Auto-GTAP combines all steps of running a GTAP-based research project into a single automated framework.
In principle, a GTAP simulation can be run with a single command. However, GTAP-based research projects usually contain more than just the model simulation. A project may also need to build a database, change aggregations, modify parameters, calculate shocks, or run multiple simulations, all the while shuffle around various files at each stage. While a single model simulation is already automated, these other task are not.
Auto-GTAP combines all these tasks into a single automated framework. By reducing the amount of manual orchestration required to run a GTAP project, Auto-GTAP will increase the speed and replicability of your analysis.
Mandatory
- Python 3.6 (other versions untested)
- Pyyaml 4.2b4 (other versions untested)
Additional requirements for GTAP-E Validation advanced example
- GEMPACK 14.0 (other versions untested, purchase information)
Additional requirements for Productivity Increase advanced example
- GEMPACK 14.0 (other versions untested, purchase information)
- GTAP Data File v10_2014 (purchase information)
cd your_project_directory
git clone https://github.com/andre-barbe/Auto-GTAP.gitpython -m venv venv ./venv/Scripts/activate- Script path:
Main.py - Source root:
src - Working directory (depends on example):
examples\productivity-increase-simple
This example includes a step to aggregate GTAP data, and thus requires the disaggregated data file from GTAP. This data file is not freely available but must be purchased from GTAP. Thus we have not included it with Auto-GTAP: users must manually acquire it and place it into Auto-GTAP.
The specific data used in this example was the PR2_v10_2014_Sep2017 release of GTAPg2. It was located in
GTPAg2\GTAP10p2\GTAP\2014
and should be placed in
Auto-GTAP\examples\productivity_increase\InputFiles\GTPAg2\GTAP10p2\GTAP\2014