Re: [Qgis-user] Planetary (Venus) coordinate reference systems

[Gabriel De Luca via QGIS-User]

Hi,

  I think that it is the main canvas trying to reproject all layers to a
common (but not related to any Earth's) CRS.  As if the datum
transformations between these systems were not defined, and the absence of
the WGS-84 pivot was insurmountable.

  In my opinion, the most convenient thing to do is to disable the
"on-the-fly" projection of the main map. In the project properties, CRS
tab, checking the option: "No CRS (or unknown/non-Earth projection)".

  In this way, the canvas becomes a kind of "universal-unitless 2D
Cartesian system". All the geometries of the layers will be displayed in
the coordinates that are stored in their vertices, and the rasters
according to their matrices of geotransform. In fact, it can contribute to
a better understanding of coordinate reference systems.

  The only things that are disabled are on-the-fly reprojection on the main
map and maybe ellipsoidal measurements. All other functionalities must be
available.


Regards,
Gabriel

El vie, 10 nov 2023 a la(s) 18:49, Diaz, Tony A. via QGIS-User (
qgis-user@lists.osgeo.org) escribió:

> I’m trying to help a student understand projections and trying to load in
> a Venus global map data. When they try to align different maps with the
> same projection, or even change projection, a consistent error message  "No
> transform available between  and Custom CRS” keeps occurring.
>
> Any tips or ideas would be helpful,
>
>
>
> Tony
>
> Caltech
>
> Pasadena, CA
>
>
>
>
> ___
> QGIS-User mailing list
> QGIS-User@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
___
QGIS-User mailing list
QGIS-User@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] python code generated during georeferencing in QGIS

[Gabriel De Luca via QGIS-User]

Hi,

  QGIS georeferencing is made in C++, so i think you can't get its workflow
in the Python console.
  You can see the code of the georeferencer in:
https://github.com/qgis/QGIS/tree/master/src/app/georeferencer

Regards,
Gabriel

El sáb, 20 may 2023 a la(s) 11:53, Mike Flannigan via QGIS-User (
qgis-user@lists.osgeo.org) escribió:

>
> Theoretically QGIS is open source code.  I have not proved that
> myself, but that is what I hear.  If that is the case, Yes,
> you can see the QGIS code used for georeferencing - if you can
> find it :-)
>
>
> Mike
>
>
> On 5/16/23 14:00, qgis-user-requ...@lists.osgeo.org wrote:
> > Message: 2
> > Date: Tue, 16 May 2023 12:55:48 +0200
> > From: Spaska Forteva
> > To:qgis-user@lists.osgeo.org
> > Subject: [Qgis-user] python code generated during georeferencing in
> >   QGIS
> > Message-ID:
> >   
> > Content-Type: text/plain; charset=UTF-8; format=flowed
> >
> > Hello,
> >
> > I have a problem and I hope for an answer here. I know how to
> > georeference a tif map with geopoints in QGIS. I have developed a shiny
> > app and want to automate/program the georeferencing for several maps
> > there (for loop). Unfortunately, I've tried a lot and can't get it
> > directly under python without errors. My question, is there any way I
> > can see the python code generated during georeferencing in QGIS? I can
> > develop something myself in the Python console, but not the steps that I
> > create with the QGIS program for georeferencing, right?
> > Is there a way to see the python code generated during georeferencing in
> > QGIS? -GPT says yes, but nothing happens to me in the console 
> >
> > Best Regards,
> > Spaska Forteva
>
> ___
> QGIS-User mailing list
> QGIS-User@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
___
QGIS-User mailing list
QGIS-User@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [QGIS-Developer] Call for co-maintainers of the QGIS hub

[Gabriel De Luca via QGIS-User]

Hi Tim,

  I could help keep an eye on new submissions and test them on my machine.
  My osgeo username is gabrieldeluca. Count on me if you think I can be of
help.

