Combinatorial decition making and optimization(CDMO)

This repository contains models, solvers, and instance data used for the CDMO project. The full technical report is included as CDMO_report.pdf — refer to it for detailed methodology, experiments and results.

Contents:

• CP/: MiniZinc constraint programming models and runner scripts.
• MIP/: Mixed-Integer Programming implementation and helpers.
• SMT/: SMT-based solvers and the runner.py orchestrator.
• Instances/: Problem instance files (.dzn/.dat).
• res/: JSON result files for CP, MIP and SMT experiments.

Summary (from the report)

• We study the Multiple Courier Planning problem and evaluate three modelling approaches: Constraint Programming (CP), Mixed Integer Programming (MIP) and SMT encodings.
• Experiments use the instances in Instances/ and results are in res/ — see CDMO_report.pdf for performance tables and comparative analysis.

Quick Start

Requirements:

• Python dependencies: see requirements.txt.
• Docker (optional) to run inside a container.

Run locally (examples):

• MIP solver: python3 MIP/MIP.py <instance_file>
• CP solver: python3 CP/run.py <instance_file>
• SMT solvers: python3 SMT/runner.py <instance_file>
• Wrapper runs: python3 run_m.py <instance_file> <method> or python3 run_m2.py <instance_file> <method>

Docker — build and run

1. Build the image (run in repository root):

   docker build -t cdmo:latest .

2. Example run (macOS — mounts workspace Instances and res):

   docker run --rm -it
   -v /Users/negin/Desktop/CDMO_feb/Instances:/src/Instances
   -v /Users/negin/Desktop/CDMO_feb/res:/src/res
   cdmo:latest
   /venv/bin/python3 run_m2.py /src/Instances/inst01.dat MIP

3. Optional helper script run_docker.sh (place in repo root):

#!/usr/bin/env bash
# run_docker.sh
INSTANCE_FILE="$1"   # e.g. inst01.dat
METHOD="$2"          # e.g. MIP, CP, SMT
IMAGE_NAME="cdmo:latest"

docker build -t "$IMAGE_NAME" .
docker run --rm -it \
  -v "$(pwd)/Instances":/src/Instances \
  -v "$(pwd)/res":/src/res \
  "$IMAGE_NAME" /venv/bin/python3 run_m2.py "/src/Instances/$INSTANCE_FILE" "$METHOD"

Make it executable:

chmod +x run_docker.sh

Usage:

./run_docker.sh inst01.dat MIP

Notes:

• Adjust the mounted host paths or the Python/venv path inside the image if your environment differs.
• If you prefer absolute host paths, replace "$(pwd)/..." with the full path to your workspace.
• The example uses run_m2.py — replace with run_m.py or a solver-specific script as needed.

Experiments & Results

• Experimental outputs are stored under res/CP, res/MIP, and res/SMT.
• Consult CDMO_report.pdf for the experimental setup, metrics, and a discussion of findings.

Notes

• Some helper scripts and runners expect instance paths or environment-specific mounts; adjust run_docker.sh and script paths to match your local setup.
• If you want, I can add example commands for each solver using specific instances.

Contact

For questions or to reproduce experiments, open an issue or contact the repository owner.