Related Plugins and Tags

QGIS Planet

MapScaping podcast: GRASS GIS probably doesn’t get the attention it deserves


In episode 86 of MapScaping podcast, Markus Neteler talks about the functionality offered (the topological vector engine, 2D and 3D raster analysis, image processing and available programming interfaces), and whom GRASS GIS is for.

Markus is the cofounder of mundialis, a Bonn based business. He is also the Chairman of the GRASS GIS Project Steering Committee.

Enjoy the podcast and the transcript!

The post MapScaping podcast: GRASS GIS probably doesn’t get the attention it deserves appeared first on GFOSS Blog | GRASS GIS and OSGeo News.

QField 1.7 Rockies hits the stage

Be ready for the cold weather with a smooth coordinate search, filters in the value relation widget, fancy new QML and HTML widgets, enhanced geometry editing functionalities and an expandable legend. Right when Autumn starts, QField 1.7 Rockies hits the stage.

As usual get it now on the play store or on github!

The days are getting shorter and the wind blows colder. It’s always good to be in a good company outside while getting your mapping work done. QField will be your reliable companion.

We know, QField 1.6 Qinling has only been out two months and with its amount of new features and stability improvements, it would have deserved a longer primetime. But we just couldn’t withhold you all the new great stuff we’ve been building lately.

So let’s welcome QField 1.7 Rockies. And yes, we mean THE Rockies, where QField is looking for plenty of new buddies.

Let’s have a look.

Merging features

Splitting of a feature has been possible for quite some time. Now the merging of features of multipolygon-layer is possible as well. Select them and merge them – easy like that. The first selected feature gets the new geometry and keeps its attributes.

Filters in the Value Relation Widget

The value relation widgets provide an easy selection of a related feature. Often it’s used for lookup tables but sometimes the related tables contain a lot of entries and the list of the possible values is long.

Using filters in the value relation drop-down can increase the efficiency in selecting the correct value. It can be configured by expressions in QGIS, so it’s possible to have the content of the drop down depend on the values entered previously in other fields.

In the screenshot above there is a Map Value Widget with “forest” and “meadow” as values. On selecting “forest”, only the trees appear in the Field “Plant Species”. On selecting “meadow” there would be listed flowers instead.

Go to coordinates in the Search

The search has not only been improved in its appearance, but it’s handling is much more comfortable with a button to clear the text and easy opening and closing.

Additionally, we added the possibility to jump to coordinates. Searching a place you know the coordinates of is now super simple. And this means that digitizing that precise geometry with known coordinates is finally possible.


QML and HTML Widget

You might remember when we introduced the QML widget in QGIS. Now it’s in QField as well. And it’s not alone. HTML widgets are supported too.

This provides a lot of possibilities to display information with texts, images and charts and it even allows you interaction.
Do you need help setting up complex forms? Don’t hesitate to get in touch with us!

Expandable legend icons

The legend items are now expandable and collapsible.

Wait a minute… Wasn’t this possible before? Yes. It was possible in earlier versions. But why it’s announced here as a new feature?

Because now it is built in a future proof manner thanks to all the people and organisations who care for QField and bought a support contract with the sustainability initiative or committed to a recurring sponsorship.

Some technical background: As you may be aware QField uses QGIS under the hood and QGIS uses Qt under the hood. Qt is currently used in version 5. Qt 5 is not that young any more and has a lot of functionality which is no longer supported by Qt. The old legend was based on the tree view, a deprecated module. Using it had some implications like the suboptimal support of HiDPI. Furthermore, these deprecated modules will disappear in the soon-to-come Qt 6.

As you can see, keeping QField at the quality we and you expect requires a lot of maintenance work. It is of utmost importance and only possible thanks to sponsoring since paying for fixing already existing features is less attractive for most people.

What will the future bring

In the last weeks, we have been highly busy on coding, testing and promoting QFieldCloud and we are very happy to be able to announce it very soon. So be prepared.

Also, keep an eye on the @QFieldForQgis and @QFieldCloud twitter accounts to stay updated.

Open Source

QField is an open source project. This means that whatever is produced is available free of charge. To anyone. Forever. This also means that everyone has the chance to contribute. You can write code, but you don’t need to. You can also help translating the app to your language or help out writing documentation or case studies or by sponsoring a new feature.

And now…

… enjoy QField 1.7 Rockies and have a nice autumn!

GRASS GIS 7.8.4 released

1. What’s new in a nutshell

Processed LiDAR data in GRASS GIS (screenshot by @neteler)As a follow-up to the previous GRASS GIS 7.8.3 we have published the new release GRASS GIS 7.8.4 with more than 170 improvements. This minor release again focuses on wxGUI fixes, especially in the animation export, the layer management, 3D visualization and the data catalogue. Many display modules received fixes as well, and the vector digitizer now works as expected.

The overview of new features in the 7.8 release series is available at new features in GRASS GIS 7.8. See also our detailed announcement with the full list of changes and bugs fixed at

Binaries/Installer download:

Source code download:

First time users may explore the first steps tutorial after installation.


The Geographic Resources Analysis Support System (, commonly referred to as GRASS GIS, is an Open Source Geographic Information System providing powerful raster, vector and geospatial processing capabilities. It can be used either as a stand-alone application or as backend for other software packages such as QGIS and R geostatistics or in the cloud. It is distributed freely under the terms of the GNU General Public License (GPL). GRASS GIS is a founding member of the Open Source Geospatial Foundation (OSGeo).

The GRASS Development Team, Oct 2020

The post GRASS GIS 7.8.4 released appeared first on GFOSS Blog | GRASS GIS and OSGeo News.

Store and visualize your raster in the Cloud with COG and QGIS

We have recently been working for the French Space Agency ( CNES ) who needed to store and visualize satellite rasters in a cloud platform. They want to access the image raw data, with no transformation, in order to fullfill deep analysis like instrument calibration. Using classic cartographic server standard like WMS or TMS is not an option because those services transform datasets in already rendered tiles.

We chose to use a quite recent format managed by GDAL, the COG (Cloud Optimize Geotiff) and target OVH cloud platform for it provides OpenStack, a open source cloud computing platform.

How it works

A COG file is a GEOTiff file which inner structure is tiled, meaning that the whole picture is divided in fixed size tile (256 x 256 pixels for instance) so you can efficiently retrieve parts of the raster. In addition to the HTTP/1.1 standard feature range request, it is possible to get specific tiles of an image through the network without downloading the entire raster.

We used a service provided by OpenStack, called Object Storage to serve the COG imagery. Object storage allows to store and retrieve file as objects using HTTP GET/POST requests.

Why not WCS ?

Web Coverage Service standard could have been an option. A WCS server can serve raw data according to a given geographic extent. It’s completely possible to deploy a container or a VPS (Virtual Private Server) running a WCS Server in a cloud plateform. The main advantages of the COG solution over WCS Server is that you don’t have to deal with the burden of deploying a server, like giving it ressources, configuring load balancing, handle updates, etc…

The beauty of COG solution is its simplicity. It is only HTTP requests, and everything else (rendering for instance) is done on the client side.

