Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[stem-dev] Notes from the STEM call on November 3, 2016

Notes from the STEM Community Call, November 3, 2016

On the call: James Kaufman, Stefan Edlund, Kun Hu, Matthias Filter, Emily Nixon, Judy Douglas

 

The next call will be in early December; date to be announced

 

Introductions: Emily Nixon, PhD student at the University of Bristol in the UK, doing mathematical modeling of disease in sheep

 

1. STEM Integration Build is up, stable

*Stefan: Will do formal Milestone with Eclipse in the next few weeks

*Date for new version: Deferred until ready to do new feature work

 

2. Ongoing Feature Work, no updates

*Nereyda (not on call): Using STEM to model pandemics in biodefense

 

3. Updates: Kun

*Primary concern of Santa Clara public health officials is TB (hard to model)

*Interest in Zika in Brazil, Southeast Asia

*Will meet with Singapore Minister of Health in December (100s of reported; multiply by 5 since only 1 in 5 symptomatic)

*Will work with 6 IBMers in Panama on collecting info in real time using mobile devices, on the street interviews (in one of 5 projects selected by IBM)

*Philippine team’s work on model on hookworm infection appears to be proceeding without difficulty

*Sent Werner slides for his presentation at Eclipse Europe

 

4. Other Work: Jamie

*Roslyn (Australia IBM) worked with Taiwan CDC, taught STEM and new dengue model using interventions (another one of the 5 projects)

 

5. Bugs: None  

     

6. Discussion

MODELING

*Emily: In looking at disease in sheep on neighboring farms, did a lattice; didn’t fit with map coordinates for UK; slow work. A quicker way?

*Jamie: Perhaps a square lattice; Chris Thoens did something like this; asks Emily to place a feature request on stem-dev requesting a lattice graph inside a polygon; advises phase has to be right 

*Emily: How to define county to county spread?

*Jamie: Typically model at highest resolution, use containment nodes, perhaps aggregate farms

*Stefan: Use migration edges between farms, connect up to a central node

*Emily: Some farms use common grazing

*Jamie: Flatten model; don’t use admin zones; label farms, roll up results, then do cross county

*Jamie: Ability to re-export a graph back to Pajet a feature request; Emily can stay with spreadsheet

 

PROFESSIONAL CONFERENCES

*Emily: Any especially helpful?

*Jamie: Epidemics (USA-based); International Society for Veterinary Epidemiology and Economics (ISDEE) in 2018 in Bangkok

 

POSSIBLE HACKATHON AT BfR

*Matthias: Has master’s student, invites Emily to BfR; will look at Chris’ salmonella model; will communicate with Emily via email

 

STEM USER INTERFACE

*Matthias: Took his master’s student 14 days to get STEM running (Java issues)

*Jamie: No internal funding for STEM, but committed to open source project; asks Matthias to articulate what he’d like

*Matthias: At IBM’s 5 selected projects, create additional videos on using STEM

*Jamie: Matthias’ student could make a You Tube video about his problems; IBM team could highlight, put on wiki; need to leverage the open source community

 

7. Items from Participants

*Stefan: After 10 years, will be putting STEM build on a new server, will run faster

*Emily: How to update? Jamie: Doc on wiki; don’t need to update every week; do NOT delete old version; all your work is there; currently a Windows problem with updates, will help Emily with this

*Judy: Lancet piece on potential Zika threat in Pakistan; Kun commented it seems on target

  

 


Back to the top