Geospatial representation are so prevalent in day to day life, such as even in simple travel related conversation to maps, aerial/satellite images etc. In digital era, geospatial data is extensively produced and consumed in ever growing proportion. Python with its free and open source libraries are giving wide variety yet simple and effective set of tools to visualise and analyse geospatial data. The current workshop is directed for beginners of Python programming language, who have basic understanding on computing and data formats. The primary objective of the workshop is to introduce and give hands on training on selected list of FOSS libraries for geospatial analysis. The workshop as a do it yourself fashion tries to solve two real world problems in Geographical Information System (GIS) and its geospatial data sources.

The workshop comprised of three components:

Component 1: Introduction

Python environment and work flow setup, an assisted task of setting up the Docker and Jupyter notebook setup. Setting up the Geographical Information System (GIS) environment with extended discussion. Setting up of GIS tools such as FOSS QGIS and Google earth. This component is comprised of four exercises. 1. Introduction to vector data, 2. Introduction to raster data, 3. binary and text file formats of geospatial data, 4. Introduction to tools of GIS, 5. Introduction to literal programming- Jupyter notebook

Component 2: work with vector data

Find characteristics of road network(type of road network, length of the type) within a 1X1 km grid. The data source is Open Street Map (OSM) road network data on a city level (60X60km size). This operation is operationally simple such as measure a line feature but computationally intensive as the operation comprised of geometry within operation on dense road network seen in urban setup. Libraries such as Shapely, Fiona, Geopandas and rtree index will be used for the fast processing of this operation. This component comprised of three exercises 1. Find distance between two points 2. Find distance between two points constrained by another vector 3. Find distance between large number of points in for loop

Component 3: work with raster data

Find cloud cover percentage over area of interest. The data source is Landsat satellite imagery. Searching cloud free Landsat images over an Area of Interest for a temporal extent of a year or more is manual and time consuming. Applying cloud cover detection algorithm could make this operation automatic. Libraries such as rasterio, Geopandas, Fiona, and libraries related to landsat algorithms will be used for this task. This component comprised of two exercises 1. Convert the imagery in geotiff into numpy arrays 2. Apply the algorithms to find the cloud cover

Note : There won't be any environment setup session separately, participants are requested to ensure below prerequisites are working properly.

Workshop Plan

1. Component 1- 40 minutes
2. Break- 5 minutes
3. Component 2- 50 minutes
4. Break- 5 minutes
5. Component 3- 50 minutes

1. Laptop 32bit/64 bit

2. Workshop material is tested on 64 bit computer, it is said to be working in 32 bit, lets experiment!

3. A copy of Docker container image from here, file from the link foss-pt-gsa_v3.tar.gz is 2.5 GB in size, will be using this container for DIY Linux

4. Install docker, I followed digital ocean install tutorial , which worked!.

5. Use docker to load downloaded tar into it as an container image, before that unzip the foss-pt-gsa_v3.tar.gz into foss-pt-gsa_v3.tar

   docker load -i  foss-pt-gsa_v3.tar 
   

6. To check the docker is loaded with new image, ensure the image foss-pt-gsa/foss-pt-gsa:version3 is listed by entering

   docker images
   

7. To run the image, enter

   docker run -dit foss-pt-gsa/foss-pt-gsa:version3
   

8. Get the CONTAINER_ID of the just started container by the command docker ps

9. To enter into the image bash

   docker exec -it CONTAINER_ID bash
   

10. After enter into the image’s bash terminal, enter following commands. the commands download the workshop github repo zip file into a working direcoty, then unzip it and get into the repo folder to start a jupter notebook server

     cd /home/ubuntu/  
     wget https://github.com/nishadhka/FOSS-Python-GeospatialAnalysis/archive/master.zip
     unzip master.zip
     cd FOSS-Python-GeospatialAnalysis-master
     jupyter notebook --ip 0.0.0.0 --no-browser --allow-root
    

11. Note down the link provided by the jupyter notebook such as example http://0.0.0.0:8889/?token=c8e944b8397b0bde97b4d9284e5e3ffc0136658fcca3ea1e

12. Logout from the docker image bash (by typing ctrl+q or closing the bash window) and in the host computer note down the image_ID of the workshop image running inside the docker by

    docker ps
    

13. Then inspect about the docker image to get to know the image’s IP address. Note down the ipaddress

    docker inspect image_ID | grep "IPAddress"
    

14. Edit the jupter server given link as into http://ipaddress:8889/?token=c8e944b8397b0bde97b4d9284e5e3ffc0136658fcca3ea1e

15. Open the link in host computer browser, it shows the Jupyternotebooks in the workshop repo and click on the file docker_test.ipynb, to run the notebook and execute its first cell to ensure all the libraries for the workshop is working properly

Windows

1. Local copy of Docker toolbox from here for windows 64 bit, for 32 bit Windows, follow this link.

2. Local copy boot2docker.iso from here, it is better to follow old method of docker toolbox instead of docker native software for Windows.

3. save the iso file in C:\Users\User.docker\machine\cache\boot2docker.iso

4. Open the program Docker quick start, It will make a separate virtual machine to run the docker container

5. Under the Docker quick start program, after the virtual machine run, check docker is working by entering command docker ps

6. Now, import the workshop container by

   docker load -i foss-pt-gsa_v3.tar
   

7. Then run the imported container image by

   docker run --name jupyter1 --rm -it --user root -d -p 8080:8080 jupyter/r-notebook:version3
   

8. again check the command docker ps, it would show the container is running by having a container ID

9. To get into the running docker

   docker exec -it container_ID bash
   

10. After enter into the image’s bash terminal, enter following commands. the commands download the workshop github repo zip file into a working direcoty, then unzip it and get into the repo folder to start a jupter notebook server

     cd /home/ubuntu/  
     wget https://github.com/nishadhka/FOSS-Python-GeospatialAnalysis/archive/master.zip
     unzip master.zip
     cd FOSS-Python-GeospatialAnalysis-master
     jupyter notebook --ip=0.0.0.0 --port=8080 --allow-root
    

11. Note down the link provided by the jupyter notebook such as example http://0.0.0.0:8080/?token=c8e944b8397b0bde97b4d9284e5e3ffc0136658fcca3ea1e

12. Replace the 0.0.0.0 with default 192.168.99.100 address of docker toolbox as per this