Skip to content

Latest commit

 

History

History
58 lines (49 loc) · 1.52 KB

File metadata and controls

58 lines (49 loc) · 1.52 KB

TeamGOATS

Zika virus forecast

###################### CONTACT INFO Kelly Schroeder kschroed@bu.edu 307-399-5248

Emily Chua ejchua@bu.edu 617-840-6331

Stephen Caron caronst@bu.edu 781-264-4383

Andrew Pineda ajpineda@bu.edu

Andre Schettino schettin@bu.edu 617-784-1217

###################### DATA -Buzzfeed Repo: -Colombia's Zika data updated weekly from 2016-01-09 to 2016-02-20; stored in the table "total" within get_data.R

-Google Trends: -Data updated daily and stored in "zika_trends"" within
get_googletrends_data.R -NOTE: Did not end up using this data

CRON SETTINGS -All updates emailed to zikaforecast@gmail.com

###################### ORDER OF R SCRIPTS

  1. get_data.R -sets the working directory -scrapes Github for Buzzfeed Colombia Zika data -divides up Colombia by department -saves all objects into "zika.RData"
  2. model_test_new.R -fits the model to the 7 weeks of available data using a Random Walk model -includes a departmental effect -visualizes the model and data time series with confidence estimates -visualizes the MCMC diagnostics and saves the MCMC output
  3. initial_forecast.R -generates an initial ensemble forecast for the 7 weeks + out through weeks 8 to 20 -uses 500 MC simulations -holds out a randomly-chosen initial forecast as our "pseudo-data"
  4. sensitivity_analysis.R -performs the sensitivity and uncertainty analyses
  5. EnKF.R -uses an Ensemble Kalman Filter to iteratively update the analysis and forecast steps on a weekly basis