Hi there Postgressors,
I have a 15-element/dimension vector (floats) data type. I also have about 10 million of these and given a query vector I would like to search these to find a number of nearest neighbours.
For this I intend to extend the current implementation of kd tree in postgresql-9.2.1/src/backend/access/spgist in file spgkdtreeproc.c. This implementation currently works on points (2D).
So till now I have implemented my user defined type in a new contrib/folder. I can install the extension and use my datatype in a table definition. So far, so good.
But I am having problems understanding the whole indexing mechanism. And how I can interface with the postgres indexing. To start with, should I be building an index from scratch or using the current GiST-based implementation?
I copied the spgkdtreeproc.c and created my own modified version (in the postgresql-9.2.1/src/backend/access/spgist) directory. How will postgres know when to use my functions instead of what is currently in place? They are named differently of course, but where is the link between the index and my functions?
Another thing which I do not get is in
Datum
spg_kd_inner_consistent(PG_FUNCTION_ARGS)
There are multiple sk_strategy defined (e.g. RTLeftStrategyNumber, RTRightStrategyNumber, RTSameStrategyNumber, etc). Who/where is this strategy being set? What does this mean? Is it just browsing through the underlying binary tree? I will have to extend this method: as I want to implement a 15-dimensional RTContainedByStrategyNumber (rather than the current BOX being used) - correct? Also where is the BOX parameter being set, by which codebase?
I have read the chapters 52-54 of the postgres documentation but I am none the wiser - if you can point me to the right documentation/tutorial I would be eternally grateful.
Many Thanks for your newbie patience,
-
Jean-Paul EbejerEarly Stage Researcher