Step by step

Here are the different steps you’d have to go through if you’re willing to navigate in a big raster image directly from the cloud.

First, let’s generate a COG file

gdal_translate inputfile.tif cogfile.tif -co TILED=YES -co COPY_SRC_OVERVIEWS=YES -co COMPRESS=DEFLATE

Install your openstack-client, it can be achieved easily with Python pip install command line

$ pip install python-openstackclient

Next, configure your openstack client in order to generate an athentification token. To do so you need to download your project specific openrc file to setup your environment)

$ source
Please enter your OpenStack Password for project myproject as user myuser:
$ openstack token issue                                 
| Field      | Value                                                                                                                                                                                   |
| expires    | 2020-07-21T08:15:12+0000                                                                                                                                                                |
| id         | xxxx_my_token_xxxx
| project_id | 97e2e750f1904b41b76f80a50dabde0a                                                                                                                                                        |
| user_id    | 18f7ccaf1a2d4344a4e35f0d84eb065e                                                                                                                                                        |

You are now good to push you COG file to the cloud instance

openstack object create MyContainer cogfile.tif --name cogfile.tif

Before starting QGIS, 2 environment variables need to be set.  (replace xxxx_my_token_xxxx with the token you’d just come to generate)

$ export SWIFT_AUTH_TOKEN=xxxx_my_token_xxxx

It can also be done directly from the QGIS Python console by setting those variable using the os.environ.

Finally, add a cloud raster data source in in QGIS

You can now navigating into your image directly reading it from the cloud

© CNES 2018, Distribution Airbus DS


While panning in the map, QGIS will download only few tiles from the image in order to cover the view extent. The display latency that you could see in the video depends essentially on:

  • The number of band of your image
  • The pixel size
  • Your internet connection (mine, the one use for the video, is not an awesome one)

Note that the white flickering that you could see when you move in the map and the raster is refreshed should be removed in next version of QGIS according to this QEP.

What’s next ?

Thanks so much to the GDAL and QGIS contributors for adding such a nice feature ! It brings lots of possibilities for organizations that have to deal with great number of big raster and just want to explore part of it.

We are already thinking about further improvments (ease authentification, better integration with processing…), so if you’re willing to fund them or just want to know more about QGIS, feel free to contact us at [email protected]. And please have a look at our support offering for QGIS.

Sol Katz Award – Thank you

On Thursday, I was awarded the 2020 Sol Katz award for Free and Open Source Software for Geospatial. I feel very honored to have been selected for this award and I’d like to take this opportunity to share a few words of thanks:

As people working in open source projects, we are constantly reminded that we are all standing on the shoulders of giants. However, particularly this year, we also see just how important small personal connections are. For me, my involvement with open source communities really took off when I joined the QGIS hackfest in Vienna in 2009 and I felt that my participation was really appreciated and welcome. I couldn’t imagine being without these connections anymore.

Thank you to the whole QGIS community, particularly my fellow PSC members both current and former: Tim, Andreas, Jürgen, Richard, Paolo, Otto, Marco Hugentobler, Alessandro, our new chair Marco Bernasocchi, and of course QGIS founder Gary Sherman for starting this awesome project and for still being around and actively promoting geospatial open source by publishing so many great books covering multiple different OSGeo projects.

I’d also like to thank my partner and my family for being incredibly understanding whenever I’m spend my time geeking out over a new programming project, data analysis, forum question, or conference talk.

Thank you also to my friends, colleagues and fellow members of the larger OSGeo community for sharing ideas, providing valuable feedback, and spreading the word about all the great work that’s going on all around us.

I’m constantly amazed by all the innovation happening to nourish and grow our community. And I’m looking forward to continue being a part of these efforts.

Thank you!


Publication de l’extension COVADIS RAPEA pour QWAT et QGEP

QWAT est une application open source de gestion des réseaux d’eau potable émanant des collectivités de Pully, le SIGE à Vevey, Morges et Lausanne.
QGEP est son homologue dédiée à la gestion des eaux usées et pluviales, initiée par le groupe utilisateur QGIS Suisse.

L’échange de données entre institutions est une pierre angulaire des politiques de l’eau. Ces échanges se basent sur des formats d’échanges standardisés. Ainsi les Cantons de Fribourg (format aquaFRI) ou de Vaud (format SIRE) conditionnent certaines subventions publiques à la transmission des données selon des formats pré-définis et permettent à ces échelons administratifs d’avoir une vision globale des réseaux humides.

Dans le cadre d’une expérimentation des outils QWAT (eau potable) et QGEP (eaux usées), Charentes Eaux a souhaité mettre en œuvre des extensions dédiées au standard d’échange de données sur les réseaux d’eau Français, le Géostandard Réseaux d’adduction d’eau potable et d’assainissement (RAEPA) défini par la Commission de validation des données pour l’information spatialisée (COVADIS).

Oslandia a été mandaté pour mettre en œuvre des instances de QWAT et QGEP, réaliser les extensions RAEPA pour chacun de ces outils, et aider Charente Eaux à charger les données des collectivités membres de ce syndicat mixte.

Le travail a été publié pour QWAT sous forme d’une extension standardisée dans le dépôt l’organisation QWAT

Pour QGEP, il n’existe pas encore de fonctionnalité pour gérer d’extension, le dépôt contient donc les définitions de données et de vues à rajouter manuellement au modèle de données.

La compatibilité des modèles de données a été évaluée et le choix a été fait de ne faire que des vues dédiées à l’export de données. Il est techniquement possible de faire des vues éditables pour permettre le chargement de données via ces vues depuis des fichiers suivant le gabarit de données RAEPA. Le niveau de simplification et d’agrégation des listes de valeurs rend ce travail peu générique dans l’état actuel du géostandard (v1.1), il est donc plus pertinent à ce stade de réaliser des scripts de chargement sans passer par ce pivot dans le cas de Charente-Eaux

Extracting trajectory-based flows between M³ prototypes

Rendering large sets of trajectory lines gets messy fast. Different aggregation approaches have been developed to address this issue. However, most approaches, such as mobility graphs or generalized flow maps, cannot handle large input datasets. Building on M³ prototypes, the following approach can be used in distributed computing environments to extracts flows from large datasets. 

This is part 3 of “Exploring massive movement datasets”.

