Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Hi, I initiated a first wiki page to discuss specifications for that tool here : http://hub.qgis.org/wiki/quantum-gis/Spatial-query-tool http://hub.qgis.org/wiki/quantum-gis/Spatial-query-tool I first wrote an analysis of Arcgis10 equivalent tool. Feel free to contribute! Regards -- View this message in context: http://osgeo-org.1560.n6.nabble.com/Spatial-queries-plugin-should-ignore-very-small-intersections-tp5025788p5026113.html Sent from the Quantum GIS - Developer mailing list archive at Nabble.com. ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Il 06/01/2013 11:20, Andreas Neumann ha scritto: true - there is a geometry validation tool to detect these invalid geometries but no cleaning tools available. We (Faunalia, thanks to Regione Toscana support) are making a cleaning tool using liblwgeom from PostGIS, and sextante. We still have minor packaging problems (we would like to have liblwgeom in osgeo4w), but for the rest the tool should be working. All the best. -- Paolo Cavallini - Faunalia www.faunalia.eu Full contact details at www.faunalia.eu/pc Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Il 06/01/2013 13:43, Régis Haubourg ha scritto: OK for starting specifications. A good start could be a Arcgis - QGIS comparison to find missing features. I really like small graphics to help user understand what the tool is doing. I put that on my list. If you already have proposals, let's post drafts here, and then finalise in a redmine ticket or in the wiki. OK with Nathan to go to core C++ rewriting. Maybe we should ask what is the global roadmap for vector menu tools. Most of actions in the menu are part of ftools. Do we start replacing them one by one by C++ core actions, or do we have a global plan to rebuild thoses actions, and have them in sync with sextante algorithms. PSC members, do you have plans for it already? Cleaning geometry is now part of geos I remember. But it does not act like arcgis 'repair geometry'. I haven't tested it yet. It creates polygon intersections when finding auto intersections. Did anybody tried it? We (Faunalia) are making a cleaning tool using liblwgeom from PostGIS, and sextante. We still have minor packaging problems (we would like to have liblwgeom in osgeo4w), but for the rest the tool should be working. All the best. -- Paolo Cavallini - Faunalia www.faunalia.eu Full contact details at www.faunalia.eu/pc Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Paolo Cavallini wrote We (Faunalia) are making a cleaning tool using liblwgeom from PostGIS, and sextante. That's good news Paolo ! How does it work? Is it automagical like in ArcGis or more like a Grass topo build where user has to choose snapping threshold? régis -- View this message in context: http://osgeo-org.1560.n6.nabble.com/Spatial-queries-plugin-should-ignore-very-small-intersections-tp5025788p5025910.html Sent from the Quantum GIS - Developer mailing list archive at Nabble.com. ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
On Mon, Jan 07, 2013 at 01:51:10AM -0800, Régis Haubourg wrote: Paolo Cavallini wrote We (Faunalia) are making a cleaning tool using liblwgeom from PostGIS, and sextante. That's good news Paolo ! How does it work? Is it automagical like in ArcGis or more like a Grass topo build where user has to choose snapping threshold? No snapping, it simply turns invalid representations into valid ones. There's no cleaning (tolerance based) performed. --strk; ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Thanks Sandro, so that means that you get multi polygons with very small polygons created when a auto-intersection is cleaned. I prefer that since it does not destroys existing node. I think qe should add another tool after the 'makevalide' tool to clean those artefacts based on a surface threshold. This is really hard to achieve for common end user, and those artefacts can be really numerous if datasets are not clean. They also will make it harder to edit a polygon manually, and will slow down spatial intersections. Anybody else needs such a tool? -- View this message in context: http://osgeo-org.1560.n6.nabble.com/Spatial-queries-plugin-should-ignore-very-small-intersections-tp5025788p5025914.html Sent from the Quantum GIS - Developer mailing list archive at Nabble.com. ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
