[stem-dev] Notes from today's STEM community call

["Judy Douglas" <judyvdouglas@xxxxxxxxxxx>]

STEM Community Call July 26, 2012

On Call: Jamie, Alfiya, Kamesh, Stefan, Matt, Kun, Chris, Werner, Judy

 

NO CALL NEXT WEEK, AUGUST 2nd

NEXT SCHEDULED CALL AUGUST 9th

 

1. Introductions 

Kamesh: Mathematical science focus; working on 2 year Singapore project on dengue analytics and monitoring using STEM

Alfiya: Master’s student focused on research and data mining; new to STEM and STEM project

2. Methods of communication, project participation for new participants

W/ Eclipse ID, can use stem-dev, bugzilla; contact via email, Skype; wiki available

Can contribute by attaching new material as patch to a bug; Eclipse committer can then commit

Can become committers via Eclipse process involving vote by existing committers

Kamesh: Procedures? Jamie: Defined by Eclipse open source license; only public data can be added to STEM    

3. 1.4.0 Release Plan

Stefan/Matt: Rebuilt Milestone 2 tested; checks out ok, stable

Will publish, upload today, Thursday, July 26, 2012

 

Features include                     
    Bug Fixes from 1.3.1
    Performance improvements
    Simplified creation of new Disease and Population models (Tools for generating new disease plugins)
    Pajek-file Importer:  
    Enable import of discrete transportation events
    Enable integration between user graphs and existing STEM graphs
    New MapView to visualize the discrete transportation events and the ability to display multiple views of the same simulation
    Polio Example
    Stochastic Modeling Framework
    Ability to Delete Items inside a Graph from Designer Perspective
    Deep copy drag and drop for models
    Refactor
    Create new mixing edges
    Improved UI for decorators view
    New Implementation of STEM Preferences
    * Time adaptive external data source models: Newly added as deliverable; TODO: Put on web plan page   
         
4. NEW DOC FOR 1.4.0
* wiki practice
* mixing edges and mixing edge graph generator (DONE) integrated in transportation models page (done). posted on newsgroup
    * map view features    
    * Pajak import tutorial (Done)
 * Best Practice for Developers, writing performance critical components in STEM (done)
 * additional doc for more realistic food SCENARIO doc is done but not in wiki yet. Scenario today.
 * Need better top level food modeling doc (draft created)

5. Food production discussion. Latest data/results

Danny not on call; work @ Almaden w/ toy model shows landscan data works well

Jamie: Will discuss in Berlin

6. ISVEE2012 BfR poster

Chris: Alex is doing this, will check on status; Jamie: Can help w/ this when in Berlin

7. BUGs of the week:
External Data source models (disease and population) must adapt time

Stefan: Making good progress on this w/ daily mosquito data; interpolating, need to be sure in sync

8. Links, conference info from Werner
On dengue: Research w/ genetically modified mosquitoes: http://www.gmx.net/themen/gesundheit/krankheiten/129dti8-brasilien-genmanipulierte-muecken-dengue and http://scienceray.com/biology/british-scientists-created-mutant-mosquitoes-to-combat-dengue/
On mosquito maps: http://www.vet-magazin.de/deutschland-magazin/aktuelles/Mueckenkarte-Deutschland.html
On Eclipse Day Poland September 13: IBM Poland may be involved; Werner will forward email from organizers to Jamie; Chris will find out if BfR would like to go; if so, Jamie will see if IBM can fund; if not, Werner may present

9. No Call on August 2nd (1st week in Berlin)

Almaden team still available by email & via stem-dev to answer queries  

Next call will be on August 9th at the usual time

10. Items from Participants

Kamesh: Getting data from Singapore on dengue incidence, also spatio-temporal data on mosquitoes; using 3 models

DISCUSSION

1. Need finer map: Currently STEM has main island (not subdivided), plus outlying islands; have polygons, need nodes; if better map available, can put into STEM

2. Need transportation information: Travel into and out of Singapore

3. Need to know how many dengue serotypes in Singapore. Four in Asia; have 30+ years of data from Thailand that shows dengue to be quasi-periodic, w/ 3-4 year periods, always a different serotype; IBM has raw data (not in STEM but public); Hopkins the best curated data set (not public); correlations bet dengue in Singapore & Thailand & (if data available) Malaysia might be of interest    

4. Need mosquito data w/ time dependence