Bjoern Milcke wrote:
...
I am not sure if Venn diagrams are really diagrams you would like to
use in OOo, maybe I am just lacking some examples. Does anybody have
some examples of data plus a corresponding Venn-diagram? I have the
impression that Venn-diagrams would rather make more sense to be
created in Draw or Impress by hand using some shapes that overlap.
And what would data look like? It would have to be a representation of
elements belonging to different sets. Even if such diagrams would make
sense, I still believe they are of little importance. Apart from the
typical static two- and three- set examples you might know from
school, I haven't seen any Venn-diagrams in newspapers or other
publications.
I am misusing the term Venn-diagram a little bit (when actually meaning
more something like an Euler diagram).
The typical case is when you have overlapping data, e.g.:
- one group of 600 patients has a disease A
- another group of 160 patients has disease B
- and 60 patients have both diseases
It is much easier for the audience to understand the distribution of the
diseases if one uses such an Euler diagram (Venn diagrams are very
particular diagrams). This becomes even more important when dealing with
more than 2 groups of overlapping data, when many overlaps become
possible. [Again, in this case, an overlap should be probably explicitly
displayed only IF > 1% of the data is present inside the overlap region.]
It is fairly common in research (especially biomedical) to have a group
of patients and describe the group in terms of chronic diseases present
in the group. [Actually every clinical trial should describe the group
of patients in terms of diseases present.]
What I often miss in those situations is the distribution and overlap of
various diseases in the group. Two persons having each only one of the
diseases A, respectively B, will be quite different from a person having
both diseases and one having non:
aka 1A + 1B vs 1(A+B) + 1(non)
It is indeed true that such graphics are somewhat sparse in the
literature, but I fear that the lack of a method to easily produce such
graphics is one of the main reasons for this. [This is not limited to
mere presence of diseases; one can easily imagine different scenarios,
like overlapping treatments and other overlapping attributes and the like.]
So I see great usefulness for this type of diagram (at least in
biomedical research).
Sincerely,
Leonard
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
