can be the same modulo ##n## without being equal as integers. How do we know that all remainders actually occur?
 
  • #4
Suppose for the sake of contradiction that ## r, s\in \{0, ..., n-1\} ## for ## r<s ## where ## c+ra\equiv c+sa\pmod {n} ##.
Since ## gcd(a, n)=1 ##, it follows that ## c+ra\equiv c+sa\pmod {n}\implies ra\equiv sa\pmod {n}\implies r\equiv s\pmod {n} ##.
Then ## n\mid (r-s)\implies n<s-r ##, which is a contradiction because ## r\leq n-1 ## and ## s\leq n-1 ## together implies ## s-r<n ##.
Thus, any two of the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## are incongruent modulo ## n ##.
Therefore, if ## gcd(a, n)=1 ##, then the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##.
 
  • #5
You have basically repeated your former version.
Math100 said:
Suppose for the sake of contradiction that ## r, s\in \{0, ..., n-1\} ## for ## r<s ## where ## c+ra\equiv c+sa\pmod {n} ##.
Since ## gcd(a, n)=1 ##, it follows that ## c+ra\equiv c+sa\pmod {n}\implies ra\equiv sa\pmod {n}\implies r\equiv s\pmod {n} ##.
## gcd(a, n)=1 ## is used in the last step because it guarantees that there is an ##\alpha ## such that ##\alpha \cdot a\equiv 1 \pmod n## that is, that ##a## has a multiplicative inverse element. Now if
\begin{align*}
ra\equiv sa\pmod {n} \Longrightarrow r\equiv r\cdot 1\equiv r\cdot a\alpha \equiv s\cdot a\alpha \equiv s\cdot 1\equiv s\pmod {n}
\end{align*}
That's why the condition is needed. It guarantees the existence of ##\alpha =a^{-1}.##
(see https://www.physicsforums.com/threads/prove-that-aa_-1-aa_-2-aa_-n-is-also-a-complete-set.1015722/)
Math100 said:
Then ## n\mid (r-s)\implies n<s-r ##, which is a contradiction because ## r\leq n-1 ## and ## s\leq n-1 ## together implies ## s-r<n ##.
Thus, any two of the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## are incongruent modulo ## n ##.
Therefore, if ## gcd(a, n)=1 ##, then the integers ## c, c+a, c+2a, c+3a, ..., c+(n-1)a ## form a complete set of residues modulo ## n ## for any ## c ##.
 
  • #6
fresh_42 said:
You have basically repeated your former version.

## gcd(a, n)=1 ## is used in the last step because it guarantees that there is an ##\alpha ## such that ##\alpha \cdot a\equiv 1 \pmod n## that is, that ##a## has a multiplicative inverse element. Now if
\begin{align*}
ra\equiv sa\pmod {n} \Longrightarrow r\equiv r\cdot 1\equiv r\cdot a\alpha \equiv s\cdot a\alpha \equiv s\cdot 1\equiv s\pmod {n}
\end{align*}
That's why the condition is needed. It guarantees the existence of ##\alpha =a^{-1}.##
(see https://www.physicsforums.com/threads/prove-that-aa_-1-aa_-2-aa_-n-is-also-a-complete-set.1015722/)
If ## gcd(a, n)=1 ## should be used in the last step, then where should I insert/include that "There exists an ## \alpha ## such that ## \alpha\cdot a\equiv 1\pmod {n}\Longleftrightarrow gcd(a, n)=1 ##"?
 
  • #7
Math100 said:
If ## gcd(a, n)=1 ## should be used in the last step, then where should I insert/include that "There exists an ## \alpha ## such that ## \alpha\cdot a\equiv 1\pmod {n}\Longleftrightarrow gcd(a, n)=1 ##"?
It was ok what you wrote. I only wanted to explain where it is actually used because it isn't apparent by itself.

## gcd(a, n)=1 ## makes ##a## a unit (multiplicative invertible element) modulo ##n##, i.e. in ##\mathbb{Z}/n\mathbb{Z}=\mathbb{Z}_n=\{0,1,2, \ldots,n-1 \}## and this allows us to cancel it from the equation ##r\cdot a \equiv s\cdot a.##

A counterexample if ## gcd(a, n)\neq 1 ## is the following:
Set ##n=12\, , \, a=8\, , \,r=6\, , \,s=3.## Here we have ##6\cdot 8 \equiv 0 \equiv 3\cdot 8 \pmod {12}## but ##r\neq s.##
 
  • Like
Likes Math100

FAQ: Prove that the integers ## c, c+a, c+2a, c+3a, ...., c+(n-1)a ## ....

1. What is the formula for finding the sum of an arithmetic sequence?

The formula for finding the sum of an arithmetic sequence is:
S = (n/2) * (2c + (n-1)a), where S is the sum, c is the first term, a is the common difference, and n is the number of terms.

2. How do you prove that the integers in an arithmetic sequence are evenly spaced?

To prove that the integers in an arithmetic sequence are evenly spaced, you can use the formula for finding the common difference (a) between consecutive terms:
a = (cn - cn-1), where cn and cn-1 are any two consecutive terms in the sequence. If the value of a is the same for all pairs of consecutive terms, then the sequence is evenly spaced.

3. Can you use the formula for an arithmetic sequence to find a specific term in the sequence?

Yes, you can use the formula cn = c + (n-1)a to find any term in an arithmetic sequence. Simply plug in the values for c, a, and n to find the desired term.

4. How do you prove that two arithmetic sequences are equivalent?

To prove that two arithmetic sequences are equivalent, you can show that they have the same first term (c) and the same common difference (a). If these values are the same for both sequences, then all of the terms in the sequences will be the same.

5. Can you use the formula for an arithmetic sequence to find the number of terms in the sequence?

Yes, you can use the formula n = (cn - c)/a + 1 to find the number of terms in an arithmetic sequence. Simply plug in the values for c, a, and cn (the last term in the sequence) and solve for n.

Back
Top