Thanks Howard, I misunderstood the pagesize parameter. It's working fine with default pagesize. I tested with my road data (~1million) and the response is quite good.
Regards, Gagan On Tue, Nov 24, 2009 at 7:48 PM, Howard Butler <[email protected]> wrote: > > On Nov 24, 2009, at 5:26 AM, Gagan Bansal wrote: > > > Hi, > > > > I was trying to use the 'pagesize' property as pagesize=1 while creating > a index > > >>>idx = index.Index('road' , pagesize=1) > > and then I insert all my bbox to the index (idx) > > >>>del idx > > > > now next time I want to read the index from file (road.idx,road.dat) > > >>>idx = index.Index('road' , pagesize=1) > > or >>>idx = index.Index('road' ) > > but I get error while reading i.e. > > rtree.core.RTreeError: Error in "Index_Create": Spatial Index Error: > InvalidPageException: Unknown page id 1 > > > > What could be the reason for this error? > > I think this error has been fixed in trunk (requires trunk libspatialindex > at this time too, unfortunately). > > > > My aim is to load the full index in single page for efficiency. > > Remember that the pagesize is a value in bytes. A pagesize=1 will never > contain the entire index. Also, ensure that your pagesize is large enough > to hold any pickled data that you might be storing in the index as well. We > have changed the pagesize to default to 4096, which is fairly reasonable. > Any larger will waste disk space, and any smaller will mean creating lots > of pages. > > Howard > _______________________________________________ > Community mailing list > [email protected] > http://lists.gispython.org/mailman/listinfo/community >
_______________________________________________ Community mailing list [email protected] http://lists.gispython.org/mailman/listinfo/community
