Closed daren-thomas closed 5 years ago
First thing I noticed: Creating the topography geometry takes forever.
First fix: using ArcScene -> Data Management Tools -> Raster -> Raster Processing -> Resample, I resampled the terrain.tif to 20m x 20m granularity and exported the raster dataset back to TIFF format. Reading in the topography is now faster.
Next problem: The geometry in zone.shp
(and probably also in district.shp
has Z coordinates. This breaks the geometry_generator.py
script.
Fixed in branch 1625-figure-out-why-the-radiation-daysim-tool-hangs-on-sabines-scenario
: POLYGON Z shapefiles can now be used (the Z component is ignored)
Next problem: zone.shp
and terrain.tif
do not intersect. The coordinate systems are different, but I believe that the routines we use should be able to abstract that away. Showing both layers in ArcGIS Desktop is impossible, it seems they're too far appart or something.
Next problem:
zone.shp
andterrain.tif
do not intersect. The coordinate systems are different, but I believe that the routines we use should be able to abstract that away. Showing both layers in ArcGIS Desktop is impossible, it seems they're too far appart or something.
@daren-thomas I encountered this issue too... but weirdly it's solved if you add layers in a certain order: if you add the terrain first, then the building, they appear to be on the same location. If you add the buildings first, the coordinate system seems to be interpreted differently and the buildings and topography can't be shown together.
ugh!
You're right! But check this out:
It seems even if it worked, there are still some buildings that are "outside" of the considered terrain...
I just checked to make sure the building our intersection is failing on is not one of those (it isn't):
Also, that's the very first building it tries... So... something is up with the coordinate system thing...
@daren-thomas you might wanna check the terrain file I sent Sabine in reply to her email today. Seemed to work better and there were no buildings outside the terrain.
Am 20.09.2018 um 16:47 schrieb Daren Thomas notifications@github.com<mailto:notifications@github.com>:
You're right! But check this out:
[image]https://user-images.githubusercontent.com/2969564/45826274-7d98ad00-bcf4-11e8-9e3c-f3f9010390a5.png
It seems even if it worked, there are still some buildings that are "outside" of the considered terrain...
I just checked to make sure the building our intersection is failing on is not one of those (it isn't):
[image]https://user-images.githubusercontent.com/2969564/45826427-c486a280-bcf4-11e8-99b2-79652aff92d4.png
— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/architecture-building-systems/CityEnergyAnalyst/issues/1625#issuecomment-423211353, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ARsFYb1PfBSuQDhoqjHBDjnSOTSrvXoPks5uc6p9gaJpZM4WyBwE.
Next thing I tried was importing the Coordinate system from zone.shp using this guide: http://desktop.arcgis.com/en/arcmap/10.3/manage-data/raster-and-images/defining-or-modifying-a-raster-coordinate-system.htm
Still the same error. Somehow the centroid points of the building footprints don't intersect with the geometry of the terrain...
Yep, same issue here. As I mentioned above, things didn't look the same on ArcGIS when I imported the shapefile first than when I opened the terrain. The reason is that ArcGIS was setting a different coordinate system for the scene in each case. So I changed the coordinate system in zone.shp and arrived at the same result you did...
@martin-mosteiro was just trying to use the reduced.rad scenario you sent me and the following problem occurrs when trying to do the base heights thing in ArcScene:
@daren-thomas Strange. Works perfectly for me... ...iff I add the terrain first and then the zone. Otherwise I still get no error message but the zone seems to disappear from the scene.
Ok, thanks guys. I just talked to Sabine and we’ll correspond here and update the files from the Polybox folder you sent Martin:
Here’s the case study after I removed the Z coordinate in the shapefile and replaced the terrain (I suppose this is the same case study you were testing): https://polybox.ethz.ch/index.php/apps/files/?dir=/2018.09.21.15.01_flattened_geometry&fileid=1099896304 Here’s the case study with Danielle’s new shapefile: https://polybox.ethz.ch/index.php/apps/files/?dir=/2018.09.21.15.01_Danielles_geometry
Thanks for your help!
I opened the 2D shapefile with the terrain Martin prepared and it seems to be ok (in QGIS).
I will now try it in CEA
I had to use swissALTI3D2016relief since the create_new_project tool would not accept the terrain.tif from Martin.
