The hands-on session is manily built around the capabilities of the Python programming language and the R statistics software.
As a general programming language, Python will serve as our gateway to API interactions. Follow the guides of this page to install a Python version suitable for your operating system.
Some graphical interfaces for Python would be useful, however they are not necessary. Just to name a few, IDLE (that comes with most Python installation on Windows and Mac), PyScripter or Eric. You can of course use any other IDEs depending on your preferences.
QGIS can be used for analysing a number of the datasets obtained within this workshop. As well as being suitable for creating visualisations and assessments of the data, a plugin has also been developed that can be used for viewing images identified from the datasources such as Flickr and Mapillary.
To install this plugin, go to Plugins->manage and Install Plugins and then under "Settings" - "Plugin repositories" add a new repository with the URL of http://cap4navi.geog.uni-heidelberg.de/link-vgi/qgis-plugins/qgis-repo.xml , along with a name (it doesn't matter what name you use, this is just so you can identify where the plugins come from) followed by clicking the "OK" button. Right click on the newly added entry (there should be a green icon next to it and the text "connected") and choose "Only show plugins from selected repository". This will make it easier to find the plugin. Also, make sure that the "Show experimental plugins" checkbox on the settings page has been checked. Now go to "Uninstalled", click on the Image Viewer plugin and select install.
To use the plugin, either click on the new icon in the toolbar (a yellow square with a small arrow and camera) or select it from the Plugins->Image Viewer menu. You will also need to have imported the csv files with the actual points in as layers. When the plugin window is open, select from the drop down box which type of image you are going to be selecting (Mapillary, Flickr or Instagram) and then make sure that the corresponding layer has been selected in the Layers panel of QGIS. Now, when you click on one of the points form this layer, the image will be shown in the plugin window.
R provides a powerful statistics environment for various tasks. You can follow numerous guides out there to download and install R.
sudo apt-get update
sudo apt-get install r-base
Download the installer from here and follow the steps.
Download the pacakge from here and follow the steps.
You can use any available IDEs. However, RStudio is recommended. You can find the installer packages here.
Install a stable version (e.g. 9.4) of PostgreSQL and PostGIS depending on your operating system.
Windows users should use the application stackbuilder from EnterpriseDB and follow the instructions there. Pay attention to install PostGIS spatial extension and pgAdmin graphical interface during the installation process as well.
You can follow many resources out there. It is recommended to install it as a package. Try for example
sudo sh -c 'echo "deb http://apt.postgresql.org/pub/repos/apt/ $(lsb_release -cs)-pgdg main" > pgdg.list'
wget --quiet -O - https://www.postgresql.org/media/keys/ACCC4CF8.asc | sudo apt-key add -
sudo apt-get update
sudo apt-get upgrade
sudo apt-get install postgresql-9.4-postgis-2.1 pgadmin3 postgresql-contrib-9.4 python-psycopg2
Mac users should use the application stackbuilder from EnterpriseDB and follow the instructions there. Pay attention to install PostGIS spatial extension and pgAdmin graphical interface during the installation process as well.
On top of Python, we are going to use some additional packages as well that makes programming easier. These additional packages are usually developed for a specific purpose (e.g. to interact with Twitter or to handle HTTP requests in general).
Tweepy is a Python wrapper built around Twitter API. It allows us to easily interact with Twitter data in a Python environment. You can install Tweepy from PyPi using easy_install or pip
pip install tweepy
This python package allows to read and write regular shapefiles in a Python environment. Use the prebuilt package from PyPi with easy_install or pip.
pip install pyshp
It is recommended to install psycopg2 from a prebuilt package for your Operating System. Some help can be found here.
Linux (Ubuntu/debian)
In general the package for Linux distributuions is called python-psycopg2. On Debian/Ubuntu, simply run
sudo apt-get install python-psycopg2
Mac OS X
fink install psycopg2-py27
or
sudo port install py27-psycopg2
Windows
Use an installer for win-psycopg from here.
Info can be found here. Install using pip
pip install flickrapi
As of June 1, 2016 access to the Instagram API has been seriously cut down by Instagram. You can still install this API wrapper and experiment with it but data access is extremely limited.
pip install python-instagram
We are going to use the following R packages. Open an R session and type run following lines.
install.packages('ggplot2')
install.packages('lubridate')
install.paclages('plyr')
install.packages('reshape2')
install.packages('ggmap')
install.packages('wordcloud')