I'm currently writing an algorithm which depends at one small point on
being able to solve a system of the form:
[a_1 b_1][x_1] = [c_1] (mod m_1)
[a_2 b_2][x_2] = [c_2] (mod m_2)
[a_3 b_3][x_3] = [c_3] (mod m_3)
where a_i,b_i,c_i \in Z and 1 < m_i \in Z
At the moment I am using the pari C library to do this, but it adds loads
of overhead because I need to convert to the pari type and convert back
(and if I am using GMP integers this is really slow) So I was hoping to
write a purpose built algorithm for doing this, but I haven't had any
luck in finding one.
An ideas?
Many thanks,
Brian.
