getting started with Simba

[geoHeil]

Hi,
do you have a getting started project for Simba?
How can I load a WKT string into Simba? Will it just work fine if I provide a Dataset[Polygon]?
And what are your maven coordinates?

git clone [email protected]:InitialDLab/Simba.git && cd Simba
git checkout standalone-2.1
sbt publishLocal

results in ~/.ivy2/local/default/simba_2.11/1.0/ivys/ivy.xml but trying to import "simba_2.11:1.0" will fail with an exception from sbt.

[dongx-psu]

Personally, I have limited experience with sbt. One way I can imagine to run this system is the following:

• Do sbt package.
• Include the built artifact in spark-submit, and also include JTS if you want to play with anything related to it (For example polygons).
• Write your program and run it on you cluster.

Another way is to modify the plugin.sbt a little bit such that the built Simba artifact will include JTS other than Spark. (More exactly sbt assembly) It should also work if you just include that artifact into your spark-submission.

Thanks for your interest on our project.

Dong

[geoHeil]

Do you have a maven example which shows basic project setup, how to load geo data into spark and then demonstrates some basic geo operations? Or should I manually parse to an Dataset[geometrty]? Dong Xie <[email protected]> schrieb am Do. 16. März 2017 um 19:37:

…

Personally, I have limited experience with sbt. One way I can imagine to run this system is the following: - Do sbt package. - Include the built artifact in spark-submit, and also include JTS if you want to play with anything related to it (For example polygons). - Write your program and run it on you cluster. Another way is to modify the plugin.sbt a little bit such that the built Simba artifact will include JTS other than Spark. It should also work if you just include that artifact into your spark-submission. Thanks for your interest on our project. Dong — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#84 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABnc9LJY5hzmberYi4G4_D0sF8HgLmfQks5rmYF7gaJpZM4Meqrj> .

[dongx-psu]

The examples are in the example package. Currently, the example shows the old way which relies on Double fields.

The current version should also work if you construct a dataset from our provided spatial classes. I will add some examples recently about that.

And we do have some effort to import data directly from GeoJSON, ShapeFile and OSM formatted file. I think we will include that soon enough after we testest the performance and stability for standalone-2.1 branch.

[dongx-psu]

New examples added for spatial class inference.

I will try to make a maven example later on.

[geoHeil]

That would be great. Thanks. Dong Xie <[email protected]> schrieb am Do. 16. März 2017 um 21:30:

…

New examples added for spatial class inference. I will try to make a maven example later on. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#84 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABnc9OPe6Sn7U_hC8YLpO7-Z6CAkq-Dbks5rmZvBgaJpZM4Meqrj> .

[geoHeil]

Please can you help me find the correct maven coordinates for simba. Georg Heiler <[email protected]> schrieb am Do. 16. März 2017 um 22:27:

…

That would be great. Thanks. Dong Xie ***@***.***> schrieb am Do. 16. März 2017 um 21:30: New examples added for spatial class inference. I will try to make a maven example later on. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#84 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABnc9OPe6Sn7U_hC8YLpO7-Z6CAkq-Dbks5rmZvBgaJpZM4Meqrj> .

[dongx-psu]

We haven't published it to any maven repo. The best way to use it is to include the built artifact in your project folder and add it to your built dependencies in the maven file. Do make sure you also include the Simba artifact in spark-submit as well.

[geoHeil]

I know that. After cloning the repo and switching to the spark 2 branch I executed sbt publishLocal. this claims to publish as: ivy2/local/default/simba_2.11/1.0/ivys/ivy.xml

Why trying to include it like "simba" %% "1.0" to the dependencies this will break the resolution of sbt.

@Skyprophet
when adding
organization := "InitialDLab" to your build.sbt then I can Import it via "InitialDLab" %% "simba" % "1.0",

#86

[dongx-psu]

Will adding organization simply solves the problem?

[geoHeil]

yes. That's why I created the pull request.