Limit to seed number

[#1]

Hello everyone
I was wondering if there were an upper limit to the seed that we can define in JMP ?  Can we go up to the infinite or nearby ?   

I am concatenating  big numbers to be able to recreate the same seed each time they occur together, but as those number are increasing with time, the seed will quickly increase as well. So far everything is running fine but I may reach values close to 10^12 or 10^13  soon enough.

thanks for your answer ! 

regards 

[#2]

It is a good question, and I don't have a precise answer. But I'm pretty sure you want to stick with integers from 1 to 2147483647 (2^31-1). After experimenting a bit:

s1 = 123559868338560;
s2 = 123456789123456;

randomreset(s1);
print("s1:",randomuniform(),randomuniform(),randomuniform());

randomreset(s2);
print("s2:",randomuniform(),randomuniform(),randomuniform());

print("bits:",hex(s1), hex(s2)); // the two numbers are almost the same, in bits

"s1:"
0.810199321946129
0.909442130941897
0.272601864300668
"s2:"
0.810199321946129
0.909442130941897
0.272601864300668
"bits:"
"42DC182183E46000"
"42DC122183E46000"

[#3 · Accepted solution]

You can use the full 32-bit range (except 0), so you should get different random results for seeds from [1, 4294967295].  Beyond that it will wrap, so note that 4294967296 wraps to 0 and uses a random seed.  Below is @Craige_Hales script with seeds 2^32+1 and 1 to show that the random values are the same with these two seeds.

s1 = 4294967297;
s2 = 1;

randomreset(s1);
print("s1:",randomuniform(),randomuniform(),randomuniform());

randomreset(s2);
print("s2:",randomuniform(),randomuniform(),randomuniform());

print("bits:",hex(s1), hex(s2)); // the two numbers are almost the same, in bits

"s1:"
0.883865864481777
0.973821102175861
0.507582586025819
"s2:"
0.883865864481777
0.973821102175861
0.507582586025819
"bits:"
"41F0000000100000"
"3FF0000000000000"

[#4]

Thanx, this is very interesting.