Fractal analysis for anomaly time series of bean related variables in Mexico

• Published in: Ciencia ergo sum Vol. 22; no. 3; pp. 233 - 241
• Main Authors: Ruiz, Olivia Delgadillo, Leos-Rodriguez, Juan Antonio, Moreno, Pedro Pablo Ramirez, Cepeda, Ricardo David Valdez
• Format: Journal Article
• Language: Spanish
• Published: 01.01.2015
• Subjects: Agricultural production; Beans; Fractals; Mexico; Time series
• ISSN: 1405-0269

Las series de anomalias de superficie cosechada, volumen de produccion, rendimiento medio anual y volumenes exportados e importados de frijol en Mexico pueden ser caracterizadas por la dimension fractal (Dv) estimada mediante la tecnica de variografia. Los resultados sugieren que las series de anomalias de variables de frijol pueden ser modeladas como series autoafines. Los valores de dimension fractal Dv [greater than] 1.5 hasta 2 expresan la presencia de variaciones de corto plazo. Los valores de Dv [greater than] 1.5 para las series de anomalias de variables de frijol sugieren un comportamiento antipersistente; es decir, estas series son impredecibles, a menos que se involucre un parametro de fractalidad en los modelos estadisticos estimados con fines de prediccion.//Mexico's anomaly time series for harvested area, production volume, annual mean yield, and exported and imported volumes of dry bean can be characterized by means of fractal dimension (Dv), estimated through variography. General results suggest those series can be properly modeled as self-affine series. Fractal dimension (Dv) values [greater than] 1.5 to 2 indicate presence of short time variation. Whereas Dv values [greater than] 1.5 for the dry bean anomaly time series suggest predominance of antipersistent variation; that is to say, all these series are unpredictable unless we involve a fractality parameter in the estimated statistical models so as to predict them. [web URL: http://cienciaergosum.uaemex.mx/index.php/ergosum/article/vi ew/635] Reprinted by permission of the Ciencia Ergo Sum Editorial Management