Analyzing Electricity Demand in Colombia: A Functional Time Series Approach

Abstract views: 48 / PDF downloads: 34


  • Jorge Barrientos Marin Departamento de Economía, Universidad de Antioquia, Colombia; & Facultad de Economía, Universidad Autónoma Latino Americana - UNAULA, Colombia,
  • Laura Marquez Marulanda Escuela de Ciencias Económicas y Administrativas, Universidad EIA, Colombia,
  • Fernando Villada Duque Departamento de Ingeniería Eléctrica, Universidad de Antioquia, Colombia.



functional data, functional time series, data smoothing, energy demand


In this work we are interested in analyzing the energy demand in Colombia for a short-term horizon, from a functional data approach. First, we make an exhaustive review of the literature on functional spaces as a potential source of statistical information. It is, of course, a theoretical reinterpretation since in practice the data are elements of a finite-dimensional space; however, very high-frequency data, properly treated, can be viewed as elements of a space of continuous functions. Second, we put such a reinterpretation into practice, by performing a spline-type smoothing of commercial energy demand, based on hourly-daily data. As a result, a function or smooth curve is obtained for each day. Finally, we expose the usefulness of this new approach for statistical analysis, modeling, and projection (or forecasting) of stochastic processes that generate high-frequency random variables.


Download data is not yet available.




How to Cite

Marin, J. B., Marulanda, L. M., & Duque, F. V. (2023). Analyzing Electricity Demand in Colombia: A Functional Time Series Approach. International Journal of Energy Economics and Policy, 13(1), 75–84.