Subgrouping

[Zsolt Udvari]

Hello!
It would be nice when can create subgroup by multiple aspect.
What I think? It exists a group, name GA. They're splitted on a subject (SA) to some subgroups (SGA1, SGA2, SGA3,...). And there's an other subject (SB), where splitted too, only other subgroups (SGB1, SGB2,...).
E.g. the students learn english language, beginner and andvanced. They're beginner and advanced subgroups (because of english). They learn second language, france, german, roman, etc., so they're three (or more) subgroups, because of 2nd language. And the two subgrouping is independent.
I hope, I was clear.

[Liviu Lalescu]

Why cannot you use the "divide years automatically" facility? It is similar, only it operates on years, not groups. But you can divide a year by 3 categories: 1) a, b, c,..., 2) advanced, beginners, 3) French, German.

I could add more than 3 categories for a year, but the generation may become too slow. So, I advise people to use tricks to have only 3 categories.

You could define a real group = a FET year, and divide the real group into 3 categories.

[Zsolt Udvari]

Hm, it seems, it will be my "dream"

[Liviu Lalescu]

Hm, it seems, it will be my "dream"

You mean you like the divide year facility and you can apply it?

[Zsolt Udvari]

Yes. And as I see, it works well for me. Maybe you can add more categories but it isn't very important (as I calculate, I have max. 5 categories on a year).

[Liviu Lalescu]

Yes. And as I see, it works well for me. Maybe you can add more categories but it isn't very important (as I calculate, I have max. 5 categories on a year).

Too many categories will slow down the generation, that is why I want people to use other tricks in this case.

[Zsolt Udvari]

It isn't very important, I can solve it with tricks and plus work. But if you apply this, I will test it...

[Liviu Lalescu]

It isn't very important, I can solve it with tricks and plus work. But if you apply this, I will test it...

I would not like to apply this, because people may abuse too many categories and the generation will get slow (or very slow).

In many categories might be zero students, so the category is useless, but FET generates its timetable. This is bad for the speed.

If I will change my mind, I'll let you know.

[Zsolt Udvari]

Why slows down? Because the number of subgroups? When yes, I think, the maximalizing the number of subgroups will be better.
When I have four categories, both categories has two groups, it has 2^4=16 subgroups. But when I have two categories with 6 groups per category, I'll have 6^2=36 subgroups. So why can't create the first case?

[Liviu Lalescu]

People might use 6 categories with 6 divisions each, 6^6 subgroups. This will slow down.

Slow down is because of number of subgroups, indeed.

[Zsolt Udvari]

And anybody wants use only 6 categories with 2 subgroups per category? This (64) is less then the current "maximum": 3 categories with 6 subgroups (216).
I think the second case slows the generation better. And this is the reason why I said that can maximalized the number of subgroups instead of categories.
I needed now in some years about 5 categories, 2-3 groups per category. The maximum 6 subgroups needed only one time.

[Liviu Lalescu]

You are right, but when I show the user 6 categories, with say: 12, 6, 6, 3, 3, 3 divisions, he expects to be able to add all of them, he will be puzzled if I say: max number of subgroups should be 216. Also, some users reported that they 12, 6, 6, 6, 6, 6, which is too much.

I hope you can use tricks. It is better, because the generation is faster.