Re: Cardinality - I really need help

From: David Cressey <>
Date: Thu, 09 Aug 2007 12:12:50 GMT
Message-ID: <6xDui.7028$I64.5168_at_trndny08>

"beginner16" <> wrote in message
> hello
> Sorry for not replying sooner, but I just started learning Smalltalk
> and it kind of a drag. Anyways, I hope someone will still read this
> reply...
> On Aug 5, 10:02 pm, Hugo Kornelis
> <> wrote:
> > On Sat, 04 Aug 2007 12:23:28 -0700, beginner16 wrote:
> > >hello
> >
> > >I would really need some help here
> >
> > >Some time ago I made a thread about cardinality and thought I
> > >understood it , but today I noticed that I actually misread a post
> > >explaining cardinality and realized that the subject still very much
> > >confuses me. Below are the original question and a reply ( the one I
> > >misread)
> >
> > (snip)
> >
> > Hi beginner16,
> >
> > I've used various methods of entitytype relationship diagramming, and
> > there are many different to depict cardinality - one of the major
> > problems when switching between diagramming styles, as that some will
> > depict minimum and maximum cardinality on one side of a relationship,
> > others on another side of a relationship, and then there are some that
> > depict minimum cardinality on one side and maximum on another side!!
> >
> > Your model looks familiar, though I don't recall at the moment how the
> > diagramming style was called when I learned it. Anyhow, if this is the
> > style I think it is, than the information in your picture says that:
> >
> > - Each E1 is associated with at most 1 E2, as indicated by a vertical
> > bar instead of a craw paw (is that the correct English translation?) at
> > the E2 side of the line (maximum cardinality); and
> > - Each E1 does not have to be associated with any E2, as indicated by a
> > circle rather than a vertical bar at the E2 side of the line (minimum
> > cardinality);
> >
> > Since there is neither a vertical bar nor a craw paw, and neither a
> > vertical bar not a circle at the reserved spots at the E1 side of the
> > line, nothing is known about how many E1 may or must be associated with
> > any given E2.
> >
> > You may have noticed that I did not phrase any of these cardinalities as
> > being a cardinality "of E1" or "of E2" in this relationship, since the
> > interpretation of the term "cardinality of E1" tends to be as ambiguous
> > as some diagramming styles.
> >
> > Best, Hugo
> Anyways, I thought that a term cardinality always means the same,
> regardless of diagramming style used. I can see how one diagramming
> styles may put min and/or max cardinality on one side and others on
> other side, but I assumed that the term itself always means max and
> min number of connections an entity ( call this entity A ) of certain
> type can have with other entities --> so if A can have min zero and
> max one connection with B entity, then cardinality of A is (0, 1). I
> find it hard to believe, that depending on diagramming style, we would
> say that cardinality of B is (0, 1)

I would have worded it differently. Connections between entities exist in the context of a relationship. If, in relationship R (A, B), each A can be connected to a minimum of zero B's and a maximum of 1 B, then the cardinality of R with respect to A is (0,1). I am not sure whether your wording and mine are equivalent or not.

Note particularly that what I'm referring to is a "cardinality of the relationship", and not a "cardinality of an entity". In the terminology I'm accustomed to, an entity has but one cardinality: the number of instances of that entity. Received on Thu Aug 09 2007 - 14:12:50 CEST

Original text of this message