forecaster

go-forecaster

go-forecaster is a Go library designed for time-series forecasting. It enables users to model and predict data with strong seasonal components, events, holidays, change points, and trends. The library offers functionalities for fitting forecast models, making future predictions, and visualizing results using Apache ECharts.​

Features

• Model Fitting: Fit time-series data to capture trends, seasonal patterns, and events.
• Prediction: Generate forecasts for future time points based on fitted models.
• Visualization: Create interactive line charts to visualize actual data alongside forecasts and confidence intervals.
• Changepoint Detection: Identify points where the time-series data exhibits abrupt changes in trend or seasonality.
• Event Components: Incorporate and analyze the impact of specific events on the time-series data.

Examples

Contains a daily seasonal component with 2 anomalous behaviors along with 2 registered change points

Contains a daily seasonal component with a trend change point which resets

Auto-changepoint detection and fit

Installation

To install go-forecaster, use go get:

go get github.com/aouyang1/go-forecaster

Usage

Here's a basic example demonstrating how to use go-forecaster:

package main

import (
    "fmt"
    "time"

    "github.com/aouyang1/go-forecaster"
)

func main() {
    // Sample time-series data
    times := []time.Time{...} // Your time data here
    values := []float64{...}  // Corresponding values

    // Initialize the forecaster with default options
    f, err := forecaster.New(nil)
    if err != nil {
        fmt.Println("Error initializing forecaster:", err)
        return
    }

    // Fit the model to the data
    err = f.Fit(times, values)
    if err != nil {
        fmt.Println("Error fitting model:", err)
        return
    }

    // Generate future time points for prediction
    futureTimes, err := f.MakeFuturePeriods(10, 24*time.Hour)
    if err != nil {
        fmt.Println("Error generating future periods:", err)
        return
    }

    // Predict future values
    results, err := f.Predict(futureTimes)
    if err != nil {
        fmt.Println("Error making predictions:", err)
        return
    }

    // Output the predictions
    for i, t := range futureTimes {
        fmt.Printf("Date: %s, Forecast: %.2f\n", t.Format("2006-01-02"), results.Forecast[i])
    }
}

This example initializes a forecaster, fits it to sample data, and predicts future values. For more detailed examples, refer to the examples directory in the repository.

Visualization

go-forecaster integrates with Apache ECharts to provide interactive visualizations of your forecasts. After fitting a model, you can generate an HTML visualization:

// Assuming 'f' is your fitted Forecaster instance
err := f.PlotFit(outputWriter, nil)
if err != nil {
    fmt.Println("Error generating plot:", err)
}

This will create an HTML file displaying the original data, the forecast, and confidence intervals.

Documentation

Comprehensive documentation is available on pkg.go.dev, detailing all available functions and types.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contributing

Contributions are welcome! If you encounter issues or have suggestions for improvements, please open an issue or submit a pull request.