This flow extraction is based on a two-step process, conceptually similar to Andrienko flow maps: first, we extract M³ prototypes from the movement data. In the second step, we determine flows between these prototypes, including information about: distribution of travel speeds and number of observed transitions. The resulting flows can be visualized, for example, to explore the popularity of different paths of movement:

After the prototypes have been computed, the flow algorithm computes transitions between pairs of prototypes. An object moving from prototype A to prototype B triggers an update of the corresponding flow. To allow for distributed processing, each node in the distributed computing environment needs a copy of the previously computed prototypes. Additionally, the raw movement data records need to be converted into trajectories. Afterwards, each trajectory is processed independently, going through its records in chronological order:

  1. Find the best matching prototype for the current record
  2. Ensure that the distance to the match is below the distance threshold and that the matched prototype is different from the previous prototype
  3. Get or create the flow between the two prototypes
  4. Ensure that the prototype and flow directions are a good match for the current record’s direction
  5. Update the flow properties: travel speed and number of transitions, as well as the previous prototype reference

This approach scales to large datasets since only the prototypes, the (intermediate) flow results, and the trajectory currently being worked on have to be kept in memory for each iteration. However, this algorithm does not allow for continuous updates. Flows would have to be recomputed (at least locally) whenever prototypes changed. Therefore, the algorithm does not support exploration of continuous data streams. However, it can be used to explore large historical datasets:

Flow example: passenger vessel speed patterns showing mean flow speeds (line color: darker colors equal higher speeds) and speed variation (line width)

If you want to dive deeper, here’s the full paper:

[1] Graser, A., Widhalm, P., & Dragaschnig, M. (2020). Extracting Patterns from Large Movement Datasets. GI_Forum – Journal of Geographic Information Science, 1-2020, 153-163. doi:10.1553/giscience2020_01_s153.

This post is part of a series. Read more about movement data in GIS.

QField 1.6 is out!

Editing multiple features at the same time, support for stylus pens, dynamic configuration of image names and much more.
QField 1.6 Qinling 秦岭 comes packed with awesome new features and an improved user experience.

We have been very busy over the last few months working on a new and shiny QField release. We have added many new features that increase efficiency on the field or allow for new workflows. In parallel, we have also been working on ironing out a series of issues and improving the overall user experience to make the app as pleasurable to use as possible. The result is QField 1.6 which has been published now.

Enough of the highlevel talking, let’s see what has been done.

Multi editing

Do you recall Geography lesson 101, Toblers first law? Everything is related to everything else. But near things are more related than distant things.

Very often there are similar objects nearby which share a property, tree species tend to group, human created objects like street light types or street paint markings tend to be of the same type at the same location.

With QField 1.6 it is now much easier to select a couple of features and change an attribute with very few taps. Identify a feature, long press an identify results, select more features and click the edit attributes button.

Stylus support

Sometimes it is just too cold to be working with fingers (although of course you can get capacitive gloves too). Or you just prefer to be working with a pen. QField 1.6 comes with support for stylus pens. If your device ships with one, give it a try.

Lock geometries

For some scenarios, especially in asset management, you only need to change attributes of existing objects and never add new features, delete features or change geometries. This can be configured through QFieldSync and set in the layer properties.

Image name configuration

Did you ever want to have the file names of your pictures to match the feature id, the layer name or any free text? The expression based configuration in QFieldSync offers now complete freedom in naming your images.

Legend and UX and legacy code

Didn’t expect to read UX and legacy code in one single title?

QML is the technology on which the QField user interface is built. QML ships a lot of user interface elements in a library called “Quick Controls”. A long time ago already it received an update from version 1 to version 2. Up to recently we still have been using some elements from version 1, which had an effect on high resolution displays not being able to properly display everything. To workaround that we introduced a lot of band aids, to improve the situation. We are very happy, that by migrating the legend and few other remaining elements to Quick Controls 2 in version 1.6, we have been able to completely drop this code.

Topological editing

QGIS can detect shared boundary by the features, so you only have to move a common vertex once, and QGIS will take care of updating the neighboring ones. So does his little college QField since this release.

Fast editing mode

For the real adventurers who know what they are doing this release brings the fast editing mode. In this mode, the features will automatically be stored on every change. The user interface is lighter and it combines perfectly with the topological editing.

Unter the hood

We have brought the whole technology stack up to speed with modern requirements. Proj and GDAL have been updated to recent versions. This helped to mitigate a couple of issues with coordinate transformations that were completely misplaced. It also paves the path for a future with datum corrections and always more important high precision measurements.

Known Issues

Unfortunately, we are experiencing a crash on startup with 32 bit devices. These devices are not that common any more, but if you have a device that is already a couple of years old it’s very well possible that it comes with a 32 bit cpu builtin. Despite the team’s hard efforts to isolate the reason, we were not able to find out what it was yet. Because of this we will not be able to update to 1.6 for these devices at the moment. We still hope that we will find a solution for this but don’t know yet when this will be.

We have updated proj to version 6. This brings plenty of bug fixes with coordinate handling. Among other things it adds support for using datum grids (gsb files) for very precise transformations, it is not yet possible to install those on the device. You will get an information message in the about dialog if your project happens to fall into this category. In this case, as a workaround switch the CRS of the project to a CRS with a known conversion that works without grid files.

What will the future bring

You guessed it already, we are not tired and have plenty of things stacked for the future. Prepare for more exciting updates for attribute forms and also for QFieldCloud which is right now being tested in our R&D labs.

Also keep an eye on the @QFieldForQgis Twitter account to stay updated.

Open Source

QField is an open source project. This means that whatever is produced is available free of charge. To anyone. Forever. This also means that everyone has the chance to contribute. You can write code, but you don’t need to. You can also help translating the app to your language or help out writing documentation or case studies or by sponsoring a new feature.

Thanks to sponsors

Various organisations have helped to make this new release become a reality. Without the support of people in organisations who believe in the future of QField and open source tool for geospatial in general. The whole team behind QField would like to thank you with a big applause!

Generating trajectories from massive movement datasets

To explore travel patterns like origin-destination relationships, we need to identify individual trips with their start/end locations and trajectories between them. Extracting these trajectories from large datasets can be challenging, particularly if the records of individual moving objects don’t fit into memory anymore and if the spatial and temporal extent varies widely (as is the case with ship data, where individual vessel journeys can take weeks while crossing multiple oceans). 

This is part 2 of “Exploring massive movement datasets”.

Roughly speaking, trip trajectories can be generated by first connecting consecutive records into continuous tracks and then splitting them at stops. This general approach applies to many different movement datasets. However, the processing details (e.g. stop detection parameters) and preprocessing steps (e.g. removing outliers) vary depending on input dataset characteristics.

For example, in our paper [1], we extracted vessel journeys from AIS data which meant that we also had to account for observation gaps when ships leave the observable (usually coastal) areas. In the accompanying 10-minute talk, I went through a 4-step trajectory exploration workflow for assessing our dataset’s potential for travel time prediction:

Click to watch the recorded talk

Like the M³ prototype computation presented in part 1, our trajectory aggregation approach is implemented in Spark. The challenges are both the massive amounts of trajectory data and the fact that operations only produce correct results if applied to a complete and chronologically sorted set of location records.This is challenging because Spark core libraries (version 2.4.5 at the time) are mostly geared towards dealing with unsorted data. This means that, when using high-level Spark core functionality incorrectly, an aggregator needs to collect and sort the entire track in the main memory of a single processing node. Consequently, when dealing with large datasets, out-of-memory errors are frequently encountered.

To solve this challenge, our implementation is based on the Secondary Sort pattern and on Spark’s aggregator concept. Secondary Sort takes care to first group records by a key (e.g. the moving object id), and only in the second step, when iterating over the records of a group, the records are sorted (e.g. chronologically). The resulting iterator can be used by an aggregator that implements the logic required to build trajectories based on gaps and stops detected in the dataset.

