Skip to main content

Forecasting Model package based on naive models

Project description

Forecast_x: Toolkit with Naive models for time series.

Forecast_x logo

PyPI version license Downloads Downloads Downloads

**If you're interested in financially supporting my open source, consider visiting this link. Your support helps tremendously with sustainability this work.

Forecast_x is a pure python package that provides different naive models for fitting multiple time series, especially in batch process, due to its powerful flexibility and easy usage.

This library can be used in several industries with focus on manufacturing processes, where forecasting models with low cost of error are needed to plan raw material consumption.

Models

Forecast_x uses the following models to produce forecast:

  • Model Naive
  • Model Seas Naive
  • Model Mean Two Periods
  • Model Mean Three Periods
  • Model Half Seas Mean
  • Model Seas Period Mean
  • Model Double Seas Mean
  • Model Seas Growth
  • Model Expo Weighted
  • Model Threefith Mean
  • Model Multi Seas Mean
  • Model Seas Double Mean Growth
  • Model Grand Mean
  • Model Smooth Grand Mean
  • Model Last Seas Mean
  • Model Current Mean Seas
  • Model Smooth Double Seas Naive
  • Model Truncated Mean
  • Model Harmonic Mean
  • Model Heronian Mean

Getting started: 10 seconds to Forecast_x

Here is how-to use Forecast_x models:

from forecast_x import forecast_x as fx

# time series observation
time_series = [51, 17, 28, 37, 52, 21, 34, 47, 38, 35, 7, 27]
freq = 12 # monthly
h = 12 # forecast months ahead

# Creating the forecast object
f = fx.forecast(time_series, freq, h)

# Applying any the model from the package
model = f.model_naive()

# The model variable would produce a list of three elements:
# - Fitting Values
# - Error
# - Forecast
model

To get only forecast of a given model you should use:

f.get_forecast('model_naive')

To allow the package to select the best fit based on multiple cross validation you should use:

model = f.best_model()
# forecast_x would select 'model_seas_period_mean' as best model based on test results
model
# Getting forecast from best model
forecast = get_forecast(model)

Installation

# or PyPI
pip install forecast_x

Dependencies

  • None.

Python Version

Supported on 3.5, 3.6 and 3.7.

License

MIT

Documentation

The official documentation will be available soon.

Citation

Citations or acknowledge on any work or project are very welcome:

Alejandro De Barros. 2018. Forecast_x: An open source forecasting tool for time series library for Python

Meta

Alejandro De Barros – (https://twitter.com/alejandrodbn) – alejandrodbn@gmail.com

Distributed under the MIT license. See LICENSE for more information.

https://github.com/alejandrodbn/forecast

Code of Conduct

Everyone interacting with this project's codebases, issue trackers, and mailing lists is expected to follow the Code of Conduct.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

forecast_x-0.12.20.tar.gz (6.6 kB view hashes)

Uploaded Source

Built Distribution

forecast_x-0.12.20-py3-none-any.whl (18.4 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page