-
Notifications
You must be signed in to change notification settings - Fork 2
RequirementsSpec
Oliver Mannion edited this page Jun 28, 2015
·
1 revision
- base file input
- ability to create individual “agents” from multiple data sets which may require some data manipulation and merging
- parameter input
- ability to load multiple tables of parameters from CSV/XLS files, which can be modified in the user interface to test different scenarios and forecast forward
- scheduling mechanism
- very simple discrete time based - state changes occur at every tick (year)
- simulation techniques
- stochastic equations (require random number generation)
- ability to clone “agents”
- output
- aggregate: charts and tables of aggregate information per year eg: freq distributions, mean, stddev, i.e. descriptive statistics and perhaps more; facility to weight and reweight the sample to obtain representativeness and for scenario testing
- individual: the complete dataset (ie: state of each individual) after each year; ability to track an individual history (state changes over time)
- NB: aggregate and individual yearly outputs will be used for
- internal verification,
- external validation/alignment/calibration to external benchmarks
- more sophisticated analysis (eg: via SAS) that can’t be performed in the simulation software
- user interface
- a GUI for end users to control input and output, and run scenarios. Should be able to do multiple runs (with same parameters but different random seeds) to obtain a robust mean estimate. It should be intuitive and “easy” for an end user to use.
- performance
- external end users should be able to perform multiple runs of a multi-year simulation on a desktop PC in a reasonable amount of time (X mins)