Creating SpatialRDD with other attributes

[ps-spark]

Hi GeoSpark Community,

I would like to create a SpatialRDD (e.g. PointRDD) from an existing RDD[(Key, Point)] that already contains the keys, I need in the JoinQuery to identify the matching objects. Is there any way to do it?

It would be similar to the carryOtherAttributes parameter but I have an RDD not the source file.

val carryOtherAttributes = true // Carry Column 2 (hotel, gas, bar...) var objectRDD = new PointRDD(sc, pointRDDInputLocation, pointRDDOffset, pointRDDSplitter, carryOtherAttributes)

Thanks for the suggestions in advance!

Kind regards,
ps-spark

[jiayuasu]

Hi,

In each geometry of the rawSpatialRDD of SpatialRDD, there is an attribute called UserData. You can concatenate all other attributes to a string and store it there. This is how GeoSpark carries other attributes.

Thanks,
Jia

[ps-spark]

Hallo @jiayuasu,

I still don't get it. I haven't found any SpatialRDD/PointRDD constructor to create a new RDD with other attributes. Could you please post an example or even better add to the Getting started?

val myRDD: RDD[(Point, String)] = ...
val mySpatialRDD = new PointRDD(myRDD, ?????)

None of the constructors that accept a RawSpatialRDD take a parameter copyOtherAttributs.

I have actually the problem that I've got a RDD[(Point, LineString)] as a result of a DistanceJoinQuery after a CRS transformation. I cannot map the resulting points back to the original points because they are in a different CRS. So I definitely need some keys to join by.

Anyway, is there any way to convert the CRS of an RDD[(Point, LineString)]?

Thank you and best wishes,
ps-spark

[jiayuasu]

@ps-spark Each object in any rawSpatialRDD or pairRDD is a Geometry. Its other attributes are stored in its UserData attributes. You also can concatenate any strings by "tab" and store it there. Please see the highlighted line in this example: https://github.com/jiayuasu/GeoSparkTemplateProject/blob/master/geospark-analysis/src/main/scala/ScalaExample.scala#L97

[ps-spark]

Hi @jiayuasu,

Thank you very much, I followed the example and it works perfectly 👍

Kind regards,
ps-spark