Is there something like this for use with an access database? Thanks
Colin ----- Original Message ----- From: "Jack Lavender" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, February 26, 2003 12:47 PM Subject: RE: Randomizing Query Output > > Please be aware that unless you keep and reseed with the key pulling a > uniform number out of a psuedo-random number generator will not give the the > nice distribution you are seeking. Not all RNGs (Random Number Generators) > are equal. I have not tested it since CF 3.0, it was poor then. > > > I suggest that you keep a tally on the side when you pull/use records. This > will allow you to run tests on the usage to demonstrate the distribution. > > Of course all of this rigor depends on if it really matters. > > > for programing references: > > The Art of Computer Programming : Seminumerical Algorithms by Donald Knuth. > Algorithms in C++ -- Sedgewick > > A dated but seminal introduction to the problem: > > Simulation Modeling and Analysis > by Averill M. Law, David W. Kelton, W. David Kelton, David M. Kelton > > good luck, > jack lavender > > > > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > Behalf Of PJ Gaenir > Sent: Wednesday, February 26, 2003 11:20 AM > To: [EMAIL PROTECTED] > Subject: Randomizing Query Output > > > I hope someone can help with this. I am in a sort of desperate timing > situation here. > > I have a testing facility that needs to pull all records of a certain spec > out of a > table, and randomize how they are retrieved. Then I loop the first X# of > them > into a list, and the test is based on this list. Since they are supposed to > be > randomly ordered, it shouldn't matter that I'm choosing the first quantity > of > them. > > The rand functions I normally use for 'semi-random' output are just not > working > I guess. The instructor has 140 questions in the pool, and he is pulling 20 > questions for the test. The questions keep coming up from the first > chapter, > which means, the first records. I will put below the two codes I've tried > for it, > neither of which has worked. Any better suggestions would be dearly > appreciated. > > ----------code2----------- - 2nd try as the first didn't work well. It does > order them > differently mind you, but is not nearly random enough to 'mix up' 100 items > so > the first 20 chosen are as likely to be from middle or end as beginning. > > <cfquery name="get"> > select id > from table > where [yadayada] > order by rand(#randrange(11,289)#) > </cfquery> > > ----------code1----------- (I first had this) > > <cfset randnum1 = "#randrange(3,289)#"> > <cfset randnum2 = "(#variables.randnum1# * #randrange(7,3963)#)"> > <cfquery name="get"> > select id > from table > where [yadayada] > order by rand (#randnum2# * ((piid / 3) * 1.73)) > </cfquery> > > The CF docs I have don't really address this. I have another randomizing > code that basically uses a place in a text field by letter, it is not > anywhere > near really random either, but that won't work in this case. > > Best regards, > PJ > > > ----------------------------------------------- > To post, send email to [EMAIL PROTECTED] > To unsubscribe: > Send UNSUBSCRIBE to [EMAIL PROTECTED] > To subscribe / unsubscribe: http://www.dfwcfug.org > > > > ----------------------------------------------- > To post, send email to [EMAIL PROTECTED] > To unsubscribe: > Send UNSUBSCRIBE to [EMAIL PROTECTED] > To subscribe / unsubscribe: http://www.dfwcfug.org > > ----------------------------------------------- To post, send email to [EMAIL PROTECTED] To unsubscribe: Send UNSUBSCRIBE to [EMAIL PROTECTED] To subscribe / unsubscribe: http://www.dfwcfug.org
