Quickstart Guide
This guide will walk you through running your first OSeMOSYS-RDM analysis.
Overview
OSeMOSYS-RDM provides a streamlined workflow for:
Base Future: Execute a single baseline scenario (“Future 0”)
RDM Experiment: Generate and evaluate multiple futures using uncertainty ranges
Postprocessing: Consolidate results into analysis-ready datasets
PRIM Analysis: Perform scenario discovery to identify key drivers
Step 1: Prepare Input Scenarios
Place your OSeMOSYS scenario/data files (GNU MathProg format) in the scenarios directory:
src/workflow/0_Scenarios/
Your scenario file should be a valid GNU MathProg data file compatible with OSeMOSYS.
Note
A reference formulation consistent with this workflow is included as model.v.5.3.txt.
Step 2: Configure the Run
Open the main configuration interface:
src/Interface_RDM.xlsx
Key Configuration Sheets
Sheet |
Purpose |
|---|---|
|
Solver selection, run toggles, region, model name |
|
Outputs to export from the model |
|
Define uncertain parameters and their ranges |
Essential Setup Parameters
In the Setup sheet, configure:
Solver: Choose from
glpk,cbc,cplex, orgurobiRun_Base_Future: Set to
Yesto run the baselineRun_RDM: Set to
Yesto run the uncertainty analysisRegion: Your model’s region identifier
Number_of_Runs: Number of futures to generate (for RDM)
Step 3: Run the Pipeline
Execute Complete Workflow
# Run the complete RDM pipeline
python run.py rdm
# Run PRIM analysis (requires RDM results)
python run.py prim
# Run both sequentially
python run.py all
Example: First Run
# Navigate to the project directory
cd osemosys-rdm
# Run with default settings
python run.py rdm
You should see output similar to:
======================================================================
AFR_RDM Pipeline Runner
======================================================================
Module: RDM
Environment: AFR-RDM-env
======================================================================
🔧 Step 1: Environment Setup
✓ Conda environment 'AFR-RDM-env' already exists.
🔧 Step 2: Dependency Management
✓ All conda packages are present.
✓ All pip packages are present.
🔧 Step 3: Git Repository Check
✓ Git repository detected.
🔧 Step 4: DVC Initialization
✓ DVC repository detected.
======================================================================
🔬 RDM Pipeline (Robust Decision Making)
======================================================================
Stages: base_future → rdm_experiment → postprocess
======================================================================
🔄 Executing RDM Pipeline...
----------------------------------------------------------------------
...
Step 4: Review Results
After execution, results are available in:
src/Results/
Output Files
File Pattern |
Description |
|---|---|
|
Consolidated model outputs |
|
Consolidated model inputs |
|
Efficient storage of all futures |
|
Efficient storage of all inputs |
What’s Next?
Customize uncertainty parameters: Edit the
Uncertainty_TablesheetAdd more scenarios: Place additional
.txtfiles in0_Scenarios/Explore PRIM results: Check
src/workflow/4_PRIM/t3b_sdiscovery/