

DONE
- implement problems function that passes through or simulates data to a given design
  - parameters: should accept different assumptions about latent period, population turnover rates
 - manual correction of data-entry errors
 - conversion of predictors
 - simulation of epidemic time series
 - NAify filter
 - continuous to discrete filter
 - midpoint rule transformation
 - imputation of missing data
 - get MLE dispersion parameter for gamma
 - parameterize simulation gamma like R glm
 - vary proportion imputed
 - determine jumpahead from jobid
 - scale up on cluster
 - summarization
  - get post processing runing fast
 - limit infective period of staff to 1 day
 - note new method of dealing with ~0-length infectious periods, new method of determining Yij
- note truncation of infectious period of staff
- tally failed estimation runs in bootstrap
- check bootstrap likelihood distribution
- reports: ecdf of loglik, proportion of failures to find estimates
 - make table 1 


MAYBE
- get build system working: current block is dependency cycle with get-ests and EpiGLM
- write up/draw up data-processing pipeline



coding TODO

 - put latest code up on openwetware

writing TODO
