[{"_1":2,"_152":-5,"_153":-5},"loaderData",{"_3":4,"_13":14,"_20":21,"_131":132},"root",{"_5":-5,"_6":-7,"_7":8,"_9":-7,"_10":11,"_12":-7},"currentUser","primaryEmail","siteVerification","6LeIxAcTAAAAAJcZVRqyHh71UMIEGNQ_MXjiZKhI","toast","enableDev",false,"incidentStatus","routes/_layout",{"_15":-5,"_16":17,"_18":17,"_19":11},"username","isAskGordonEnabled",true,"isUsingNewTemplate","showMinimalLanding","routes/_layout.r.$namespace.$reponame",{"_22":23,"_24":-5,"_25":11,"_26":11,"_27":28,"_29":30,"_31":-5,"_32":11,"_33":34,"_83":-5,"_84":85,"_124":-5,"_125":11,"_126":17,"_127":11,"_128":-5,"_129":130},"canonicalUrl","https://hub.docker.com/r/nuest/geobia2016-qgis-model/tags","mcpServer","canSeeDockerfile","canSeeAutobuilds","namespace","nuest","reponame","geobia2016-qgis-model","cloudBuildSource","isAutobuildsAvailable","repo",{"_35":36,"_37":38,"_40":41,"_42":43,"_44":11,"_45":46,"_47":48,"_49":50,"_52":30,"_27":28,"_53":54,"_55":39,"_56":57,"_58":59,"_60":61,"_62":-5,"_63":57,"_64":65,"_66":11,"_67":28,"_68":69,"_74":17,"_75":76,"_80":28,"_81":82},"categories",[],"content_types",[39],"image","date_registered","2016-09-14T07:25:46.742732Z","description","...","is_private","last_updated","2018-01-03T13:15:26.270559Z","last_modified","2024-10-16T13:48:34.145251Z","media_types",[51],"application/vnd.docker.container.image.v1+json","name","pull_count",1626,"repository_type","star_count",0,"status",1,"status_description","active","source","collaborator_count","full_description","# Docker container for QGIS models\n\nThis project contains Docker containers to run [QGIS models](https://docs.qgis.org/2.14/en/docs/user_manual/processing/modeler.html) based on a model file and a minimal Python script from the command line. The Python script contains the actual model call, including all the required input data files (which a generic script cannot guess).\n\nThere are two ways to use the image. First, run it and mount the required workspacea.\nSecond, create a new Dockerfile to build an image that embeds the data.\n\nThere are two image configurations in this repository: \nThe first one is based on Ubuntu with a variant based on the UbuntuGIS repository. It was originally based on [Todd Stavish](https://github.com/toddstavish/Dockerfiles/tree/master/QGIS)'s work. Thanks, Todd!\nThe second is based on the QGIS Desktop container by [Kartoza](https://github.com/kartoza/docker-qgis-desktop) (Thanks!).\n\n_NB_: All commands in this document are executed from within the repository's root directory unless otherwise noted.\n\n## Example workflow: NDVI\n\nA small example workflow is contained in this repository at `workflows/example`. It calculates an [NDVI](https://en.wikipedia.org/wiki/Normalized_Difference_Vegetation_Index) for a [Sentinel satellite image](https://sentinel.esa.int/web/sentinel/sentinel-data-access) of the city center of Münster, Germany.\n\n### With Docker Hub image\n\nThere also is an automated build [on Docker Hub](https://hub.docker.com/r/nuest/docker-qgis-model), which you can use to execute the example and then extract the result with the `docker cp` command:\n\n```bash\ndocker run --name qgis_example nuest/docker-qgis-model:example\ndocker cp qgis_example:/workspace/results example_results\ndocker rm qgis_example\n```\n\nThe following gif shows an execution of the example container.\n\n![console gif of example execution](https://media.giphy.com/media/26BRxuXyhkReDH4dO/giphy.gif)\n\n### With local image\n\nAlternatively, you can build the base images as described below in the section \"Build the container\" and then build the example container locally.\n\nIn the directory `workspace/example`, run the following commands to build the image including the example data, run the container to execute the analysis, and then extract the output files to a local directory relative to the current path. The last command removes the image from local storage.\n\n```bash\ndocker build -t qgis-model-example .\ndocker run --name qgis_example qgis-model-example\n\ndocker cp qgis_example:/workspace/results example_results\ntree example-results\ndocker rm qgis_example\n```\n\nTake a look at the console - it contains several useful log statements. The directory `./example_results` contains the resulting GeoTIFF (`result.tif`) and a JPG preview file (`result.jpg`).\n\n### With external data\n\nA working example for calculating an NDVI based on a GeoTIFF is in the directory `/example`. To run it, first build the Ubuntu container and then run it with the following commands (executed from the root of this project):\n\n```bash\ndocker build -t docker-qgis-model:trusty -f ubuntu/trusty/Dockerfile ubuntu/.