If you want to dive deeper, here’s the full paper:

[1] Graser, A., Dragaschnig, M., Widhalm, P., Koller, H., & Brändle, N. (2020). Exploratory Trajectory Analysis for Massive Historical AIS Datasets. In: 21st IEEE International Conference on Mobile Data Management (MDM) 2020. doi:10.1109/MDM48529.2020.00059

This post is part of a series. Read more about movement data in GIS.

Movement data in GIS #31: exploring massive movement datasets

Exploring large movement datasets is hard because visualizations of movement data quickly get cluttered and hard to interpret. Therefore, we need to aggregate the data. Density maps are commonly used since they are readily available and quick to compute but they provide only very limited insight. In contrast, meaningful aggregations that can help discover patterns are computationally expensive and therefore slow to generate.

This post serves as a starting point for a series of new approaches to exploring massive movement data. This series will summarize parts of my PhD research and – for those of you who are interested in more details – there will be links to the relevant papers.

Starting with the raw location records, we use different forms of aggregation to learn more about what information a movement dataset contains:

  1. Summarizing movement using prototypes by aggregating raw location records using our flexible M³ Massive Movement Model [1]
  2. Generating trajectories by connecting consecutive records into continuous tracks and splitting them into meaningful trajectories [2]
  3. Extracting flows by summarizing trajectory-based transitions between prototypes [3]

Besides clever aggregation approaches, massive movement datasets also require appropriate computing resources. To ensure that we can efficiently explore large datasets, we have implemented the above mentioned aggregation steps in Spark. This enables us to run the computations on general purpose computing clusters that can be scaled according to the dataset size.

In the next post, we’ll look at how to summarize movement using M³ prototypes. So stay tuned!

But if you don’t want to wait, these are the original papers:

[1] Graser. A., Widhalm, P., & Dragaschnig, M. (2020). The M³ massive movement model: a distributed incrementally updatable solution for big movement data exploration. International Journal of Geographical Information Science. doi:10.1080/13658816.2020.1776293.
[2] Graser, A., Dragaschnig, M., Widhalm, P., Koller, H., & Brändle, N. (2020). Exploratory Trajectory Analysis for Massive Historical AIS Datasets. In: 21st IEEE International Conference on Mobile Data Management (MDM) 2020. doi:10.1109/MDM48529.2020.00059
[3] Graser, A., Widhalm, P., & Dragaschnig, M. (2020). Extracting Patterns from Large Movement Datasets. GI_Forum – Journal of Geographic Information Science, 1-2020, 153-163. doi:10.1553/giscience2020_01_s153.

This post is part of a series. Read more about movement data in GIS.

Spatial on air: talking Python on the MapScaping Podcast

Podcasts have become huge. I’m an avid listener of podcasts myself. I particularly enjoy formats that take the time to talk about unconventional topics in detail.

My first podcast experience was on the QGIS podcast hosted by Tim Sutton in 2014. Unfortunately, it seems like the podcast episodes are not online anymore.

Recently, I had the pleasure to join the MapScaping Podcast by Daniel O’Donohue to talk about Python for Geospatial: 

Other guests Daniel has already interviewed include:

Another geospatial podcast I really enjoy is The Mappyist Hour by Silas and Todd. Unfortunately, it’s a bit silent there now but it’s definitely worth to listen into their episode archive. One of my favorites is Episode 9 where Linda Stevens (Hecht) discusses her career at ESRI, the future of GIS, and the role of Open Source Spatial in that future:

If you listen to and want to recommend other spatial podcasts, please share them in the comments!

Movement data in GIS #29: power your web apps with movement data using mobilitydb-sqlalchemy

This is a guest post by Bommakanti Krishna Chaitanya @chaitan94


This post introduces mobilitydb-sqlalchemy, a tool I’m developing to make it easier for developers to use movement data in web applications. Many web developers use Object Relational Mappers such as SQLAlchemy to read/write Python objects from/to a database.

Mobilitydb-sqlalchemy integrates the moving objects database MobilityDB into SQLAlchemy and Flask. This is an important step towards dealing with trajectory data using appropriate spatiotemporal data structures rather than plain spatial points or polylines.

To make it even better, mobilitydb-sqlalchemy also supports MovingPandas. This makes it possible to write MovingPandas trajectory objects directly to MobilityDB.

For this post, I have made a demo application which you can find live at The code for this demo app is open source and available on GitHub. Feel free to explore both the demo app and code!

In the following sections, I will explain the most important parts of this demo app, to show how to use mobilitydb-sqlalchemy in your own webapp. If you want to reproduce this demo, you can clone the demo repository and do a “docker-compose up –build” as it automatically sets up this docker image for you along with running the backend and frontend. Just follow the instructions in for more details.

Declaring your models

For the demo, we used a very simple table – with just two columns – an id and a tgeompoint column for the trip data. Using mobilitydb-sqlalchemy this is as simple as defining any regular table:

from flask_sqlalchemy import SQLAlchemy
from mobilitydb_sqlalchemy import TGeomPoint

db = SQLAlchemy()

class Trips(db.Model):
   __tablename__ = "trips"
   trip_id = db.Column(db.Integer, primary_key=True)
   trip = db.Column(TGeomPoint)

Note: The library also allows you to use the Trajectory class from MovingPandas as well. More about this is explained later in this tutorial.

Populating data

When adding data to the table, mobilitydb-sqlalchemy expects data in the tgeompoint column to be a time indexed pandas dataframe, with two columns – one for the spatial data  called “geometry” with Shapely Point objects and one for the temporal data “t” as regular python datetime objects.

from datetime import datetime
from shapely.geometry import Point

# Prepare and insert the data
# Typically it won’t be hardcoded like this, but it might be coming from 
# other data sources like a different database or maybe csv files
df = pd.DataFrame(
       {"geometry": Point(0, 0), "t": datetime(2012, 1, 1, 8, 0, 0),},
       {"geometry": Point(2, 0), "t": datetime(2012, 1, 1, 8, 10, 0),},
       {"geometry": Point(2, -1.9), "t": datetime(2012, 1, 1, 8, 15, 0),},

trip = Trips(trip_id=1, trip=df)

Writing queries

In the demo, you see two modes. Both modes were designed specifically to explain how functions defined within MobilityDB can be leveraged by our webapp.

