Blyx

Backtest

Forecast evaluation framework · 0.1.1

Evaluate forecasts in time order with a gap for delayed observations and disjoint test windows.

Install Backtest in your editor

Extension RahulChaube.blyx-backtest · 0.1.0

The extension includes a runnable Blyx example, function completions and the framework source. For the modular project, extract the starter, open its folder and run blyxc native-run main.blyx with the native toolchain installed.

Create an empty .blyx file, type blyx-backtest-demo and accept the completion. Choose Blyx: Run File to run the example.

For an offline installation, choose Extensions: Install from VSIX for Blyx Language first, then for Backtest.

Download the companion Blyx Language VSIX

Install with Conda

Create an environment containing Backtest 0.1.1, its Blyx source and a runnable example. Run the second command with Blyx 0.2.1, Rust and a host linker installed.

conda create -y --prefix ./blyx-env --override-channels -c https://rahulchaube1.github.io/blyx-packages blyx-backtest=0.1.1=0
blyxc native-run ./blyx-env/share/blyx/backtest/main.blyx

To use this package in your application, copy share/blyx/backtest/vendor/backtest from the environment into your project's vendor directory.

Install the full collection with Homebrew or Scoop

Import into a Blyx project

Install Blyx 0.2.1, Rust 1.85 or newer and a host linker. From the directory containing your entry file, clone this tagged release.

git clone --depth 1 --branch v0.1.1 https://github.com/Rahulchaube1/blyx-backtest.git vendor/backtest

Save this as main.blyx beside vendor and run blyxc native-run main.blyx.

use vendor::backtest::backtest;

fn main() {
    let demand = [10, 20, 30, 40, 50, 60, 70, 80];
    let baselines = bt_baselines(demand, 3, 2, 1, 2);
    println!(at(baselines, 0));
    println!(at(baselines, 1));
    println!(at(baselines, 2));
    let model = bt_score(demand, [45, 55, 65, 75], 3, 2, 1);
    println!(at(model, 1));
}

Expected output

4
25
30
5

API and input requirements

Choose an initial training history, a positive horizon and a nonnegative gap. A zero training window expands the history. A positive window caps its length. The final incomplete test horizon is omitted.

APIUse
bt_fold_count(n, initial, horizon, gap), bt_fold(n, initial, horizon, gap, window, fold)Plan training and test windows with zero-based, exclusive end indices.
bt_baselines(series, initial, horizon, gap, window)Score a last-value and training-mean baseline on identical observations.
bt_score(series, forecasts, initial, horizon, gap)Calculate mean absolute and signed error of submitted forecasts.

Develop and contribute

Open the tagged source to read the implementation and examples. Include the smallest reproducible Blyx program, compiler version and expected result when reporting an issue.