\ndocker run --rm -it -v $(pwd)/example/:/workspace qgis-model-ubuntu:trusty\n```\n\nOr using the image from Docker Hub:\n\n```bash\ndocker run --name qgis_example_hub -v /workspace/example/:/workspace nuest/docker-qgis-model:trusty\n```\n\n### With OSGeo-Live or local Linux OS\n\nThe [OSGeo-Live](http://live.osgeo.org/en/index.html) DVD or VM, also have all software you need to execute the workflow. If you use QGIS under Linux, this might even work on your own desktop computer.\n\nIf you want to give the example a try with OSGeo-Live, [download](http://live.osgeo.org/en/download.html) and start it and run the following commands in a terminal.\nThe steps are: clone the repository, copy the model file to the required location, and run the model:\n\n```bash\ngit clone https://github.com/nuest/docker-qgis-model.git\ncd docker-qgis-model/workspace/example\ncp models/compute_NDVI.model ~/.qgis2/processing/models/docker.model\npython model.py\n```\n\nTake a look at the data in the directory `/tmp/results/` (the full path is contained in the logs shown in the console).\n\nTo make this work, the file `model.py` uses default values for some of the environment variables that are originally defined in the Dockerfile.\n\n## Preparations to package a workflow\n\nPrepare a directory with the following contents. In the remainder of these instructions, we will assume it is called `data`.\n\n* `/models/*.model` - your QGIS workflow model (one only, or configure name via environment variable `QGIS_MODELFILE`)\n * when the container is executed, this file is copied into the model directory of the root user under the name `model.docker` so that it must be referenced in the python file as `\"modeler:docker\"`\n* `model.py` - a minimal Python file with the run command referencing the required data files, see template below. Depending on your model, the actual call to `processing.runalg(...)` will look very different. For example, it might container several inputs and/or outputs.\n* all data files\n* any additional user scripts in `scrips/*.py`\n\n**Python file template:**\n\n```bash\n#!/usr/bin/python\n\n# Run preparation file\nimport sys\nsys.path.append(\"/qgis/util\")\nimport prepare\n\n# Import and initialize Processing framework\nfrom processing.core.Processing import Processing\nProcessing.initialize()\nimport processing\n\n# Run model\ninput=\"/workspace/data/input.file\"\noutput=\"/workspace/data/output.file\"\n\nprint \"Start processing...\"\nprocessing.runalg(\"modeler:docker\",input,output)\nprint \"Processing complete\"\n```\n\n## Run the model with a mounted workspace\n\nBuild the container (see below) and start it with the following command, mounting your `workspace` directory to `/workspace` and replacing `` with either `debian` or `ubuntu`. In the latter case we recommend to explicitly select the Ubuntu version and thereby the QGIS version by appending either the tag `:trusty` or `:xenial`. If you want to publish your whole model in a self-contained image, see next section.\n\n```bash\ndocker run --rm -it -v /:/workspace qgis-model-\n```\n\n### Configuration\n\nIf you want to run the model manually (i.e. to debug etc.) append `/bin/bash` to the command to override the default command, then execute `./qgis/model.sh` manually.\n\nThe used options are as follows:\n\n* `--rm` will remove the container as soon as it ends\n* `-it` ensures you can see the stdout logs\n* The startup script will copy all files named `*.model` from the mounted directory\n\nPotentially useful additional options are as these:\n\n* `--name=\"qgis\"` to name the running container for easier identification\n* To *execute a specific model*, you can overwrite the environment variable `QGIS_MODELFILE` which has the default value `/workspace/models/*.model`. This is useful if you have a working directory with more than one model file. Specify the environment variable when you run the container: `docker run --rm -it -v /:/workspace -e QGIS_MODELFILE=/workspace/models/mymodel.model qgis-model-`\n* If you want to run the model manually (i.e. to debug etc.) add the parameter `--entrypoint=/bin/bash` to the command (before the image name) to override the default entrypoint and get a bash shell, then execute `./qgis/model.sh` manually.\n\n### Access log while running\n\nAfter the model has started you can access the current state of the containers log file with the command `docker exec`:\n\n```bash\ndocker ps\n# note the name of the container running the command /qgis/model.sh\ndocker exec cat /qgis/qgis.log\n```\n\nAlternatively to `cat`, you can use `less` or any other tools available in the container.