1. All trips mode – In this mode, we extract all trip data, along with distance travelled within each trip, and the average speed in that trip, both computed by MobilityDB itself using the ‘length’, ‘speed’ and ‘twAvg’ functions. This example also shows that MobilityDB functions can be chained to form more complicated queries.


trips = db.session.query(

2. Spatial query mode – In this mode, we extract only selective trip data, filtered by a user-selected region of interest. We then make a query to MobilityDB to extract only the trips which pass through the specified region. We use MobilityDB’s ‘intersects’ function to achieve this filtering at the database level itself.


trips = db.session.query(
   func.intersects(Point(lat, lng).buffer(0.01).wkb, Trips.trip),

Using MovingPandas Trajectory objects

Mobilitydb-sqlalchemy also provides first-class support for MovingPandas Trajectory objects, which can be installed as an optional dependency of this library. Using this Trajectory class instead of plain DataFrames allows us to make use of much richer functionality over trajectory data like analysis speed, interpolation, splitting and simplification of trajectory points, calculating bounding boxes, etc. To make use of this feature, you have set the use_movingpandas flag to True while declaring your model, as shown in the below code snippet.

class TripsWithMovingPandas(db.Model):
   __tablename__ = "trips"
   trip_id = db.Column(db.Integer, primary_key=True)
   trip = db.Column(TGeomPoint(use_movingpandas=True))

Now when you query over this table, you automatically get the data parsed into Trajectory objects without having to do anything else. This also works during insertion of data – you can directly assign your movingpandas Trajectory objects to the trip column. In the below code snippet we show how inserting and querying works with movingpandas mode.

from datetime import datetime
from shapely.geometry import Point

# Prepare and insert the data
# Typically it won’t be hardcoded like this, but it might be coming from 
# other data sources like a different database or maybe csv files
df = pd.DataFrame(
       {"geometry": Point(0, 0), "t": datetime(2012, 1, 1, 8, 0, 0),},
       {"geometry": Point(2, 0), "t": datetime(2012, 1, 1, 8, 10, 0),},
       {"geometry": Point(2, -1.9), "t": datetime(2012, 1, 1, 8, 15, 0),},

geo_df = GeoDataFrame(df)
traj = mpd.Trajectory(geo_df, 1)

trip = Trips(trip_id=1, trip=traj)

# Querying over this table would automatically map the resulting tgeompoint 
# column to movingpandas’ Trajectory class
result = db.session.query(TripsWithMovingPandas).filter(
   TripsWithMovingPandas.trip_id == 1

# <class 'movingpandas.trajectory.Trajectory'>

Bonus: trajectory data serialization

Along with mobilitydb-sqlalchemy, recently I have also released trajectory data serialization/compression libraries based on Google’s Encoded Polyline Format Algorithm, for python and javascript called trajectory and trajectory.js respectively. These libraries let you send trajectory data in a compressed format, resulting in smaller payloads if sending your data through human-readable serialization formats like JSON. In some of the internal APIs we use at Adonmo, we have seen this reduce our response sizes by more than half (>50%) sometimes upto 90%.

Want to learn more about mobilitydb-sqlalchemy? Check out the quick start & documentation.

This post is part of a series. Read more about movement data in GIS.

Who is behind QGIS at Oslandia ?

You are using QGIS and look for support services to improve your experience and solve problems ? Oslandia is here to help you with our full QGIS editor service range ! Discover our team members below.

You will probably interact first with our pre-sales engineer Bertrand Parpoil. He leads Geographical Information System projects for 15 years for large corporations, public administrations or hi-tech SME. Bertrand will listen to your needs and explore your use cases, to offer you the best set of services.

Régis Haubourg also takes part in the first steps of projects to analyze your usages and improve them. GIS Expert, he knows QGIS by heart and will make the most of its capabilities. As QGIS Community Manager at Oslandia, he is very active in the QGIS community of developers and contributors. He is president of the Francophone OSGeo local chapter ( OSGeo-fr ), QGIS voting member, organizes the French QGIS day conference in Montpellier, and participates to QGIS community meetings. Before joining Oslandia, he led the migration to QGIS and PostGIS at the Adour-Garonne Water Agency, and now guides our clients with their GIS migrations to OpenSource solutions. Régis is also a great asset when working on water, hydrology and other specific thematic subjects.

Loïc Bartoletti develops QGIS, specializing in features corresponding to his fields of interests : network management, topography, urbanism, architecture… We find him contributing to advanced vector editing in QGIS, writing Python plugins, namely for DICT management. Pushing CAD and migrations from CAD tools to GIS and QGIS is one of his major goals. He will develop your custom applications, combining technical expertise and functional competences. When bored, Loïc packages software on FreeBSD.

Vincent Mora is senior developer in Python and C++, as well as PostGIS expert. He has a strong experience in numerical simulation. He likes coupling GIS (PostGIS, QGIS) with 3D numerical computing for risk management or production optimization. Vincent is an official QGIS committer and can directly integrate your needs into the core of the software. He is also GDAL committer and optimizes low-level layers of your applications. Among numerous activities, Vincent serves as lead developer of the development team for Hydra Software, a tool dedicated to unified hydraulics and hydrology modelling and simulation based on QGIS.

Hugo Mercier is an officiel QGIS committer too for several years. He regularly talks in international conferences on PostGIS, QGIS and other OpenSource GIS softwares. He will implement your needs with new QGIS features, develop innovative plugins ( like QGeoloGIS ) and design and build your new custom applications, solving all kind of technological challenges.

Paul Blottière completes our QGIS committers : very active on core development, Paul has refactored the QGIS server component to bring it to an industry-grade quality level. He also designed and implemented the infrastructure allowing to guarantee QGIS server performances. He dedicated himself to QGIS server OGC certification, especially for WMS (1.3). Thanks to this work QGIS is now a reference OGC implementation.

Julien Cabièces recently joined Oslandia, and quickly dived into QGIS : he contributes to the core of this Desktop GIS, on the server component, as well as applications linked to numerical simulation. Coming from a satellite imagery company with industrial applications, he uses his flexibility to answer all your needs. He brings quality and professionalism to your projects, minimizing risks for large production deployments.

You may also meet Vincent Picavet. Oslandia’s founder is a voting member, and is involved in the project’s evolution and the organization of the community.

Aside from these core contributors, all other Oslandia members also master QGIS integrate this tool into their day-to-day projects.

Bertrand, Régis, Loïc, Vincent (x2), Hugo, Paul et Julien are in tune with you and will be happy to work together for your migrations, application development, and all your desires to contribute to the QGIS ecosystem. Do not hesitate to contact us !

QGIS Snapping improvements

A few months ago, we proposed to the QGIS grant program to make improvements to the snap cache in QGIS. The community vote selected our project which was funded by Developments are now mostly finished.

In short, snapping is crucial for editing geospatial features. It is the only way to ensuring they are topologically related, ie, connected vertices have exactly the same coordinates even if manual digitizing on screen is imprecise by nature.  Snapping correctly supposes QGIS have in memory an indexed cache of the geometries to snap to. And maintainting this cache when data is modified, sometimes by another user or database logic, can be a real challenge. This it exactly what this work adresses.

The proposal was divided into two different tasks:

  • Manage circular dependencies
  • Relax the snap cache index build

Manage cicular data dependencies

Data dependencies

Data dependency is an existing feature that allows you to configure QGIS to reload layers (and their snapping cache) when a layer is modified.

It is useful when you store your data in a database and you set up triggers to maintain consistency between the different tables of your data model.

