(a^b) mod c

hi all, i need your help. i want to use (a^b) mod c

but in matlab if i use mod((a^b),c);

it's return inf

because a,b,c are too big.

are you know how to do that with another way, in matlab?? for example use a loop ??

[Oleg Komarov]

- What do you mean by too big?

- Are they scalars?

Answer 1

[Jan]

If b is not prime, but x * y:

(a^b) mod c = (a^x mod c)^y mod c

But if b is prime, this does not help to avoid the overflow. Then I'd try FEX:John's VPI or FEX:Ben Barrowes' Multiple Precision Toolbox.

EDITED: A loop method:

function r = ApowerBmodC(a, b, c)
r = 1;
for i = 1:b
  r = mod(a * r, c);
end

Then "ApowerBmodC(1234, 5678, 1256)" replies 1112. For these numbers the "(a^x mod c)^y mod c" does not help, because 5678 = 2*17*167 and "mod(1234^2, 1256)^17" is 4.38e45 such that trailing figures are all zero due to rounding.

[zikrullah muhammad]

thank you, but if b prime what must i do??

[Jan]

Have you looked at the two posted links?

[Sean de Wolski]

VPI is back !!

Answer 2

[Oleg Komarov]

If you're looking for elementwise elevation to a power then:

mod(a.^b,c)

[zikrullah muhammad]

i need loop, so the number not bigger than realmax

for example:

mod((1234^5678),1256)---> it's return NaN