Exposició GRASS

Ramon Oller: Monocentricity or directional heterogeneity: a conditionally parametric approach.

Resum
In many urban spatial models the distance from a localized point plays a prominent role in the explanation of outcome variables. The basic Monocentric City Model implies a decreasing relationship between distance to the Central Business District (CBD) and population density, housing prices, land rent and income. One goal in these models is to measure the effect of distance on the output variable. Most approaches assume directional homogeneity, i.e. the distance effect does not depend on the direction in which it is being measured. Only few works have analyzed the problem of directional heterogeneity. We present a new approach to deal with the issue. It presents several advantages over existing approaches. The location and the number of areas where the distance effect varies are determined endogenously. Another advantage is that it provides a procedure to test the level of monocentricity of a spatial distribution. An application to the spatial distribution of wage income in Madrid municipality is performed to illustrate the main characteristics of the method and to show the advantages.
____________________________________________________________________________________________

12:00

David Moriña: Paquet 'survsim' de R per a la simulació de dades de supervivència.

Resum
Presentarem el paquet 'survsim' de R per a la simulació de dades de supervivència en diferents situacions, incloent supervivència clàssica, esdeveniments recurrents i esdeveniments múltiples. La principal rutina de simulació permet a l'usuari introduir un nombre arbitrari de distribucions, cadascuna corresponent a un nou esdeveniment o episodi, amb els seus paràmetres, a escollir entre Weibull, log-logística o log-normal. En tots els casos el paquet permet la introducció d’heterogeneïtat individual a través d'un efecte aleatori global per a tots els episodis o esdeveniments o bé un específic per a cadascun.