For instance, say you have topological informations containing lines and nodes. Nodes are part of lines and lines go through nodes. Then, you move a node in QGIS, and save your modifications to the database. In order to keep the data consistent, a trigger updates the geometry of the line going through the modified node.

Node 2 is modified, Line 1 is updated accordingly

QGIS, as a database client, has no information that the line layer currently displayed in the canvas needs to be refreshed after the trigger. Although the map canvas will be up to date, because QGIS fetches data for display without any caching system, the snapping cache is not and you’ll end up with ghost snapping highlights issues.

Snapping highlights (light red) differ from real line (orange)

Defining a dependency between nodes and lines layers tells QGIS that it has to refresh the line layer when a node is modified.

Dependencies configuration: Lines layer will be refreshed whenever Nodes layer is modified

It also have to work the other way, modifying a line should update the nodes to ensure they still are on the line.

Circular data dependencies

So here we are, lines depend on nodes which depend on lines which depend on nodes which…

That’s what circular dependencies is about. This specific behavior was previously forbidden and needed a special way to deal with it. Thanks to this recent development, it is now possible.

It’s also possible to add the layer itself as one of its own dependencies. It helps dealing with specific cases where one feature modification could lead to a modification of another feature in the same layer (to keep consistency on road networks for instance).

Road 2 is modified, Road 1 is updated accordingly

This feature is available in the next QGIS LTR version 3.10.

Relax the snapping cache index build

If you work in QGIS with huge projects displaying a lot of vector data, and you enable snapping while editing these data, you probably already met this dialog:

Snap indexing dialog

This dialog informs you that data are currently being indexed so you can snap on them while you will edit feature geometry. And for big projects, this dialog can last for a really long time. Let’s work on speeding it up!

What’s a snap index?

Let’s say you want to move a line and snap it onto another one. While you drag your line with the mouse, QGIS will look for an existing geometry beneath the mouse cursor (with a certain pixel tolerance) every time you move your mouse. Without spatial index, QGIS will have to go through every geometry in your layer to check if the given geometry is beneath the cursor position. This would be very ineffective.

In order to prevent this, QGIS keeps an index where vector data are stored in a way that it can quickly find out what geometry is beneath the mouse cursor. The building of this data structure takes time and that is what the progress dialog is about.

Firstly: Parallelize snap index build

If you want to be able to snap on all layers in your project, then QGIS will have to build one snap index for each layer. This operation was made sequentially meaning that if you have for instance 20 layers and the index building last approximatively 3 seconds for each, then the whole index building will last 1 minute. We made modifications to QGIS so that index building could be done in parallel. As a result, the total index building time could theoretically be 3 seconds!

4 layers snap index being built in parallel

However, parallel operations are limited by the number of CPU cores of your machine, meaning that if you have 4 cores (core i7 for instance) then the total time will be up to 4 times faster than when the building is sequential (and last 15 seconds in our example).

Secondly: relax the snap build

For big projects, parallelizing index building is not enough and still takes too much time. Futhermore, to reduce snap index building, an existing optimisation was to build the spatial index for a specific area of interest (determined according to the displayed area and layer size). As a consequence, when you’ve done waiting for an index currently building and you move the map or zoom in/out, you could possibly trigger another snap index building and wait again.

So, the idea was to avoid waiting at all. Snap index is now built whenever it needs to (when you first enable snapping, when you move or zoom) but the user doesn’t have to wait for the build to be over and can continue what it was doing (creating feature, moving…). Snapping highlights will be missing when the index is currently being built and will appear gradually as soon as they finished. That’s what we call the relaxing mode.

No waiting dialog, snapping highlights appears as soon as snap index is ready

This feature has been merged into current QGIS master and will be present in future QGIS 3.12 release. We keep working on this feature in order to make it more stable and efficient.

What’s next

We’ll continue to improve this feature in the coming days, if you have the chance to test it and encounter issues please let us know on the QGIS tracker. If you think about a missing feature or just want to know more about QGIS, feel free to contact us at [email protected]. And please have a look at our support offering for QGIS.

Many thanks to QGIS grant program for funding these new features. Thanks also to all the people involved in reviewing the code and helping to better understand the existing mechanism.


GRASS GIS 7.8.2 released

GRASS GIS 7.8.2 released with updated PROJ 6 and GDAL 3 support

What’s new in a nutshell

As a follow-up to the recent GRASS GIS 7.8.1 we have pusblished the new stable release GRASS GIS 7.8.2.
Besides other improvements, the release contains important PROJ 4/5/6 related datum handling fixes, wxGUI fixes and a fix for the vector import from PostGIS databases.

An overview of the new features in the 7.8 release series is available at new features in GRASS GIS 7.8.

Binaries/Installer download:

Source code download:

See also our detailed announcement:

First time users may explore the first steps tutorial after installation.


The Geographic Resources Analysis Support System (, commonly referred to as GRASS GIS, is an Open Source Geographic Information System providing powerful raster, vector and geospatial processing capabilities in a single integrated software suite. GRASS GIS includes tools for spatial modeling, visualization of raster and vector data, management and analysis of geospatial data, and the processing of satellite and aerial imagery. It also provides the capability to produce sophisticated presentation graphics and hardcopy maps. GRASS GIS has been translated into about twenty languages and supports a huge array of data formats. It can be used either as a stand-alone application or as backend for other software packages such as QGIS and R geostatistics. It is distributed freely under the terms of the GNU General Public License (GPL). GRASS GIS is a founding member of the Open Source Geospatial Foundation (OSGeo).

The GRASS Development Team, December 2019

The post GRASS GIS 7.8.2 released appeared first on GFOSS Blog | GRASS GIS and OSGeo News.

Folium vs. hvplot for interactive maps of Point GeoDataFrames

In the previous post, I showed how Folium can be used to create interactive maps of GeoPandas GeoDataFrames. Today’s post continues this theme. Specifically, it compares Folium to another dataviz library called hvplot. hvplot also recently added support for GeoDataFrames, so it’s interesting to see how these different solutions compare.

Minimum viable

The following snippets show the minimum code I found to put a GeoDataFrame of Points onto a map with either Folium or hvplot.

Folium does not automatically zoom to the data extent and I didn’t find a way to add the whole GeoDataFrame of Points without looping through the rows individually:

Hvplot on the other hand registers the hvplot function directly with the GeoDataFrame. This makes it as convenient to use as the original GeoPandas plot function. It also zooms to the data extent:

Standard interaction and zoom to area of interest

The following snippets ensure that the map is set to a useful extent and the map tools enable panning and zooming.

With Folium, we have to set the map center and the zoom. The map tools are Leaflet defaults, so panning and zooming work as expected:

Since hvplot does not come with mouse wheel zoom enabled by default, we need to set that:

Color by attribute

Finally, for many maps, we want to show the point location as well as an attribute value.

To create a continuous color ramp for a numeric value, we can use branca.colormap to define the marker fill color:

In hvplot, it is sufficient to specify the attribute of interest:

I’m pretty impressed with hvplot. The integration with GeoPandas is very smooth. Just don’t forget to set the geo=True parameter if you want to plot lat/lon geometries.

