Practical useful features

[mattlange]

In my special topic (Learning Spaces) I am analyzing the dependecy of questions of a questionary. There will be used association rules of the form

P can solve Task 1 and Task 2 => P can solve Task 3
P cannot solve Task 1 and cannot solve Task 2 => P cannot solve Task 3

Some features would be very helpful:

1. Researching association rules
   a) Setting up support and confidence BEFORE the research of association rules is starting
   b) Setting up the maximal numbers of attributes in the premise of an implication/association rule
   c) Giving some information about the necessary time and the progress of research (actually the user don't know whether it's computed in one hour or in 3 months)
2. Re-Plotting the lattice
   User can add an rule and lattice will be computed new (as example: user can interprete an association rule as implcation because of confidence 99% und want to look up the new lattice)
3. Context working
   a) Can load contexts in csv-form
   b) Can invert the context
   c) Can delete by button all attributes with support <s

[blu2lz]

1c) is already there, progress bars are displayed in 0.6. I will think about a possibility to calculate the remaining time.

[blu2lz]

Since this is split up into multiple issues which are all in future, I suppose it is okay to move this to future, too. I leave it open to work through it again after fixing related issues to ensure that we did not overlook anything.