Everything runs fine until the radiation tool... here's the error:
C:\Users\iA\Anaconda2\envs\cea\python.exe C:/Users/iA/Documents/GitHub/CityEnergyAnalyst/cea/resources/radiation_daysim/radiation_main.py
DEBUG:Fiona:GDAL_DATA: C:\Program Files (x86)\PostgreSQL\10\gdal-data
DEBUG:Fiona:PROJ_LIB: C:\Users\iA\Anaconda2\envs\cea\lib\site-packages\osgeo\data\proj
DEBUG:Fiona:GDAL_DATA: C:\Program Files (x86)\PostgreSQL\10\gdal-data
DEBUG:Fiona:PROJ_LIB: C:\Users\iA\Anaconda2\envs\cea\lib\site-packages\osgeo\data\proj
DEBUG:Fiona:Userencoding: None
DEBUG:Fiona:Projection not found (cogr_crs was NULL)
DEBUG:Fiona:Projection not found (cogr_crs was NULL)
DEBUG:Fiona:Index: 0
DEBUG:fiona.collection:Flushed buffer
DEBUG:fiona.collection:Stopped session
creating 3D geometry and surfaces
DEBUG:Fiona:GDAL_DATA: C:\Program Files (x86)\PostgreSQL\10\gdal-data
DEBUG:Fiona:PROJ_LIB: C:\Users\iA\Anaconda2\envs\cea\lib\site-packages\osgeo\data\proj
DEBUG:Fiona:Userencoding: None
DEBUG:Fiona:Got coordinate system
INFO:Fiona:Failed to auto identify EPSG: 7
DEBUG:Fiona:Params: +proj=somerc +lat_0=46.95240555555556 +lon_0=7.439583333333333 +k_0=1 +x_0=600000 +y_0=200000 +ellps=bessel +units=m +no_defs
DEBUG:Fiona:Got coordinate system
DEBUG:Fiona:Index: 0
DEBUG:fiona.collection:Flushed buffer
DEBUG:fiona.collection:Stopped session
DEBUG:Fiona:GDAL_DATA: C:\Program Files (x86)\PostgreSQL\10\gdal-data
DEBUG:Fiona:PROJ_LIB: C:\Users\iA\Anaconda2\envs\cea\lib\site-packages\osgeo\data\proj
DEBUG:Fiona:Userencoding: None
DEBUG:Fiona:Got coordinate system
INFO:Fiona:Failed to auto identify EPSG: 7
DEBUG:Fiona:Params: +proj=somerc +lat_0=46.95240555555556 +lon_0=7.439583333333333 +k_0=1 +x_0=600000 +y_0=200000 +ellps=bessel +units=m +no_defs
DEBUG:Fiona:Got coordinate system
DEBUG:Fiona:Index: 0
DEBUG:fiona.collection:Flushed buffer
DEBUG:fiona.collection:Stopped session
DEBUG:Fiona:GDAL_DATA: C:\Program Files (x86)\PostgreSQL\10\gdal-data
DEBUG:Fiona:PROJ_LIB: C:\Users\iA\Anaconda2\envs\cea\lib\site-packages\osgeo\data\proj
DEBUG:Fiona:Userencoding: None
DEBUG:Fiona:Projection not found (cogr_crs was NULL)
DEBUG:Fiona:Projection not found (cogr_crs was NULL)
DEBUG:Fiona:Index: 0
DEBUG:fiona.collection:Flushed buffer
DEBUG:fiona.collection:Stopped session
Failed to boolean cut
Failed to boolean cut
[...]
Failed to boolean cut
Traceback (most recent call last):
File "C:/Users/iA/Documents/GitHub/CityEnergyAnalyst/cea/resources/radiation_daysim/radiation_main.py", line 231, in
Process finished with exit code 1
Where does the error lie?
PS: I cannot access the polybox... it says I don't have the permission
@sabinepython Sorry, here are the correct links
@sabinepython do you have new files? the ones martin links to don't seem to work properly for me. i get a totally different errror...
@daren-thomas same here, I tried Martin's files on Friday and it didn't work out neither. I'm now working on a new reduced sample with 10 buildings in the middle of the district. https://www.dropbox.com/sh/yja9zlseciawc2c/AABkLMW8aXzD9u2bJb9wvweOa?dl=0
Concerning the re-projection, my shapefiles are already in CH1903/LV03 ...
Or is CH1903/LV03 the root of the problem and I should use another one? Could you recommend any PCS for Switzerland I could use and fulfills CEA's requirements?
Oh it seems to work better with CH1903+/LV95 :)
@daren-thomas So now I changed the terrain created by Martin and the district & zone shapefiles to CH1903+/LV95, they overlap nicely in QGIS :) Using these files I'm getting the error below when running the create_new_file:
`Running create-new-project with occupancy-types = C:\Users\iA\Dropbox\SemesterProject_SP\CEA\test_sample10/inputs/building-properties/occupancy.dbf
Warning 1: Cannot find ellipsoid.csv
Running create-new-project with zone = C:\Users\iA\Dropbox\SemesterProject_SP\CEA\test_sample10/inputs/building-geometry/zone.shp
Running create-new-project with terrain = C:\Users\iA\Dropbox\SemesterProject_SP\CEA\test_sample10/inputs/topography/terrain.tif
Running create-new-project with terrain = C:\Users\iA\Dropbox\SemesterProject_SP\CEA\test_sample10/inputs/topography/terrain.tif
Running create-new-project with output-path = C:\Users\iA\Dropbox\SemesterProject_SP
Warning 1: Cannot find prime_meridian.csv
Traceback (most recent call last):
File "C:/Users/iA/Documents/GitHub/CityEnergyAnalyst/cea/datamanagement/create_new_project.py", line 141, in
Process finished with exit code 1`
any ideas of where the bug lies?
@sabinepython can you make your terrain and shape files available?
They're in the link I shared before terrain: https://www.dropbox.com/s/lez7bv9d5my6fcm/terrain.tif?dl=0 shapefiles: https://www.dropbox.com/sh/ryg7hw34pio7xvq/AABhDS39QivlsjKN9kgKBJXAa?dl=0
@sabinepython based on your scenario as posted here: https://www.dropbox.com/sh/yja9zlseciawc2c/AABkLMW8aXzD9u2bJb9wvweOa?dl=0
I just managed to run the radiation simulation. I did encounter a problem with building Z1947 - It has 0 floors above ground and 0 height above ground and the CEA was not able to handle that. I just removed it and the simulation ran through. I did not get the error you had above, though. Instead, I got an error failed lofting
.
What you have above seems to be an installation error. I can help you with installing no problem!
@daren-thomas Could you come over for the installation issue this afternoon or even now? If after that there are still some issues we can discuss them tomorrow after Integrated Design.
@daren-thomas could you please share your modified files (without Z1947)? thanks
Sabine Python's semester project involves a rather big scenario that seems to hang with the radiation-daysim tool. Why? I don't know.
The scenario can be obtained from Martin, Danielle, Sabine or Daren.