Undergrad Primality test from pascal triangle

Click For Summary
In Pascal's triangle, a row number is prime if all coefficients (except for the 1s) are divisible by the row number. The discussion highlights the inefficiency of checking binomial coefficients for primality, particularly for large numbers like 15361, as it requires calculating n/2 coefficients. A proposed method involves dividing the coefficients into parts to check for divisibility, but this approach remains computationally expensive. Modern primality tests are significantly faster, relying on checking divisibility up to the square root of n, which is more efficient than the discussed method. Overall, the proposed technique is deemed extremely inefficient for practical primality testing.
rajeshmarndi
Messages
319
Reaction score
0
In pascal triangle, if all the elements of a row(except 1 in both end) are divisible by the row number, then the row number is a prime.

Or if the coefficient of binomial expansion (except 1's) are divisible by the power, then the power is a prime.
It is inefficient to check all the coefficient upto the middle term(since the coefficient repeat, thereafter) with the power of the binomial expansion, for divisibility.

But if we take this way,

Say 15361, i.e

(1+1)15361 = 1 + C(15361,1) + C(15361,2) + ... + C(15361,7680) + C(15361,7681) + ... + C(15361,15360) + 1
So, if 15361 is a prime, then it will also be divisible, to the sum of coefficient, too. But it is highly possible, composite number will also divide the sum of coefficient.
But if we take summation of coefficient in parts, then it becomes almost zero, that each summation part,which will have different values, will also be divisible by power of composite numbers.

That is, in the above example.

If we divide ,7680 coefficient(upto middle term), of the above examples, into say 12 equal parts, i.e each part will have 640 coefficent and each summation part will be of different values. Then, the possibility become highly, that only a power of prime will only divide all the 12 parts summation.C(15361,1)+...+C(15361,640) = will be divisible by the prime power
C(15361,641)+...+C(15361,1280) = will be divisible by the prime power
C(15361,1281)+...+C(15361,1920) = will be divisible by the prime power
C(15361,1921)+...+C(15361,2560) = will be divisible by the prime power
C(15361,2561)+...+C(15361,3200) = will be divisible by the prime power
C(15361,3201)+...+C(15361,3840) = will be divisible by the prime power
C(15361,3841)+...+C(15361,4480) = will be divisible by the prime power
C(15361,4481)+...+C(15361,5120) = will be divisible by the prime power
C(15361,5121)+...+C(15361,5760) = will be divisible by the prime power
C(15361,5761)+...+C(15361,6400) = will be divisible by the prime power
C(15361,6401)+...+C(15361,7040) = will be divisible by the prime power
C(15361,7041)+...+C(15361,7680) = will be divisible by the prime power
So, what is the efficieny of such check, for primality.

Thanks.
 
Mathematics news on Phys.org
rajeshmarndi said:
So, what is the efficieny of such check, for primality.
Extremely bad. To check the primality of n, you have to calculate n/2 binomial coefficients. That works for n=10000, it might work for n=1010, but it won't work for 1040, where n/2 exceeds the total number of computing steps every computer in the world combined ever did.

An approach that is still extremely slow, but orders of magnitude better: just test for divisibility from 2 to ##\sqrt n##. Instead of n/2, the number of calculations scales with ##\sqrt n##.

Modern tests for prime numbers can check 100-digit numbers in a second.
 
mfb said:
Extremely bad. To check the primality of n, you have to calculate n/2 binomial coefficients.
I do not know if there is a formula for summation of coefficent of
C(15361,1)+...+C(15361,640) =

If there is, then you are NOT calculating n/2 binomial coefficent.

You will be only calculating how many parts you are dividing n/2. And then checking those, divisibility with n.
 
There is no useful formula that would speed up calculations notably. It doesn't help to be better by a factor of 10, 1000 or even 1010 if you need a speedup of more than 101000 to be competitive.
 
Here is a little puzzle from the book 100 Geometric Games by Pierre Berloquin. The side of a small square is one meter long and the side of a larger square one and a half meters long. One vertex of the large square is at the center of the small square. The side of the large square cuts two sides of the small square into one- third parts and two-thirds parts. What is the area where the squares overlap?

Similar threads

  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 7 ·
Replies
7
Views
1K
  • · Replies 16 ·
Replies
16
Views
3K
  • · Replies 66 ·
3
Replies
66
Views
7K
Replies
6
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 125 ·
5
Replies
125
Views
19K
  • · Replies 61 ·
3
Replies
61
Views
10K
Replies
4
Views
4K
  • · Replies 2 ·
Replies
2
Views
3K