| Randomu() behavior - BAD! [message #68604] |
Fri, 13 November 2009 11:01 |
Conor
Messages: 138
Registered: February 2007
|
Senior Member |
|
|
So I did a simple test and have a better idea of how the seed
initialization works with randomu(). In my mind, this is not good
news! It seems that IDL does have some sort of hidden global "seed"
variable. The first time you pass an undeclared variable to randomu
and let it initialize the seed, it is storing the seed it generated in
this hidden variable, and using that same hidden variable the next
time randomu() is called with an undeclared variable. Why is this a
problem? Because if you are calling randomu() with both a declared
seed variable and undeclared seed variables, it will end up returning
the same set of random numbers. This will happen of course if you
have one routine that is generating random variables, and you are
calling subroutines that also generate random variables. If you don't
pass your seed value to the subroutine, then it will be working with
an undeclared variable and will end up re-generating the exact same
random numbers you've already used!!!! Here's an example:
print,randomu(seed,1)
print,randomu(first_var_undeclared,1)
print,randomu(seed,1)
print,randomu(second_var_undeclared,1)
print,randomu(seed,1)
print,randomu(third_var_undeclared,1)
print,randomu(seed,1)
Outputs:
0.456874
0.528637
0.528637
0.783314
0.783314
0.0769241
0.0769241
Notice that there is no *visible* relation between my calls to randomu
() with the seed variable and my calls to randomu with a series of
undeclared variables. However since my seed variable was seeded by
the same hidden variable used in all calls to randomu(), I end up
getting duplicate results from subsequent calls to randomu() - not
good! Moreover this hidden variable is global and is used in calls in
subroutines as well. Try out these routines:
pro test2,seed=seed
print,randomu(seed,1)
end
pro test
test2,seed=seed
print,randomu(seed,1)
test2
print,randomu(seed,1)
test2
print,randomu(seed,1)
test2
end
When I run 'test' I get the following output:
0.0108795
0.292802
0.292802
0.121008
0.121008
Just to really prove the point, I can then duplicate these numbers by
calling randomu() from the main command line using the same seed
variable left over from my previous example:
print,randomu(seed,1)
print,randomu(seed,1)
print,randomu(seed,1)
0.0108795
0.292802
0.121008
I really wish I had understood this behavior before (it is *NOT*
documented). I have had routines which keep track of their own seed
and call subroutines which use an undeclared seed. That means that my
main routine and my subroutine are re-using the same random numbers,
which is decidedly bad. Ultimately this is my fault - I should have
been passing the seed back and forth between subroutines. However, it
would have been nice to have this documented and to have a nice
warning about this behavior, because naively I don't expect different
programs and functions to be able to communicate expect as I enable
them to communicate. Basically RSI has created a hidden global
variable and not told anybody about it - that seems like a big lapse
on their parts.
For the lazy programmer you could turn this around and make it a good
thing. I'm sure it would be horrible programming practice, but if you
just always passed an undeclared variable to randomu() as a seed, it
would automatically use this hidden global variable and you wouldn't
have to worry about passing your seed back and forth :) That would
save you some trouble until the next version update when they change
the behavior again without letting anyone know :(
Also, this still doesn't explain the problem I was having in the other
thread.
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