Folium seems less straightforward for this use case. Maybe I missed some option similar to the Choropleth function that I showed in the previous post.

Interactive plots for GeoPandas GeoDataFrames of LineStrings

GeoPandas makes it easy to create basic visualizations of GeoDataFrames:

However, if we want interactive plots, we need additional libraries. Folium (which is built on Leaflet) is a great option. However, all examples for plotting GeoDataFrames that I found focused on point or polygon data. So here is what I found to work for GeoDataFrames of LineStrings:

First, some imports:

import pandas as pd
import geopandas
import folium

Loading the data:

graph = geopandas.read_file('data/population_test-routes-geom.csv') = {'init' :'epsg:4326'}

Creating the map using folium.Choropleth:

m = folium.Map([48.2, 16.4], zoom_start=10)



I also tried using folium.PolyLine which seemed like the more obvious choice but does not seem to accept GeoDataFrames as input. Instead, it expects a list of coordinate pairs and of course it expects them to be in the opposite order that Shapely.LineString.coords provides … Oh the joys of geodata!

In any case, I had to limit the number of features that get plotted because Folium refuses to plot all 8778 features at once. I decided to filter by line length because drawing really short lines is pointless for my overview visualization anyway.

(Fr) Rechercher une adresse avec QGIS

Sorry, this entry is only available in French.

QGIS Versioning now supports foreign keys!

QGIS-versioning is a QGIS and PostGIS plugin dedicated to data versioning and history management. It supports :

  • Keeping full table history with all modifications
  • Transparent access to current data
  • Versioning tables with branches
  • Work offline
  • Work on a data subset
  • Conflict management with a GUI

QGIS versioning conflict management

In a previous blog article we detailed how QGIS versioning can manage data history, branches, and work offline with PostGIS-stored data and QGIS. We recently added foreign key support to QGIS versioning so you can now historize any complex database schema.

This QGIS plugin is available in the official QGIS plugin repository, and you can fork it on GitHub too !

Foreign key support


When a user decides to historize its PostgreSQL database with QGIS-versioning, the plugin alters the existing database schema and adds new fields in order to track down the different versions of a single table row. Every access to these versioned tables are subsequently made through updatable views in order to automatically fill in the new versioning fields.

Up to now, it was not possible to deal with primary keys and foreign keys : the original tables had to be constraints-free.  This limitation has been lifted thanks to this contribution.

To make it simple, the solution is to remove all constraints from the original database and transform them into a set of SQL check triggers installed on the working copy databases (SQLite or PostgreSQL). As verifications are made on the client side, it’s impossible to propagate invalid modifications on your base server when you “commit” updates.

Behind the curtains

When you choose to historize an existing database, a few fields are added to the existing table. Among these fields, versioning_ididentifies  one specific version of a row. For one existing row, there are several versions of this row, each with a different versioning_id but with the same original primary key field. As a consequence, that field cannot satisfy the unique constraint, so it cannot be a key, therefore no foreign key neither.

We therefore have to drop the primary key and foreign key constraints when historizing the table. Before removing them, constraints definitions are stored in a dedicated table so that these constraints can be checked later.

When the user checks out a specific table on a specific branch, QGIS-versioning uses that constraint table to build constraint checking triggers in the working copy. The way constraints are built depends on the checkout type (you can checkout in a SQLite file, in the master PostgreSQL database or in another PostgreSQL database).

What do we check ?

That’s where the fun begins ! The first thing we have to check is key uniqueness or foreign key referencing an existing key on insert or update. Remember that there are no primary key and foreign key anymore, we dropped them when activating historization. We keep the term for better understanding.

You also have to deal with deleting or updating a referenced row and the different ways of propagating the modification : cascade, set default, set null, or simply failure, as explained in PostgreSQL Foreign keys documentation .

Nevermind all that, this problem has been solved for you and everything is done automatically in QGIS-versioning. Before you ask, yes foreign keys spanning on multiple fields are also supported.

What’s new in QGIS ?

You will get a new message you probably already know about, when you try to make an invalid modification committing your changes to the master database

Error when foreign key constraint is violated

Partial checkout

One existing Qgis-versioning feature is partial checkout. It allows a user to select a subset of data to checkout in its working copy. It avoids downloading gigabytes of data you do not care about. You can, for instance, checkout features within a given spatial extent.

So far, so good. But if you have only a part of your data, you cannot ensure that modifying a data field as primary key will keep uniqueness. In this particular case, QGIS-versioning will trigger errors on commit, pointing out the invalid rows you have to modify so the unique constraint remains valid.

Error when committing non unique key after a partial checkout


There is a lot to check when you intend to replace the existing constraint system with your own constraint system based on triggers. In order to ensure QGIS-Versioning stability and reliability, we put some special effort on building a test set that cover all use cases and possible exceptions.

What’s next

There is now no known limitations on using QGIS-versioning on any of your database. If you think about a missing feature or just want to know more about QGIS and QGIS-versioning, feel free to contact us at [email protected]. And please have a look at our support offering for QGIS.

Many thanks to eHealth Africa who helped us develop these new features. eHealth Africa is a non-governmental organization based in Nigeria. Their mission is to build stronger health systems through the design and implementation of data-driven solutions.

QGIS 3 and performance analysis


Since last year we (the QGIS communtity) have been using QGIS-Server-PerfSuite to run performance tests on a daily basis. This way, we’re able to monitor and avoid regressions according to some test scenarios for several QGIS Server releases (currently 2.18, 3.4, 3.6 and master branches). However, there are still many questions about performance from a general point of view:

  • What is the performance of QGIS Server compared to QGIS Desktop?
  • What are the implications of feature simplification for polygons and lines?
  • Does the symbology have a strong impact on performance and in which proportion?

Of course, it’s a broad and complex topic because of the numerous possibilities offered by the rendering engine of QGIS. In this article we’ll look at typical use cases with geometries coming from a PostgreSQL database.


The first way to monitor performance is to measure the rendering time. To do so, the Map canvas refreshis activated in the Settings of QGIS Desktop. In this way we can get the rendering time from within the Rendering tab of log messages in QGIS Desktop, as well as from log messages written by QGIS Server.

The rendering time retrieved with this method allows to get the total amount of time spent in rendering for each layer (see the source code).

But in the case of QGIS Server another interesting measure is the total time spent for a specific request, which may be read from log messages too. There are indeed more operations achieved for a single WMS request than a simple rendering in QGIS Desktop:

The rendering time extracted from QGIS Desktop corresponds to the core rendering time displayed in the sequence diagram above. Moreover, to be perfectly comparable, the rendering engine must be configured in the same way in both cases. In this way, and thanks to PyQGIS API, we can retrieve the necessary information from the Python console in QGIS Desktop, like the extent or the canvas size, in order to configure the GetMap WMS request with the appropriate WIDTH,, HEIGHT , and BBOX parameters.

Another way to examine the performance is to use a profiler in order to inspect stack traces. These traces may be represented as a FlameGraph. In this case, debug symbols are necessary, meaning that the rendering time is not representative anymore. Indeed, QGIS has to be compiled in Debug mode.