I made an eliminate tool (similar to ESRI's eliminate). It dissolves common boundaries of selected polygons and adjacent polygons with either largest area or largest common boundary. If you have multi polygons you would need to transform them into single polygons beforehand. The eliminate tool is part of fTools in current master. Bernhard Am 07.01.2013 11:18, schrieb Régis Haubourg: Thanks Sandro, so that means that you get multi polygons with very small polygons created when a auto-intersection is cleaned. I prefer that since it does not destroys existing node. I think qe should add another tool after the 'makevalide' tool to clean those artefacts based on a surface threshold. This is really hard to achieve for common end user, and those artefacts can be really numerous if datasets are not clean. They also will make it harder to edit a polygon manually, and will slow down spatial intersections. Anybody else needs such a tool? ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Il 07/01/2013 11:08, Sandro Santilli ha scritto: No snapping, it simply turns invalid representations into valid ones. There's no cleaning (tolerance based) performed. we can do this with another command; the two may be chained into a sextante cleanup model. -- Paolo Cavallini - Faunalia www.faunalia.eu Full contact details at www.faunalia.eu/pc Nuovi corsi QGIS e PostGIS: http://www.faunalia.it/calendario ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Hi Andreas, +1 for me. Tolerance threshold should be square units ( area) for polygon - polygons intersects, and simple units (length) for line-polygon intersections. Beyond that, I had in mind to fund a GUI and performance evolution of spatial query tool. It is not very fast, and user interface is really hard to understand for simple users. And it is the first feature I use to help newcomers to understand GIS tools capacities. It also often fails when using non shp formats, or having bad geometries. We should couple this tool with a geometry cleaning tool, raising a list of errors and automated corrections made. So means adding a clean option to 'check geometry' tool. Régis -- View this message in context: http://osgeo-org.1560.n6.nabble.com/Spatial-queries-plugin-should-ignore-very-small-intersections-tp5025788p5025809.html Sent from the Quantum GIS - Developer mailing list archive at Nabble.com. ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
Am 06.01.2013 11:09, schrieb Régis Haubourg: Hi Andreas, +1 for me. Tolerance threshold should be square units ( area) for polygon - polygons intersects, and simple units (length) for line-polygon intersections. yes Beyond that, I had in mind to fund a GUI and performance evolution of spatial query tool. It is not very fast, and user interface is really hard to understand for simple users. And it is the first feature I use to help newcomers to understand GIS tools capacities. It also often fails when using non shp formats, or having bad geometries. We should couple this tool with a geometry cleaning tool, raising a list of errors and automated corrections made. So means adding a clean option to 'check geometry' tool. true - there is a geometry validation tool to detect these invalid geometries but no cleaning tools available. So how do we proceed here. Do you want to start writing a specification for improving the spatial query plugin and the cleaning tools? In my opinion it also should be a core tool in QGIS, not a plugin. These spatial queries are really core GIS functions, not optional extensions. Andreas ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
On Sun, Jan 6, 2013 at 8:20 PM, Andreas Neumann a.neum...@carto.net wrote: In my opinion it also should be a core tool in QGIS, not a plugin. These spatial queries are really core GIS functions, not optional extensions. I agree this one of the reasons that I brought up the topic of moving code plugins to core features a couple of days ago. - Nathan ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
Re: [Qgis-developer] Spatial queries plugin should ignore very small intersections
On Sun, 6 Jan 2013 04:43:10 -0800 (PST) Régis Haubourg regis.haubo...@eau-adour-garonne.fr wrote: Maybe we should ask what is the global roadmap for vector menu tools. Most of actions in the menu are part of ftools. Do we start replacing them one by one by C++ core actions, or do we have a global plan to rebuild thoses actions, and have them in sync with sextante algorithms. AFAIK, some geoprocessing functions from fTools also available in QGIS analysis library. But they needs review and optimization. C++ implementation still slow with relatively large datasets. -- Alexander Bruy ___ Qgis-developer mailing list Qgis-developer@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/qgis-developer
