[AccessD] miracle required apparently

[Paul Wolstenholme]

Stuart,

You are the first person I know of to try that without a computer.

I'm not sure I can add much to your solution.  I expect I started the main
query in a qbe grid but not all the logic is easy to express that way.
The next step is to switch to SQL view and expand on the WHERE clause by
hand with your logic properly formatted (knowing that a save in Access will
trash the formatting - so do keep a formatted copy that is readable by
humans elsewhere).

As you are probably aware, your WHERE clause can use pretty much any
statements and any levels of parentheses you like that makes sense to you.
It doesn't need to be entered in canonical form - Access does that
conversion internally.

To take this further you could include preferences by creating a quality
field to sort on and selecting the top results.

Paul Wolstenholme


On Wed, 18 Oct 2023 at 14:21, Stuart McLachlan <stuart at lexacorp.com.pg>
wrote:

> I used to do those logic problems all the time with pen and paper.
>
> That one is pretty simple.
> Every value of every parameter is named
> A 5x5 grid and simple case of deduction/elimination.
> :)
>
> On 18 Oct 2023 at 12:10, Paul Wolstenholme wrote:
>
> > There is a similar problem at
> > https://www.rd.com/article/einsteins-riddle-solve-it/
> >
> > I had a colleague who wrote code (probably in C or a derivative) that
> > took about a day to execute in 1999 and gave an answer. I wrote an
> > Access 95 query that took about a second to reveal the possible
> > variations that are encompassed in the ambiguities of the problem
> > statement.
> >
> > My technique was first to form a table of all the permutations (in
> > this case of 5 things, meaning 120 entries). Next I formed a query for
> > each thing that could lie in any of those permutations. Finally I
> > formed a query that started as the cross product of each of those
> > queries but was linked according to the rules of the problem. (After
> > that I showed the final query's results in a form).
> >
> > I expect my former colleague is still plotting his revenge.
> >
> > Paul Wolstenholme
> >
> > On Wed, 18 Oct 2023 at 08:14, Paul Hartland via AccessD <
> > accessd at databaseadvisors.com> wrote:
> >
> > > I don't have much time until the weekend for knocking up a quick
> > > test, but think that scenario is pretty doable with 4 lots of times,
> > > 4 lots of classes, upto 8 participants and not being the same time
> > > or class zone more than once
> > >
> > > Paul
> > >
> > > On Tue, 17 Oct 2023, 02:22 Steve Schapel, <
> > > steve at datamanagementsolutions.biz>
> > > wrote:
> > >
> > > > Hi all
> > > >
> > > > I'm trying to do something that I initially thought would be
> > > > reasonably easy.  But alas, so far success has eluded me.  Any
> > > > insights accepted with much gratitude!
> > > >
> > > > The goal:  Assign a number of Participants to a number of
> > > > Activities over a number of Sessions.
> > > >
> > > > Very simple example:
> > > >      4 Sessions
> > > >      4 Activities
> > > >      8 Participants (therefore 2 per Activity)
> > > >
> > > > To illustrate:
> > > > Let's say the sessions are 9am, 10am, 11am, 12pm.
> > > > Let's sat the activities are chess, tai chi, bowls, diving
> > > > Let's say the participants are A, B, C, D, E, F, G, H
> > > >
> > > > The stipulation is that each participant should do each activity
> > > > one time.
> > > >
> > > > By trial and error brute force, I know that there is at least one
> > > > solution, namely:
> > > >
> > > >                       chess       tai chi       bowls
> > > >                       diving
> > > >    9am           A & E        C & F        D & G        B & H
> > > > 10am           C & G       A & H       B & E         D & F
> > > > 11am           D & H      B & G        A & F         C & E
> > > > 12pm           B & F       D & E        C & H        A & G
> > > >
> > > > HOWEVER, I have tried multiple angles of looping through nested
> > > > (and sometimes randomised) recordsets based on the core data
> > > > elements (sessions, activities, and participants), to write the
> > > > assignments to the available slots in a schedule table, and to my
> > > > shock (and horror) we always reach the point in the procedure
> > > > where it gets stuck, due to trying to assign a participant to two
> > > > activities in the same session, but with no valid alternative slot
> > > > available.
> > > >
> > > > There is no problem if the model calls for the super simple option
> > > > of only one participant for each activity.  But otherwise, no
> > > > dice, so far.
> > > >
> > > > There MUST be a way to make this work?  Surely?
> > > >
> > > > Thanks a lot.
> > > >
> > > > Regards
> > > > Steve
> > > > --
> > >
> > >
> > --
> > AccessD mailing list
> > AccessD at databaseadvisors.com
> > https://databaseadvisors.com/mailman/listinfo/accessd
> > Website: http://www.databaseadvisors.com
> >
>
>
> --
> AccessD mailing list
> AccessD at databaseadvisors.com
> https://databaseadvisors.com/mailman/listinfo/accessd
> Website: http://www.databaseadvisors.com
>