\n\n## Create a self-contained image\n\nThe previous run command mounts a directory of the host computer to the container, which is suitable for model development. If you want to publish a self-contained Docker image, you can create a minimal Dockerfile based on the images created above, which simply copies your data into the container, then build and execute that image.\n\nBe aware that you to access the output of the process you must not use `--rm` but keep the container to extract the data and delete it afterwards. See above in section \"Example with embedded data\" for the required commands.\n\n## Build the container\n\n### Ubuntu\n\nSee directory `/ubuntu/Dockerfile.` for the respective Dockerfile.\n\nExecute the following command in the root directory `/` of this repository to build the container and name it.\n\n```bash\ndocker build -t qgis-model: -f ubuntu/Dockerfile. ./ubuntu\n```\n\nThe build context is set to `./ubuntu` so that the required model file can be copied. The Dockerfile name is configured explicitly with the `-f` option.\n\nThe following command builds the image for Ubuntu 16.04 and tags it as being the \"latest\".\n\n```bash\ndocker build -t qgis-model:xenial -t qgis-model:latest -f Dockerfile.xenial ./ubuntu\n```\n\n### Kartoza\n\nSee directory `/kartoza` for the Dockerfile, which uses [kartoza/qgis-desktop](https://hub.docker.com/r/kartoza/qgis-desktop/) by Tim Sutton, using the tag `2.8` as the base image. Find the Dockerfiles on GitHub: [https://github.com/kartoza/docker-qgis-desktop/tree/master/2.8](https://github.com/kartoza/docker-qgis-desktop/tree/master/2.8). The Kartoza images are based on `Ubuntu 16.04`. The following commands build the image with a tag and then run a container on that tag with the example workspace.\n\n```bash\ndocker build -t qgis-model-kartoza kartoza/.\n\ndocker run --rm -it -v $(pwd)/workspace/example/:/workspace qgis-model-kartoza\n```\n\nYou can also run an interactive version of this container (with QGIS user interface) by adding the following parameters to the `docker run` call, for details see [the base images starth.sh script](https://github.com/kartoza/docker-qgis-desktop/blob/develop/2.14/start.sh).\n\n```bash\nxhost +local:docker\ndocker run -it --rm -v /:/home/ -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY qgis-model-kartoza /start.sh\nxhost -local:docker\n```\n\n## LICENSE\n\nThis project is published under the Apache License, Version 2.0.\n\nCopyright 2016 Daniel Nüst.\n","has_starred","hub_user","immutable_tags_settings",{"_70":11,"_71":72},"enabled","rules",[73],".*","is_automated","permissions",{"_77":17,"_78":11,"_79":11},"read","write","admin","user","storage_size",3360424972,"orgData","recentTags",[86],{"_87":39,"_88":89,"_90":91,"_92":93,"_94":95,"_45":113,"_114":89,"_115":28,"_116":51,"_52":117,"_118":119,"_120":104,"_121":113,"_122":61,"_123":17},"content_type","creator",30166,"full_size",696534254,"id",4837058,"images",[96],{"_97":98,"_99":100,"_101":102,"_103":104,"_105":106,"_107":108,"_109":102,"_110":-5,"_111":91,"_58":61,"_112":-5},"architecture","amd64","digest","sha256:9c3978f85c81c82fd055650985f381d5d27dd7fb7e48ad8531dcaadd1de2fa58","features","","last_pulled","2026-09-06T22:42:31.268795585Z","last_pushed","2018-01-03T13:15:26Z","os","linux","os_features","os_version","size","variant","2018-01-03T13:15:25.968008Z","last_updater","last_updater_username","media_type","latest","repository",915472,"tag_last_pulled","tag_last_pushed","tag_status","v2","sbxKitMetadata","isPotentialSpam","isHubSbxKitsEnabled","displaySecurityMetadata","tagAttestations","tagsHaveAttestations",[],"routes/_layout.r.$namespace.$reponame.tags",{"_33":133,"_140":141,"_129":150,"_151":11},{"_35":134,"_37":135,"_40":41,"_42":43,"_44":11,"_45":46,"_47":48,"_49":136,"_52":30,"_27":28,"_53":54,"_55":39,"_56":57,"_58":59,"_60":61,"_62":-5,"_63":57,"_64":65,"_66":11,"_67":28,"_68":137,"_74":17,"_75":139,"_80":28,"_81":82},[],[39],[51],{"_70":11,"_71":138},[73],{"_77":17,"_78":11,"_79":11},"tags",{"_142":59,"_143":-5,"_144":-5,"_145":146},"count","next","previous","results",[147],{"_87":39,"_88":89,"_90":91,"_92":93,"_94":148,"_45":113,"_114":89,"_115":28,"_116":51,"_52":117,"_118":119,"_120":104,"_121":113,"_122":61,"_123":17},[149],{"_97":98,"_99":100,"_101":102,"_103":104,"_105":106,"_107":108,"_109":102,"_110":-5,"_111":91,"_58":61,"_112":-5},[],"canDeleteTags","actionData","errors"]

nuest/geobia2016-qgis-model

Tags for nuest/geobia2016-qgis-model