Regards,
Gabriel

El mar, 4 abr 2023 a la(s) 12:40, Tim Sutton via QGIS-Developer (
qgis-develo...@lists.osgeo.org) escribió:

> Dear QGIS enthusiasts
>
> As some of you may be aware, we have been developing a space for users to
> share their QML styles, processing models, 3D objects, etc.
>
> You can find already a rich collection of great resources here:
>
> https://plugins.qgis.org/ (open the 'Hub' menu up top).
>
> It would be really great to have co-reviewers help me to deal with the new
> submissions as they come in. If you are interested to help me, please let
> me know and I will give you the access needed (I'll need your OSGEO id).
>
> Thanks!
>
> Regards
>
> Tim
>
> --
>
> --
>
> Tim Sutton
> Kartoza Co-Founder
> Visit http://kartoza.com to find out about open source:
>  * Desktop GIS programming services
>  * Geospatial web development
> * GIS Training
> * Consulting Services
> Tim is a member of the QGIS Project Steering Committee
>
> ---
> ___
> QGIS-Developer mailing list
> qgis-develo...@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
>
___
QGIS-User mailing list
QGIS-User@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] [EXTERNAL] Re: Geopackage slow on NAS if not read-only

[Gabriel De Luca via QGIS-User]

Hi Árni,

You can try setting the SQLITE_USE_OGR_VFS=YES configuration option, as
mentioned at https://github.com/qgis/QGIS/issues/27899.
Be sure to read the implications of changing the default (NO) option:
https://gdal.org/drivers/vector/sqlite.html#configuration-options
You can change the config option in a environment variable or a config
file: https://gdal.org/user/configoptions.html#configuration-options

Regards,
Gabriel


El jue, 16 mar 2023 a la(s) 11:08, Árni Geirsson via QGIS-User (
qgis-user@lists.osgeo.org) escribió:

> I do appreciate the good advice to use DBMS, I am doing that too and
> understand the advantages. I would never consider using a geopackage if I
> thought that multiple people would want to edit them at the same time. But
> that is not the case and that is not my problem. Sometimes, I just want to
> keep geodata in a file on a network share and run the risk (in my case
> insignificant) that the file will be corrupted by multi user access. This I
> can do with shapefiles, but unlike the shapefiles, I have to set the
> read-only flag on the geopackages stored on the network to get decent
> rendering speed, without ever wanting to edit the files. I think wanting to
> use file based storage for geodata in a networked environment is not
> unreasonable, even if there is some danger of corruption. All file based
> formats are probably subject to that risk, but I repeat, that issue is not
> what I am talking about, just the rendering speed of data in a geopackage
> that is not set to read-only.
> I suspect that there is no solution and that I will just have to live with
> this - as I have done for years. It is just a bit annoying. And I am
> surprised how little it is discussed.
>
> Árni
>
>
> On Thu, 16 Mar 2023 at 13:46, Sadowski Jarosław via QGIS-User <
> qgis-user@lists.osgeo.org> wrote:
>
>> +1
>>
>>
>>
>> Don’t use file formats to edit by multiple users. I was occurring some
>> critical problems, as for example disappearing objects from database or
>> LONG read for first time ☹
>>
>>
>>
>> Use PostgreSQL or other solutions like:
>>
>> Mergin Maps: Collect, Store and Analyze your Geo-Data Easily
>> 
>>
>> GIS Support » GIS.Box (gis--support-pl.translate.goog)
>> 
>>
>>
>>
>>
>> *_*
>>
>> *Jarosław Sadowski*
>> Kierownik Zespołu ds. Ochrony Środowiska | *Biuro Strategii i
>> Planowania, Projektowania i Inżynierii Podprogramu Kolejowego*
>> *Environmental Protection Team Leader **| Railway Subprogramme Strategy
>> & Planning, Design & Engineering Department*
>> e: jaroslaw.sadow...@cpk.pl
>> m: +48 532 720 230
>>
>> *From:* QGIS-User  *On Behalf Of *Bo
>> Victor Thomsen via QGIS-User
>> *Sent:* Thursday, March 16, 2023 2:35 PM
>> *To:* qgis-user@lists.osgeo.org
>> *Subject:* Re: [Qgis-user] [EXTERNAL] Re: Geopackage slow on NAS if not
>> read-only
>>
>>
>>
>> UWAGA: Ta wiadomość pochodzi spoza CPK Sp. z o.o. Proszę pomóż zadbać o
>> bezpieczeństwo naszej organizacji. Zastanów się, zanim otworzysz link lub
>> załącznik.
>>
>>
>>
>> When any user on your network access the geopackage layer and (maybe
>> unnecessarily) puts the layer in edit mode, there will be created 2 extra
>> files in the same directory as the gpkg file resides in. And QGIS behaves
>> different regarding *both* read and write operations when these file are
>> present. This is probably the explanation of the longer reading times, even
>> if nobody actually is editing the layer. When the user stops the editing
>> mode for the layer, the 2 files disappear.
>>
>> You can check if this is the explanation:
>>
>>1. Open the gpkg layer in read mode on one computer, check the access
>>speed of the same layer on *another* computer.
>>2. Set the layer in edit mode on the first computer, check the access
>>speed to the layer on the second computer
>>3. Revert the edit mode on the first computer, check the access speed
>>to the layer on the second computer.
>>
>> I am guessing, that situations 1 and 3 are fast, while situation 2 is
>> slow.
>>
>> First of all: Never, ever try to implement some kind of multi-user
>> editing on a file based format, where the file resides on a networked
>> drive. It will never, ever work reliably. At some point 2 users will try to
>> edit the same layer at the same time and it will go kaboom (I can read from
>> your mails that you are aware about this). This goes for - probably - every
>> file based format on a network drive.
>>
>> Secondly: What about making the gpkg file read-only at the network share
>> level for most of the users. And only granting write access to the gpkg
>> file for certain users that are instructed in *not* setting the layer in
>> edit mode unless it's strictly necessary ?
>>
>> The best solution: Install Postgres/PostGIS on your NAS
>>
>> Med venlig hilsen / Best regards
>>
>>
>>
>> Bo 

