Two things come to mind (Both on the Bufferer transformer):
The buffer amount might be set too high, which will create very wide buffers.
And the End Cap Style should be set to None (no extension on the end of the roads).
Hope this helps.
Two things come to mind (Both on the Bufferer transformer):
The buffer amount might be set too high, which will create very wide buffers.
And the End Cap Style should be set to None (no extension on the end of the roads).
Hope this helps.
Attached is a template file (FME 2015) with an example.road-buffers.fmwt
If you're reprojecting to a coordinate system that uses feet, then the simply use the Bufferer, not the GeographicBufferer, which assumes your input data is in GeographicCoordinates (lat, long)
Two things come to mind (Both on the Bufferer transformer):
The buffer amount might be set too high, which will create very wide buffers.
And the End Cap Style should be set to None (no extension on the end of the roads).
Hope this helps.
So I am using FL83-WF as the coordinate system in the reprojector. Then it filters by road type and then into a geographic buffer set to 2 and feet just to see if it will work. Have endcap set to none. It still creates a round shape. i am not sure what I am doing wrong. before the buffer it looks like a street layout in a city
If you're reprojecting to a coordinate system that uses feet, then the simply use the Bufferer, not the GeographicBufferer, which assumes your input data is in GeographicCoordinates (lat, long)
I am getting same issues. I switched to a normal buffer and have it set at 2 for distance and endcap of no. everything else is the factory defaults in the tool. I have used buffers before and never had these issues. Attached are some images:
beforebuffer.jpg
after-buffer.jpg
So I am using FL83-WF as the coordinate system in the reprojector. Then it filters by road type and then into a geographic buffer set to 2 and feet just to see if it will work. Have endcap set to none. It still creates a round shape. i am not sure what I am doing wrong. before the buffer it looks like a street layout in a city
Does that Shape file contain lines or just a large amount of points?
In the case of points each point will be buffered and you will end up with cirkels.
I am getting same issues. I switched to a normal buffer and have it set at 2 for distance and endcap of no. everything else is the factory defaults in the tool. I have used buffers before and never had these issues. Attached are some images:
beforebuffer.jpg
after-buffer.jpg
That definitely looks like your buffer distance is wonky. Is the beforebuffer image before or after the reprojection? Does your original data have the projection defined? If your data was originally in lat/long, but didn't have a defined projection, then the reprojected will simply cast it into FL83-WF rather than reprojecting.
As @erik_jan points out, points that are buffered will result in circles, very short lines (many many times shorter than your buffer length) will have pretty much the same effect. Those very short lines may be caused by the reprojection if, as @jdh mentions, the source coordinate system is undefined. Can you route the data to the Data Inspector after the reprojection and have it show a background map, so you can check whether it's in the right place?
As @erik_jan points out, points that are buffered will result in circles, very short lines (many many times shorter than your buffer length) will have pretty much the same effect. Those very short lines may be caused by the reprojection if, as @jdh mentions, the source coordinate system is undefined. Can you route the data to the Data Inspector after the reprojection and have it show a background map, so you can check whether it's in the right place?
It is in a projection of lat/long. I have inspected before and after the reprojector and it seems to be right. I even looked after the filter just before the filter and it seems to be ok. Its just going through the buffer. If I join the line segments togeather some and make a larger line, do you think that this will take care of the issue? I really thought this would be a easy workbench to build and I am having more issues than I thought.
That definitely looks like your buffer distance is wonky. Is the beforebuffer image before or after the reprojection? Does your original data have the projection defined? If your data was originally in lat/long, but didn't have a defined projection, then the reprojected will simply cast it into FL83-WF rather than reprojecting.
Ya, i am not sure what to do because i want to extend the segments left and right to get a correct distance width for the roads from the centerline data.
A LineJoiner, grouped by road type and number of lanes might improve the speed of the process and the outcome.
Could you share a sample of your data, so we could have a look at the potential issues?
Ya, i am not sure what to do because i want to extend the segments left and right to get a correct distance width for the roads from the centerline data.
Could you provide a sample of the original data? A few roads would be sufficient.
A LineJoiner, grouped by road type and number of lanes might improve the speed of the process and the outcome.
Could you share a sample of your data, so we could have a look at the potential issues?
test1.zip
Here is a sample of my data. It is in a lat/lon coordinate system. I took just 300 segments and turned into a shape file. Category is what I am using the attribute filter on if that is important
test1.zip
Here is a sample of my data. It is in a lat/lon coordinate system. I took just 300 segments and turned into a shape file. Category is what I am using the attribute filter on if that is important
Hi @kouri1986
I tried this and it looks absolutely fine. Attached is a workspace (FME 2016) that creates the buffers.
esrishape2none.fmw
Then I have no clue, maybe my program is corrupt. I will reinstall and see if it works. Thank you for looking at it
@Eric_JAN. I reinstalled FME and everything seemed to work the way it was suppose to.
@Eric_JAN. I reinstalled FME and everything seemed to work the way it was suppose to.
Nice that this has been solved, but strange it was not working before!
