Code and Experimental Data for “A Time-Quality Coordinated Matching Method for Decentralized Trading of Multi-source Homogeneous Data Assets”

Description

This dataset contains the source code and experimental materials supporting the manuscript “A Time-Quality Coordinated Matching Method for Decentralized Circulation and Trading of Multi-source Homogeneous Data Assets.” The archive includes the Solidity implementation of Time-Quality Coordinated Matching (TQCM), internal control mechanisms, external batch baselines, Hardhat tests, replay-generation scripts, processed calibration summaries, experimental outputs, statistical results, and figure-generation code. Version 2 adds scalability replays at 80, 200, 400, and 800 events across five scenarios, ten seeds, and four mechanisms, together with completion-rate and out-of-gas boundary reporting. It also adds deterministic tests of candidate-access and consecutive-fill limits, including residual cancellation, order-state transition, and locked-asset release verification. The evaluation further covers effective-price formation, waiting compensation, seller-quality regulation, price protection, partial-fill updates, quality-input robustness, multi-seed mechanism comparison, on-chain execution cost, and parameter sensitivity. Third-party raw Binance BTCUSDT and FI-2010 data are not redistributed. Processed calibration summaries, replay inputs, results, and plotting inputs are included for direct reproduction.

Steps to reproduce

1. Download and extract TQCM_release_v2.zip. 2. Install dependencies and compile the contracts: npm ci npx hardhat compile Use npm install if npm ci is unavailable. 3. Run the mechanism-level Hardhat tests listed in README.md. These verify effective prices, price protection, deterministic matching, partial fills, quality isolation, robustness, and parameter sensitivity. 4. Reproduce the 30-seed main and controlled experiments with: test/UnifiedMultiSeedReplay.test.js Included replay files cover balanced, stale_pressure, quality_competition, guard_stress, and partial_fill_pressure. Exact PowerShell and Linux/macOS commands are provided in README.md. 5. Generate the statistical results: python scripts/summarize_final_statistics.py This reproduces descriptive statistics, bootstrap confidence intervals, paired Wilcoxon tests, Holm-adjusted p-values, and effect sizes. 6. Generate manuscript Figures 3–5 and 7: python tools/plotting/four_key_figures/run_four_figures.py --project-root . --formats png svg pdf Generate their vertical variants with: python tools/plotting/four_key_figures/run_four_figures.py --project-root . --formats png svg pdf --vertical --keep-existing 7. Reproduce the scalability experiment with the included 80-, 200-, 400-, and 800-event inputs: npx hardhat test test/ScalabilityReplay.test.js --no-compile The experiment covers five scenarios, ten seeds, and four mechanisms per scale. Large runs may be executed in shards as described in README.md. Merge results and generate manuscript Figure 6: python scripts/merge_scalability_results.py python scripts/plot_scalability_results.py --layout all 8. Reproduce the bounded-execution stress tests in Windows PowerShell: $env:EXPERIMENT_BLOCK_GAS_LIMIT="120000000" $env:EXPERIMENT_TX_GAS_LIMIT="100000000" npx.cmd hardhat test .\test\BoundedExecutionStress.test.js --no-compile Output: data/bounded_execution_results/b2_bounded_execution_latest.json 9. Processed calibration summaries, replay inputs, experimental outputs, scalability results, statistical results, and plotting inputs are included in the archive. Third-party Binance BTCUSDT and FI-2010 raw data are not redistributed and are needed only to repeat preprocessing from the original sources. See README.md for source and preprocessing instructions. The EXPERIMENT_BLOCK_GAS_LIMIT and EXPERIMENT_TX_GAS_LIMIT variables adjust only the local Hardhat measurement ceiling and do not modify contract code or represent public-network gas limits.

Institutions

• China Jiliang University

  Zhejiang, Hangzhou
• Beijing Language and Culture University

  Beijing, Beijing
• Southern New Hampshire University

  New Hampshire, Manchester

Categories

Licence