https://github.com/nens/raster-tools

Collection of cli tools for raster-processing.
https://github.com/nens/raster-tools

Last synced: 4 days ago
JSON representation

Collection of cli tools for raster-processing.

• Host: GitHub
• URL: https://github.com/nens/raster-tools
• Owner: nens
• License: gpl-3.0
• Created: 2014-01-15T13:47:35.000Z (almost 11 years ago)
• Default Branch: master
• Last Pushed: 2022-12-06T09:23:24.000Z (almost 2 years ago)
• Last Synced: 2024-04-12T01:12:29.476Z (7 months ago)
• Language: Python
• Size: 692 KB
• Stars: 6
• Watchers: 34
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.rst
  • Changelog: CHANGES.rst
  • License: LICENSE.rst

Awesome Lists containing this project

README

        
raste-tools

============

A collection of raster tools.

Development installation

------------------------

For development, you can use a docker-compose setup::

    $ docker-compose build --build-arg uid=`id -u` --build-arg gid=`id -g` lib

    $ docker-compose up --no-start

    $ docker-compose start

    $ docker-compose exec lib bash

Create a virtualenv, install dependencies & package, run tests::

    # note that Dockerfile prepends .venv/bin to $PATH

    (docker)$ virtualenv --system-site-packages .venv 

    (docker)$ pip install -r requirements.txt --index-url https://packages.lizard.net

    (docker)$ pip install -e .[test]

    (docker)$ pytest

Update packages::

    

    (docker)$ rm -rf .venv

    (docker)$ virtualenv --system-site-packages .venv

    (docker)$ pip install -e . --index-url https://packages.lizard.net

    (docker)$ pip freeze | grep -v raster-tools > requirements.txt

Now you are ready to run the raster tools in the container.

Mapping an extra local folder

-----------------------------

First, clone this repo and make some required directories::

    $ git clone [email protected]:nens/raster-tools

    $ cd raster-tools

Create a docker-compose.override.yaml to map local filesystems:

.. code-block:: yaml

    version: '3'

    services:

      lib:

        volumes:

          - /some/local/path:/some/container/path

Task server installation

------------------------

One time server provisioning::

    ansible-playbook ansible/provision.yml -i ansible/task.yml

Deploying new versions::

    ansible-playbook ansible/deploy.yml -i ansible/task.yml 

Be aware that if you provision a server for the first time, the authentication

needs to be setup. We may need a credentials file to access private packages on

packages.lizard.net, to be put in `deploy/files/nens_netrc` file.

To enable users to use the raster-tools scripts, append the absolute path to

``.venv/bin`` to the PATH in ``/etc/environment``.

Filling nodata in rasters

-------------------------

Use fillnodata to fill nodata regions in rasters. The input to the algorithm is

the edge surrounding the region. The raster saved as the target argument only

contains the complementary cells.

Rasterizing landuse tables

--------------------------

For rasterization of landuse tables from a postgres datasource a special

wrapper command is available at bin/rasterize-landuse, use --help for args.

Creating streamlines

--------------------

Run the following scripts for streamline calculation::

    flow-fil  # depression filling

    flow-dir  # direction calculation

    flow-acc  # accumulation

    flow-vec  # make shapefiles

    flow-rst  # make rasters from shapefiles

Streamlines have been rasterized in the past to be able to visualize them with

good performance on any zoomlevel. A number of tricks can be to make them look

like a vector dataset:

- Rasterize at sufficiently high resolution so that supersampling is never

  needed

- (Pre)aggregate using a `maximum` algorithm for lower resolutions

- Use binary dilation after serving image tiles on one or more higher classes

  to create a 'wider stroke' effect

Multiprocessing

---------------

A number of scripts have a ``--part`` option to run the script on a subset of the

source features, e.g. --part=2/3 to run on the second part of three parts. To

use xargs to run some script for example on 4 processes, use::

    xargs -a <(echo -e '1/4\n2/4\n3/4\n4/4') -L 1 -P 4 your_script --your_args --part