I've been waiting 6 hours and had no timetable

Started by AZ, September 12, 2012, 04:40:14 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

AZ

hi
I launched FET before going to school; when I came back home it was stil computing!
Is there a way to easily find mistakes in putting constraints?
activity n 764 is the last placed by FET
No: 764, Id:199, Insegnanti: Conterno, Materia: Laboratori tecnologici, Chiavi dell'attività: , Studenti: 2 A
No: 765, Id:200, Insegnanti: Conterno, Materia: Laboratori tecnologici, Chiavi dell'attività: , Studenti: 2 A
how can I understand the problems with 765?
I'm looking for a some kind of debugging strategy, not to continously ask for help ...

thanks
Antonio

Liviu Lalescu

Unfortunately, there is no general strategy and FET cannot indicate the problem in general. It is like a puzzle.

I hope other users can give more advice than me.

An increasing approach might be good (to make the timetable increasingly difficult, but still possible).

Volker Dirr

check the (unfished) timetable and try to place the activity manualy. normaly you can the the problem then.
(it also help sometimes to increase the max number of hours, so the timetable might be set complete (of course incorrect), but you can think again manualy to find the bug.)
(it also help if you delete some other years (and(or constraints) and genarate again. (because you can see the bug faster by that.)

mbarsan

Hi.
I want to suggest another strategy that helped me to find some error inside the design of data.
Go to menu Advanced/ Spread the activities... and select ALL the 4 options. Furthermore set all weights to 100%.
Remember to use the next function, to remove redundant constraints.
When you try to generate with your data (see first post), you get following message:
Time constraint
Minimum number of days between activities
Weight (percentage)=100%
Number of activities=3
Activity with id=733 (T:Gasparri, S:Ecologia applicata, St:5 D)
Activity with id=734 (T:Gasparri, S:Ecologia applicata, St:5 D)
Activity with id=735 (T:Gasparri, S:Ecologia applicata, St:5 D)
Minimum number of days=1
Consecutive if same day=yes
cannot be respected because teacher Gasparri has at most 2 available days from teacher not available, breaks and teacher max days per week. Please lower the weight of this constraint to a value below 100% (it depends on your situation, if 0% is too little, make it 90%, 95% or even 99.75%. Even a large weight should not slow down much the program. A situation where you may need to make it larger than 0% is for instance if you have 5 activities with 4 possible days. You want to spread them 1, 1, 1 and 2, not 2, 2 and 1)


Well this is self explanatory: you constrained Gasparri to work only 2 days per week. You have to choose if Gasparri will work 3 days per week or if he has only two subactivities in 5D group (2+1 instead of 1+1+1)

When you go to see next message, FET says:

Time constraint
Minimum number of days between activities
Weight (percentage)=100%
Number of activities=6
Activity with id=1133 (T:Bisarello, S:Laboratori tecnologici ed esercitazioni, AT:MAX3_lab_tecn, St:1 G)
Activity with id=1134 (T:Bisarello, S:Laboratori tecnologici ed esercitazioni, AT:MAX3_lab_tecn, St:1 G)
Activity with id=1135 (T:Bisarello, S:Laboratori tecnologici ed esercitazioni, AT:MAX3_lab_tecn, St:1 G)
Activity with id=1136 (T:Bisarello, S:Laboratori tecnologici ed esercitazioni, AT:MAX3_lab_tecn, St:1 G)
Activity with id=1137 (T:Bisarello, S:Laboratori tecnologici ed esercitazioni, AT:MAX3_lab_tecn, St:1 G)
Activity with id=1138 (T:Bisarello, S:Laboratori tecnologici ed esercitazioni, AT:MAX3_lab_tecn, St:1 G)
Minimum number of days=1
Consecutive if same day=yes
cannot be respected because teacher Bisarello has at most 5 available days from teacher not available, breaks and teacher max days per week. Please lower the weight of this constraint to a value below 100% (it depends on your situation, if 0% is too little, make it 90%, 95% or even 99.75%. Even a large weight should not slow down much the program. A situation where you may need to make it larger than 0% is for instance if you have 5 activities with 4 possible days. You want to spread them 1, 1, 1 and 2, not 2, 2 and 1)


This is the same situation for another teacher who works 5 days per week. It is impossible for him to have 6 activities with minimum number of 1 day between.

You may improve your data and then optimize again using a lower weight instead of 100%

Often this trick doesn't help anyway. But sometime this will be...
Hope this may help!

AZ

thanks a lot for your useful  suggestions;
we really appreciated them;
unfortunately, they didn't work!
we are three teachers working on this timetable, everyone with his own mind (not to influence each other), but a solution is far to come!
anyway, thanks a lot.
Antonio