On 30/05/07, Rui Carvalho <[log in to unmask]> wrote:
> Power-law distributions often have exponential cut-offs at the tail, which
> mean that the straight lines become curves. This is a well know phenomenon
> in real networks and one that I would expect to be present in street
> networks as physical constrains are bound to restrain growth at large
> scales.
Sorry for not being specific, I wanted to say that the whole tails
seem to be curved, much more that your plots in your paper (Carvalho
and Penn, 2004).
> One of the problems I had in the Physica A paper was that we do not know
> what a street is. There are several representations out there, but there
> is no uniquely accepted concept?
This is a issue, in fact these "power laws" (if they are power laws)
are literally created by the generalisation process. There are at
least 4 aggregation methods: axial lines, named streets, continuity
lines (angles) and road category, not mentioning a possible
combination of all of them.
Maybe it is time to have some epistemology in those physicist papers
and explain better what one is assuming before the experiments and the
effects of representation in the process.
Regards,
Lucas Figueiredo
|