Re: [Qgis-user] What QGIS plugin development should I work on this year?

[Gabriel De Luca via QGIS-User]

Case 1) would be welcome.

I do a little manually that transformation. I convert the values of
latitude, longitude and ellipsoidal height to geocentric Cartesian
coordinates. Then I find the transformation parameters between the local
system and the geocentric one.

The transformation pipeline is then made up of: a 3D affine transformation
from the local to the geocentric system, the conversion from the geocentric
to the geodetic or geographic system, and finally the projection to some
cartographic system (usually a custom one, but can be projected to any).

The geoscience plugin, on the other hand, is based on the construction of a
derived from a projected reference system. In other words, it calculates
the affine transformation parameters in 2D between a local system and a
projected one and uses them in the construction of a new derived system. The
derived from projected CRS is only compatible with the WKT2:2019 version,
and that limits it in some GDAL and QGIS functions.

I collaborated with the author in the construction of the WKT for derived
from projected CRSes and the calculation of parameters is done with a
script that I had written when I began to need to carry out this type of
transformation. It only requires numpy and is very efficient at finding
transformation parameters. The algorithm on which it is based is not mine.

You can use the script if you consider it convenient, I published it with a
MIT license at https://github.com/gabriel-de-luca/simil.

Regards,
Gabriel De Luca

El jue, 26 ene 2023 a la(s) 14:08, Greg Troxel via QGIS-User (
qgis-user@lists.osgeo.org) escribió:

