It is an R-package allows a reproducible research for non-parametric clustering and downscaling of daily solar irradiation time-series. The current version includes:
-
SIR_Data
Constructor function of objects ofSIRData
class. Once the user creates aSIRData
object from his data, he no longer need other inputs, ALL the rest work will done automatically. -
DPGMMclus
S3 Method for non-parametric Bayesian Dirichlet-Gaussian mixture model clustering of daily clearness index distributions. It can be also used to perform any data clustering of class matrix other than irradiance data. It generate an object of classclusData
containing the clustering outputs. -
parClusGen
Constructor function of objects ofgenData
class, needed for the generation of hight resolution solar irradiance data. -
GenData
Function to generate high resolution solar irradiance time-series. It requires object ofgenData
class as input. -
clPlot
Function to generate plots of the resulting classes.
The author is happy to answer your questions and is open to future collaborations on this topic. Please contact: [email protected] or [email protected].
Users can install the development version of SolarClusGnr
:
with either the remotes package:
remotes::install_github("frimane/SolarClusGnr")
or with the devtools package:
devtools::install_github("frimane/SolarClusGnr")
The original paper describing the methods implemented in SolarClusGnr
is:
Frimane, Â., Soubdhan, T., Bright, J.M., Aggour, M., 2019. Nonparametric bayesian-based recognition of solar irradiance conditions: Application to the generation of high temporal resolution synthetic solar irradiance data. Solar Energy 182, 462-479. URL:http://www.sciencedirect.com/science/article/pii/S0038092X19301781, doi:https://doi.org/10.1016/j.solener.2019.02.052.
The BibTex entry:
@article{Frimane2019,
title = "Nonparametric Bayesian-based recognition of solar irradiance conditions: Application to the generation of high temporal resolution synthetic solar irradiance data",
journal = "Solar Energy",
volume = "182",
pages = "462-479",
year = "2019",
issn = "0038-092X",
doi = "https://doi.org/10.1016/j.solener.2019.02.052",
url = "http://www.sciencedirect.com/science/article/pii/S0038092X19301781",
author = "{\^A}zeddine Frimane and Ted Soubdhan and Jamie M. Bright and Mohammed Aggour",
keywords = "Solar irradiance, Clustering, Clearness index, Bayesian nonparametric, Synthetic irradiance",
}
This package is free and open source software under MIT-license.