Exploratory analysis tools
This step of the framework workflow is needed to explore data features, assess overall data quality, and choose appropriate error metrics.
When forecast data is represented using the framework data structures, we can define APIs and algorithms for the following forecast exploratory analysis tools:
Prediction-realization diagram - shows actuals vs forecasts scatterplot, visualizes data features across many series and many methods
Fixed origin graph - shows point forecasts produced by different methods at some given fixed origin
Fixed horizon graph - shows point forecasts produced by different methods at different origins but with same horizon
Fanchart - visualises interval forecasts for a given series, method, and origin
The forvision package for R implements the above tools and in some cases can allow some interactivity (e.g., the use of a range selector via the dygraphs package).
To cite this website, please use the following reference:
Sai, C., Davydenko, A., & Shcherbakov, M. (date). The Forvision Project. Retrieved from https://forvis.github.io/
© 2020 Sai, C., Davydenko, A., & Shcherbakov, M. All Rights Reserved. All rights reserved. Short sections of text, not exceed two paragraphs, may be quoted without explicit permission, provided that full acknowledgement is given.