> Great place to ask IMHO.
>
> These is a combination of a suggestion and a fish for an answer that
> tells me no work is needed.  I don't know how widespread the audience is
> relative to your standards.
>
> 1)
>
> Consider someone with a local coordinate system establish with a total
> station or similar classical methods.  No assumptions on orientation of
> axes, except up is up.  Let's assume there are perhaps hundreds of
> points.  Perhaps this is just a data collector dump and perhaps it is
> gama output.
>
> Further assume the person has absolute positions (lat/lon/HAE) probably
> via GPS RTK, and perhaps standard deviations.
>
> I want to be able to enter the correspondence, and compute a
> transformation that can be not only used in qgis but stored in a file to
> be used with proj/gdal.  And, I want to store the corresponding points
> so that I can re-open, change some, and add some, without starting over.
>
> There is a geoscience plugin that mostly does this but it seemed awkward
> for reasons I don't remember this minute.
>
>
> 2)
>
> Similar, but assume someone has a survey plan with angles and distances,
> and somehow (in a new plugin or separate) converts them to a local
> system (because one can ~never assume the angles are true to any
> particular geodetic coordinate system).  Given 1, this is about entering
> the angles/distances and turning it into a layer, I think.
>
> There is a cadastral plugin but it seems to be about taking a shape and
> producing documents, vs taking documents and producing a shape.
>
>
> 3)
>
> With RTK, one can take 30s, 60s, whatever position traces on a mark.
> Then one can average those.  Multiple such observations over multiple
> days can then be combined to form an estimate for the mark.  There are
> interesting questions here about finding the mark by being close, vs
> labeling, and not labeling is a faster workflow.   I'm doing this
> manually but it would be cool to be more automated.
>
>
> ___
> QGIS-User mailing list
> QGIS-User@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
___
QGIS-User mailing list
QGIS-User@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Geo Raster Background

[Gabriel De Luca]

 Hi, I usually do a brief adjustment on pixel values prior to
georeferencing, to make sure that no pixel has the value I need to assign
to nodata.

Depending on the source dataset, I usually try to replace the zeros with
ones, and reserve the zero for nodata.

El vie, 1 de oct. de 2021 a la(s) 15:34, L.W. (eagl...@gmx.de) escribió:

> Hi,
>
> nope this isn't what I want ... it is the same as setting transparent
> color manually.
>
> I am using PDF and TIF for georeferencing.
>
> Regards
>
> Without no-data-0 or transparent-color:
>
> Set transparent-color to 0,0,0 or use no-data-0:
>
>
> Am 01.10.2021 um 19:26 schrieb Charles Dixon-Paver:
>
> I just loaded an unreferenced png with black content into qgis, then set
> the nodata value to 9, then did a fake georeference with arbitrary
> coordinates, and the result seemed to be rotated without any background
> pixels. Using 3.20 on Windows. The geoereferencer has a checkbox for
> setting nodata to 0, so uncheck that first. Otherwise, if you're still
> unable to get the "triangles" to change, my previous messages should still
> apply.
>
> On Fri, 1 Oct 2021 at 18:27, L.W.  wrote:
>
>> yes, maybe to orange, a color that is not in the original image, so I can
>> do this color transparent.
>>
>>
>> Am 01.10.2021 um 18:25 schrieb David Strip:
>>
>> When georeferencing, aren't the areas outside the original image, eg, the
>> "triangles" resulting from rotation, treated as no data? Wouldn't this
>> issue be resolved by setting the no data value to something other than
>> black?
>> Just guessing here, obviously.
>>
>>
>> On 10/1/2021 2:36 AM, L.W. wrote:
>>
>> Thanks for answering, but I think this is not the way I want to go ...
>>
>> Or, I do not have written detailed.
>>
>> I have a raster image e.g. a white rectangle, after setting the points,
>> QGIS rotats this image, let's say, by 45 degrees.
>>
>> The corners of the imported image are now black triangles.
>>
>> I do want to have e.g. red triangles ...
>>
>> Regards
>>
>>
>> Am 01.10.2021 um 09:50 schrieb Charles Dixon-Paver:
>>
>> There's not really any way for a GIS system to automatically
>> differentiate between black background pixels and black data pixels. QGIS
>> supports users setting custom pixel values to render as transparent pixels,
>> but it's left up to users to ensure that their data and background values
>> are different.
>>
>> From my view, there are a couple of options when it comes to fixing this
>> issue in your data. One is to use an image processing tool like GIMP or
>> Photoshop effectively, but that requires retaining the geographic
>> information and knowing how to reprocess the raster accordingly.
>>
>> The simplest solution I'm aware of is probably to follow this workflow in
>> QGIS:
>>
>>- Start a new QGIS project
>>- Set all black pixels to transparent in the transparency tab of the
>>raster layer properties
>>- Digitize a "bounding area" - you can do this with a memory layer
>>- Move your vector layer to render underneath the raster
>>- Give the vector a simple black symbology
>>- Use the "Convert map to raster" tool from the processing toolbox
>>
>> There are a wide number of variations on that workflow that are possible
>> of course, but I think that's the easiest to get started.
>>
>>
>> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Georeferencing a DXF file

[Gabriel De Luca]

Hi Walt,

  There are some options to georeference a vector layer:

  Since it is 2D, you can georeference it to a projected coordinate system
with ogr2ogr command line application, in the same way that a raster layer
can be georeferenced from control points. You can use any transformation
method and differences will be compensated if the number of control points
provided is higher than the minimum requirement for that method.
  Vector Bender plugin can georeference the layer in QGIS, but if you
define more control points than strictly necessary the transformation will
not be linear also not compensated, but triangulated into as many linear
transformations as control points more than three.
  If you have the coordinates in both systems (the local cartesian of the
original drawing and the geographic but projected one) you can align the
drawing in AutoCAD and export to dxf, import in QGIS and just assign the
proper CRS. Also you can translate, rotate and scale in QGIS with
geoprocessing algorithms or with QAD plugin.

  If you need to georeference to geographic coordinates instead of a
projected system, you can use a linear but 3D transformation to geocentric
and then a conversion to geographic. It's probably not what you want to do. But
you also shouldn't use a 2D linear transformation to go from the drawing
system to a geographic one.

Regards,
Gabriel.

On Wed, Feb 3, 2021 at 4:13 PM Walt Ludwick  wrote:

> I've received a 3D AutoCAD file from an architect that i managed to
> convert into 2D .dxf format and import into my QGIS project, but it is
> out-of-scale and out-of-place, so i guess i need to add some georeferencing
> coordinates to the file.
>
> This should be easy enough, since it's an accurate representation of a
> building that's already in the project -with an additional wing tacked on-
> so i'm thinking i could just use a few corners of that existing buildings
> as ground control points, and map them to the corresponding corners on this
> .dxf model, then maybe all other co-ordinates could be automatically
> generated by some tool, if i could just find the right one for the job.
>
> Is there in fact any such easy way to do this thing?
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] error loading tiff layer

[Gabriel De Luca]

El vie., 31 de ene. de 2020 a la(s) 08:10, Humberto Nel Florez Ramos (
hnflorezra...@gmail.com) escribió:

> regards.
> I could not load a layer in tiff format. The error that qgis shows is:
>
> "No transform is available between Unknown CRS: ENGCRS [" PCS Name =
> unnamed ", EDATUM [" "], CS [Cartesia ... and Unknown CRS: GEOGCRS ["
> unknown ", DATUM [" unknown ", ELLIPSOID [" WGS 8 ... .
> No coordinate operations are available between these two reference systems
> ".
>
> The file has a reference system, which is magna-sirgas from Bogota (3116),
> I have tried the magna-sirgas (3116) project, and wgs84 (4326) showing the
> same error.
>
>
Humberto,

ENGCRS is a WKT2 Engineering CRS keyword. So the file seems to be a GeoTIFF
with a Engineering CRS.

While EPSG:3116 Magna SIRGAS is a projected CRS.

Engineering (or Topocentric) CRS may be converted to a Geocentric CRS as
described in the Section 2.2.2 of the Geomatics Guidance Note 7, part 2
Coordinate Conversions & Transformations including Formulas [1].

I think that QGIS doesn't have support for this kind of CRS, but I think
that PROJ does. So maybe you can find a pipeline string for the conversion
from the Engineering CRS to the Projected CRS with projinfo, and then
transform the GeoTIFF file with gdalwarp program (-ct parameter for the
pipeline string).

Once you have the GeoTIFF file reprojected to SIRGAS, you will have no
problem loading it into QGIS.

If you want to copy the WKT2 string, we can try to find the pipeline string
to convert it to EPSG:3116.


Kind Regards.
Gabriel


[1] Page 99 of http://www.epsg.org/Portals/0/373-07-02.pdf
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Micro-Geometry Problems

[Gabriel De Luca]

About how to remedy these situations, you can use the v.clean GRASS
provider algorithm, selecting the snap cleaning tool.

Kind Regards,

Gabriel

El mié., 15 de ene. de 2020 a la(s) 16:10, Gabriel De Luca (
pablogabrieldel...@gmail.com) escribió:

> When digitizing, coordinates are taken from the canvas and stored to the
> layer. If the canvas (project) CRS is not the same as the layer CRS, a
> transformation and/or conversion must be done between the taken and stored
> coordinates. That transformation and/or conversion performed between the
> previous vertex to be rendered, and the rendered coordinate digitized to be
> stored, may lead a round trip error.
>
> If you are digitizing from a canvas CRS to the same layer CRS, and
> snapping to vertex, I think that you will not see any difference between
> both vertices.
>
> If you are snapping to intersections, you can see a difference also
> working in the same CRS for the canvas and the layer, because an
> intersection calculated coordinate may not be a rational number and at some
> decimal place it must be truncated.
>
> I don't know if any rendering option could truncate the coordinate of a
> vertex when rendering it, so when snapping to that vertex (in the same CRS
> for the canvas and the layer) could return a truncated coordinate for the
> digitized vertex.
>
> El mar., 14 de ene. de 2020 a la(s) 18:59, 
> escribió:
>
>> All-
>>
>>
>>
>> I have digitized a land use/land cover layer as individual polygons.  The
>> snap options were used and, for most vertices, the snap was precise.  A
>> vertex in polygon A has a precise match to a vertex in neighboring polygon
>> B.
>>
>>
>>
>> But in 582 out of 2799 polygons, there are mis-matches in adjoining
>> vertices that are only evident if you zoom into the layer at a
>> magnification (i.e., a scale) of from 2:1 to 4 to 1.  I have spent
>> a couple of weeks trying to remedy these by hand using the Vertex Tool and
>> other means.  My question is this: is there an automated way in QIS to
>> detect and remedy these situations?
>>
>>
>>
>> Thanks,
>>
>> Steve Miler
>> ___
>> Qgis-user mailing list
>> Qgis-user@lists.osgeo.org
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user

Re: [Qgis-user] Micro-Geometry Problems

[Gabriel De Luca]

When digitizing, coordinates are taken from the canvas and stored to the
layer. If the canvas (project) CRS is not the same as the layer CRS, a
transformation and/or conversion must be done between the taken and stored
coordinates. That transformation and/or conversion performed between the
previous vertex to be rendered, and the rendered coordinate digitized to be
stored, may lead a round trip error.

If you are digitizing from a canvas CRS to the same layer CRS, and snapping
to vertex, I think that you will not see any difference between both
vertices.

If you are snapping to intersections, you can see a difference also working
in the same CRS for the canvas and the layer, because an intersection
calculated coordinate may not be a rational number and at some decimal
place it must be truncated.

I don't know if any rendering option could truncate the coordinate of a
vertex when rendering it, so when snapping to that vertex (in the same CRS
for the canvas and the layer) could return a truncated coordinate for the
digitized vertex.

El mar., 14 de ene. de 2020 a la(s) 18:59,  escribió:

> All-
>
>
>
> I have digitized a land use/land cover layer as individual polygons.  The
> snap options were used and, for most vertices, the snap was precise.  A
> vertex in polygon A has a precise match to a vertex in neighboring polygon
> B.
>
>
>
> But in 582 out of 2799 polygons, there are mis-matches in adjoining
> vertices that are only evident if you zoom into the layer at a
> magnification (i.e., a scale) of from 2:1 to 4 to 1.  I have spent
> a couple of weeks trying to remedy these by hand using the Vertex Tool and
> other means.  My question is this: is there an automated way in QIS to
> detect and remedy these situations?
>
>
>
> Thanks,
>
> Steve Miler
> ___
> Qgis-user mailing list
> Qgis-user@lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
___
Qgis-user mailing list
Qgis-user@lists.osgeo.org
List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user