For these tests we use the same dataset as that for the daily performance tests, which is a layer of polygons with 282,776 features.

Feature simplification deactivated

Let’s first have a look at the rendering time and the FlameGraph when the simplification is deactivated. In QGIS Desktop, the mean rendering time is 2591 ms. Using to the PyQGIS API we are able to get the extent and the size of the map to render the map again but using a GetMap WMS request this time.

In this case, the rendering time is 2469 ms and the total request time is 2540 ms. For the record, the first GetMap request is ignored because in this case, the whole QGIS project is read and cached, meaning that the total request time is much higher. But according to those results, the rendering time for QGIS Desktop and QGIS Server are utterly similar, which makes sense considering that the same rendering engine is used, but it is still very reassuring :).

Now, let’s take a look to the FlameGraph to detect where most of the time is spent.


Undoubtedly the FlameGraph’s are similar in both cases, meaning that if we want to improve the performance of QGIS Server we need to improve the performance of the core rendering engine, also used in QGIS Desktop. In our case the main method is QgsMapRendererParallelJob::renderLayerStatic where most of the time is spent in:

Methods Desktop % Server %
QgsExpressionContext::setFeature 6.39 6.82
QgsFeatureIterator::nextFeature 28.77 28.41
QgsFeatureRenderer::renderFeature 29.01 27.05

Basically, it may be simplified like:

Clearly, the rendering takes about 30% of the total amount of time. In this case geometry simplification could potentially help.

Feature simplification activated

Geometry simplification, available for both polygons and lines layers, may be activated and configured through layer’s Properties in the Rendering tab. Several parameters may be set:

  • Simplification may be deactivated
  • Threshold for a more drastic simplification
  • Algorithm
  • Provider simplification
  • Scale

Once the simplification activated, we varied the threshold as well as the algorithm in order to detect performance jumps:

The following conclusions can be drawn:

  • The Visvalingam algorithm should be avoided because it begins to be efficient with a high threshold, meaning a significant lack of precision in geometries
  • The ideal threshold for Snap To Grid and Distance algorithms seems to be 1.05. Indeed, considering that it’s a very low threshold, the precision of geometries is still pretty good for a major improvement in rendering time though

For now, these tests have been run on the full extent of the layer. However, we still have a Maximum scale parameter to test, so we’ve decreased the scale of the layer:

And in this case, results are pretty interesting too:

Several conclusions can be drawn:

  • Visvalingam algorithm should be avoided at low scale too
  • Snap To Grid seems counter-productive at low scale
  • Distance algorithm seems to be a good option


For these tests we also use the same dataset as that for daily performance tests, which is a layer of lines with 125,782 features.

Feature simplification activated

In the same way as for polygons we have tested the effect of the geometric simplification on the rendering time, as well as algorithms and thresholds:

In this case we have exactly the same conclusion as for polygons: the Distance algorithm should be preferred with a threshold of 1.05.

For QGIS Server the mean rendering time is about 1180 ms with geometry simplification compared to 1108 ms for QGIS Desktop, which is totally consistent. And looking at the FlameGraph we note that once again most of the time is spent in accessing the PostgreSQL database (about 30%) and rendering features (about 40%).







Another parameter which has an obvious impact on performance is the symbology used to draw the layers. Some features are known to be time consuming, but we’ve felt that a a thorough study was necessary to verify it.


Firstly, we’ve studied the influence of the width as well as the Single Symbol type on the rendering time.

Some points are noteworthy:

Simple Line is clearly the less time consuming

– Beyond the default 0.26 line width, rendering time begins to raise consequently with a clear jump in performance


Another interesting feature is the Draw effects option, allowing to add some fancy effects (shadow, glow, …).

However, this feature is known to be particularly CPU consuming. Actually, rendering all the 125,782 lines took so long that we had to to change to a lower scale, with just some a few dozen lines. Results are unequivocal:


The last thing we wanted to test for symbology is the effect of the Categorized classification. Here are the results for some classifications with geometry simplification activated:

  • No classification: 1108 ms
  • A simple classification using the column “classification” (8 symbols): 1148 ms
  • A classification based on a stupid expression “classification x 3″ (8 symbols): 1261 ms
  • A classification based on string comparison “toponyme like ‘Ruisseau*'” (2 symbols): 1380 ms
  • A classification with a specific width line for each category (8 symbols): 1850 ms

Considering that a simple classification does not add an excessive extra-cost, it seems that the classification process itself is not very time consuming. However, as soon as an expression is used, we can observe a slight jump in performance.


Another important part to study regarding performance is labeling and the underlying positioning. For this test we decreased the scale and varied the Placement parameter without tuning anything.

Clearly, the parallel labeling is much more time consuming than the other placements. However, as previously stated, we used the default parameters for each positioning, meaning that the number of labels really drawn on the map differs from a placement to another.


The last kind of geometries we have to study is points. Similarly to polygons and lines, we used the same dataset as that of performance tests, that is a layer with 435588 points.

In the case of points geometries geometry simplification is of course not available. So we are going to focus on symbology and the impact of marker size.

Obviously Font Marker must be used carefully because of the underlying jump in performance, as well as SVG Symbols. Moreover, contrary to Simple Marker, an increase of the size implies a drastic augmentation in time rendering.

General conclusion

Based on this factual study, several conclusions can be drawn.

Globally, FlameGraph for QGIS Desktop and QGIS Server are completely similar as well as rendering time.

It means that if we want to improve the performance of QGIS Server, we have to work on the desktop configuration and the rendering engine of the QGIS core library.

Extracting generic conclusions from our tests is very difficult, because it clearly depends on the underlying data. But let’s try to suggest some recommendations :).

Firstly, geometry simplification seems pretty efficient with lines and polygons as soon as the algorithm is chosen cautiously, and as long as your features include many vertices. It seems that the Distance algorithm with a 1.05 threshold is a good choice, with both high and low scale. However, it’s not a magic solution!

Secondly, a special care is needed with regards to symbology. Indeed, in some cases, a clear jump in performance is notable. For example, fancy effects and Font Marker SVG Symbol have to be used with caution if you’re picky on rendering time.

Thirdly, we have to be aware of the extra cost caused by labeling, especially the Parallel  placement for line geometries. On this subject, a not very well-known parameter allows to drastically reduce labeling time: the PAL candidates option. Actually, we may decrease the labeling time by reducing the number of candidates. For an explicit use case, you can take a look at the daily reports.

In any case, improving server performance in a substantial way means improving the QGIS core library directly.

Especially, we noticed thanks to FlameGraph that most of the time is spent in drawing features and managing the data from the PostgreSQL database. By the way, a legitimate question is: “How much time do we spend on waiting for the database?”. To be continued 😉

If you hit performance issues on your specific configuration or want to improve QGIS awesomeness, we provide a unique QGIS support offer at thanks to our team of specialists!

  • <<
  • Page 2 of 13 ( 249 posts )
  • >>
  • gis

Back to Top

Sustaining Members