You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We see that the clipping for waterways has more or less happened (seen in the bottom NoData Cells). However the lake body is not clipped correctly. Driving logic is here
Likely problem is that filtering for waterway is just not sufficient here. Also zooming around the rendered map we see that not a rivers are well clipped suggesting that not even all rivers are captured.
The goal is not total coverage for every tag but good visual improvement to illustrate capability.
I'd suggest testing it on California subset with sparkSubmitJobWithMain ingest.ToblerPyramid --orc s3://geotrellis-test/dg-srtm/california.orc --layer ca-tobler --tiny --render --partitions 1584 using this PR #8
To run this in decent amount of time its sufficient to downgrade 100 machines to 20 in the build.sbt.
A somewhat outdate docker container is available with that branch at: docker run -it --rm -p 8000:8000 -p 4040:4040 -v ~/.aws:/home/hadoop/.aws quay.io/geodocker/jupyter-geopyspark:tobler-demo
The text was updated successfully, but these errors were encountered:
Looking at current render output:
We see that the clipping for waterways has more or less happened (seen in the bottom NoData Cells). However the lake body is not clipped correctly. Driving logic is here
Likely problem is that filtering for
waterway
is just not sufficient here. Also zooming around the rendered map we see that not a rivers are well clipped suggesting that not even all rivers are captured.This should be changed to capture lakes and as many rivers as possible, starting by inspecting the OSM wiki for water: http://wiki.openstreetmap.org/wiki/Key:water
The goal is not total coverage for every tag but good visual improvement to illustrate capability.
I'd suggest testing it on California subset with sparkSubmitJobWithMain ingest.ToblerPyramid --orc s3://geotrellis-test/dg-srtm/california.orc --layer ca-tobler --tiny --render --partitions 1584 using this PR #8
To run this in decent amount of time its sufficient to downgrade 100 machines to 20 in the build.sbt.
To explore the data I'd suggesting using GeoPySpark and browse around the OSM features in dataframe using this notebook as starting point: https://gist.github.com/echeipesh/445a0d5e397964e55a2350f654e0309a
This notebook requires:
feature/vector-pipe
branch.A somewhat outdate docker container is available with that branch at:
docker run -it --rm -p 8000:8000 -p 4040:4040 -v ~/.aws:/home/hadoop/.aws quay.io/geodocker/jupyter-geopyspark:tobler-demo
The text was updated successfully, but these errors were encountered: