Skip to content

Classification of diurnal time series patterns. Used with HUUM for parameter adjustment.

License

Notifications You must be signed in to change notification settings

nclwater/HUUM_classification

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

README

This project contains a script which classifies a given (sub-daily) time series's diurnal pattern into four main classes. These are:

  • Two Peaks
  • Morning Peak
  • Evening Peak
  • Long Peak

Plus, strictly speaking, two more:

  • Unsuitable: Time series which are, usually due to overall daily demand value, outside plausible bounds.
  • Unclassifyable: Time series, which, while not unsuitable, cannot be classified according to the logic implemented.

It has not been uploaded to the Python Package Index, so a manual install via pip is necessary.

The documentation of the logic behind the classification can be found in the thesis Agent Based Modelling of city-wide Water Demand (not yet published).

Created by Sven Berendsen as part of his PhD project at Newcastle University, (C) 2023. Licenced under the terms of the Apache License 2.0.

ToDo

  • Introduce more classes to cover the currently unclassifyable cases.
  • Professionalise Code: Currently the code is "clean prototype" level, i.e. has little in-code documentation (but very sensible function and object names).
  • Improve the derivation of the characteristic points according to previously defined ones.
  • Remove the vestiges of matplotlib usage.

Licence

This code is originally (C) Sven Berendsen, 2023. Published under the terms of the Apache Licence, version 2.0.

About

Classification of diurnal time series patterns. Used with HUUM for parameter adjustment.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages