Tree Construction + Primal-Dual Routing

This repository contains a single-script runner Tree_construction_Primal_Dual.py. It supports multiple datasets and routing methods, and automatically generates:

• a plot_output/ folder
• result CSV files

Environment

Option 1: Conda

conda env create -f environment.yml
conda activate pd-routing

Option 2: pip

pip install -r requirements.txt

Inputs

Required arguments (all runs):

--dataset_input: one of
    uniform_distribution_data,
    GMM_data,
    meituan_dataset_weekday,
    meituan_dataset_weekdend_day,
    worst_instance_greedy

--routing_method: one of 
    PD_DFS, 
    PD_Greedy, 
    PD_DGreedy, 
    Greedy_baseline

Dataset-specific required arguments:

uniform_distribution_data requires:
    --st_pair_size_input, --depot_size_input, --level_size_input, --seed_value_list_input

GMM_data requires:
    --GMM_cluster_input, --GMM_sigma_input,
    --st_pair_size_input, --depot_size_input, --level_size_input, --seed_value_list_input

meituan_dataset_weekday / meituan_dataset_weekdend_day requires:
    --depot_size_input, --level_size_input

worst_instance_greedy requires:
    --st_pair_size_input

Routing_method-specific required arguments:

PD_DFS, PD_Greedy, PD_DGreedy requires: 
    --use_constant_mst: 0 (disable) or 1 (enable)
    --mst_multiplier: 0 if disabled, otherwise a positive integer (e.g., 7)

Output

After each run, the script creates:

plot_output/ 
CSV result files (saved by the script)

Examples

uniform_distribution_data + PD_DFS:

python Tree_combination_Primal_Dual.py \
--dataset_input uniform_distribution_data \
--routing_method PD_DFS \
--use_constant_mst 1 --mst_multiplier 7 \
--st_pair_size_input 1000 \
--depot_size_input 90 \
--level_size_input 3 \
--seed_value_list_input 1,2,3,4,5

GMM_data + PD_DFS:

python Tree_combination_Primal_Dual.py \
--dataset_input GMM_data \
--routing_method PD_DFS \
--use_constant_mst 0 --mst_multiplier 0 \
--st_pair_size_input 1000 \
--depot_size_input 90 \
--level_size_input 3 \
--seed_value_list_input 1,2,3,4,5 \
--GMM_cluster_input 5 \
--GMM_sigma_input 5

worst_instance_greedy + PD_DFS:

python Tree_combination_Primal_Dual.py \
--dataset_input worst_instance_greedy \
--routing_method PD_DFS \
--use_constant_mst 0 --mst_multiplier 0 \
--st_pair_size_input 1000

meituan_dataset_weekday + PD_DGreedy:

python Tree_combination_Primal_Dual.py \
--dataset_input meituan_dataset_weekday \
--routing_method PD_DGreedy \
--use_constant_mst 0 --mst_multiplier 0 \
--depot_size_input 120 \
--level_size_input 3

uniform_distribution_data + Greedy_baseline:

python Tree_combination_Primal_Dual.py \
--dataset_input uniform_distribution_data \
--routing_method Greedy_baseline \
--st_pair_size_input 1000 \
--depot_size_input 90 \
--level_size_input 3 \
--seed_value_list_input 1,2,3,4,5

Notes

Please run commands from the repository root.
The script writes outputs to plot_output/ and CSV files as implemented in the script