Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

I have a spatial trajectory dataset which basically contains about 10E6 (lat,lon,time)-tuples, grouped together by an identifier so they form trajectories.

Currently, SQLite is used to store the raw data.

Lat/Lon are in WGS84 but span only small areas (i.e. European country size).

I would like to store it in a geodatabase for easy descriptive & exploratory analysis, eg. calculating the overall and average distance for trajectories, and basic data preprocessing, eg. dismissing all tuples that are outside a polygon, for example.

Thus, I thought PostGIS would be optimal since I have worked with it before.

Some questions arise though: Should I just store the tuples as POINTS with GEOMETRY data type, or should I use MULTIPOINTS for whole trajectories?

And what about performance? Performance is not critical since the data is used for a research project and not for a production application, but I don't want to wait 5 hours to calculate distances between a million point-pairs.

What about indexing? What type of indices should I use and on what type of data?

I know this question is kind of general but I would like to see if people have experience with storing trajectory data and maybe they can give me some advice.

share|improve this question
1  
why multipoint when you can use lines ? –  simplexio Feb 5 '13 at 11:39
    
I don't know if LINES would be appropriate since I later need to export the data for more sophisticated modelling in R and for visualization. For those two purposes I will need single point data. But maybe LINES are stored more efficiently and still allow me to export data as points? –  wnstnsmth Feb 5 '13 at 12:15
    
if i understood what you do is that you have two points and one line of data for them, and you calculate stuff like avarage distance between points etc... so why not use two point linestring to store data –  simplexio Feb 5 '13 at 12:49

1 Answer 1

up vote 4 down vote accepted
+50

About different geometry types: From your description it looks like you should absolutely store your trajectories as linestrings. If you store them as points or multipoints you will have to build linestrings in runtime if you don't only want to do the calculations on the points defining the trajectories but also what is between the points. an Example (in meter-based projection)

ST_Distance('MULTIPOINT((10 10, 10 20))'::geometry,'POINT(0 15)'::geometry)

Will return: 11.1803398874989 which is the distance from the closest point in the multipoint to the point. while:

select ST_Distance('LINESTRING(10 10, 10 20)'::geometry,'POINT(0 15)'::geometry)

will return 10 which is the distance from the point to the closest point to the closest point on the line.

The same things about intersections test. If you only uses the points, a trajectory passing over a polygon without points in the polygon, will not be found if you only calculate on the vertex.points and don't use the points to define linestrings.

About projections you should absolutely transform (with ST_Transform) your data to a local projection if possible. It is faster and you will get a lot more functions to choose from.

About R you should be aware that postgreSQL have have a procedural language handling R directly in the database: PL/R

About indexing: The spatial index used in postgis is GIST which builds an index of the bounding boxes.

share|improve this answer
    
If appropriate, there is also a package for connecting R to Postgresql (cran.r-project.org/web/packages/RPostgreSQL/index.html) which can be a convenient way of reading/writing from the database for non-spatial analysis (as R does not handle geometries). –  djq Feb 13 '13 at 15:14
    
Well, R does actually handle spatial data, you just need to load the required packages. –  wnstnsmth Feb 15 '13 at 9:31
    
@Nicklas Avén Thanks for this exhaustive answer. I should have been a bit more clear. I do not need to process inter-trajectory calculations such as distances between trajectories, nor do I need to check whether a trajectory as a whole pass over a certain polygon, just whether some points of the trajectory are outside a polygon (to remove then for preprocessing) - pretty simple stuff actually. But still I'll accept it since it gives some interesting hints. –  wnstnsmth Feb 15 '13 at 9:34

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.