[Babase] Ranker Meeting

Lacey Maryott lacey.maryott at duke.edu
Wed Apr 25 09:03:51 EDT 2007


Hello all :)

    We are writing to update everyone on the status of ranker, and the 
meeting we had today with Tyler Brock and his senior thesis committee 
members (Susan, Jun Yang, and Jeffrey Forbes, another CS faculty member 
at Duke).  We'll start by describing the capabilities of the program and 
interface, then talk about what needs to happen before it will be at a 
stage where we can utilize it.

The program and interface have the ability to
- Create matrices and rankings from the parts and interact data on 
dyadic agonisms
- Save and Load Rankings
- Write to and Read the Database
- Start new rankings/matrices without logging out
- Do and undo manual ranks
- Load previous rankings
- Auto load of most recent ranking
and possibly the most exciting functions:
- Automatic ranking
- A function that allows the user to re-rank individuals in the matrix 
by drag and drop on the screen.

*The interface requires a log in and that log in will be the same as 
your PPA login.  As noted above, in this program, individuals along the 
top of the matrices, as well as down the left hand side of the matrices 
can be moved via "drag and drop" which "revolutionizes" our ability to 
quickly and easily minimize entries below the diagonal by moving 
individuals within a matrix.  This is very exciting and has the 
potential to save a lot of time as well as having the potential to 
reduce human error.

Some things that need to happen before the program can be used are:
- When you choose the option to save the rankings you have assigned to a 
rank report, a text file is created with an sname and a rank number 
beside it. Currently the numbering starts with a rank of 0, whereas we 
need ranks to start at 1
- At this point, the program only has the capability to rank groups in 1 
month intervals, whereas it would be necessary to be able to do months 
at a time.
- A scroll bar needs to be added such that matrices of large groups of 
individuals can be viewed on a normal computer screen

These are all things that Tyler feels can be managed within the time he 
has remaining here at Duke along with documentation and commenting on 
and finalizing the code for the program.

Karl,
    in reference to some of the questions you had sent me to the meeting 
with;
- once finalized, Tyler will be sending you the code with 
comments/information such that the code is useful/meaningful to you.
- In terms of getting the program into Papio, he just needs to sign the 
new jar files he has altered, and he will then send you the jar and jnlp 
files ( I think he said it was 4 files total) so that you can put them 
up on Papio.

Summary: the new ranker program should be available as a link from 
papio.biology.duke.edu within a couple of weeks. It will allow us to 
assign ranks in the same way that the old version of ranker did in 
FoxPro. It also has an additional tool that implements an automatic 
ranking program. My guess is that, at this point, we will not use this 
automatic ranking program very much. It represents a first stab at what 
will be, in the long run, a complicated thing to make happen (and it 
lays important groundwork for taking this forward). It is exciting to 
see the first attempt and think about the possibilities, but the major 
step from the babase user perspective is a functioning, enhanced version 
of the ranker that we know and love that will work with our new database.

We are sure that as users start to work with it, many things will come 
to light that we will want to modify; we'll keep a list of these things 
and go forward from there.



Cheers,
Lacey and Susan


-- 
Lacey Maryott
Department of Biology
Duke University
ph: 919-660-7306
fax: 919-660-7293
Lacey.Maryott at duke.edu 



More information about the Babase mailing list