Skip to content

Latest commit

 

History

History
83 lines (61 loc) · 2.33 KB

File metadata and controls

83 lines (61 loc) · 2.33 KB

Pipeline

Invest-GT is easiest to operationalize as a staged research pipeline rather than as a monolithic model. The goal is to move from explicit assumptions to reproducible runs and structured evaluation.

Recommended pipeline

00. Inputs and world modeling

Define the baseline world representation used by all downstream work:

  • actors, institutions, alliances, blocs
  • macro variables and policy variables
  • event types and constraints
  • asset universe or allocation targets
  • scenario assumptions and time horizon

Primary workspace:

  • pipeline/00_inputs/

01. Hypothesis definition

Write down the claims the system is supposed to test before building runtime logic.

Primary workspace:

  • pipeline/01_hypotheses/
  • hypotheses/

02. Agent design

Specify actors, incentives, constraints, beliefs, and strategy spaces.

Primary workspace:

  • pipeline/02_agents/
  • agents/

03. Model design

Define payoff structures, update rules, interaction forms, and propagation logic.

Primary workspace:

  • pipeline/03_models/
  • models/

04. Runtime assembly

Connect the separate components through explicit interfaces and configuration rules.

Primary workspace:

  • pipeline/04_runtime/
  • runtime/
  • configs/

05. Simulations

Execute scenarios, sweeps, and experiment variants.

Primary workspace:

  • pipeline/05_simulations/
  • simulations/

06. Tests and validation

Compare outputs against consistency checks, benchmarks, and historical anchors.

Primary workspace:

  • pipeline/06_tests/
  • tests/
  • notebooks/

07. Reports

Translate outputs into usable research findings and limitations.

Primary workspace:

  • pipeline/07_reports/
  • reports/
  • docs/

Practical flow

  1. define or update the world state in pipeline/00_inputs/
  2. define a scenario or research question
  3. write a hypothesis in pipeline/01_hypotheses/
  4. specify actors and decision logic
  5. formalize the interaction model
  6. assemble and run the runtime
  7. test and interpret the results
  8. summarize the run and its limitations

Why this structure fits Invest-GT

This project sits at the intersection of politics, strategic interaction, and investment behavior. That work becomes hard to reason about when assumptions, scenarios, model logic, and outcomes are mixed together. A pipeline-first structure keeps those concerns legible and versionable.