Digital Geography

6. April 2015

Introducing: My personal application list

Once again I switched from Windows back to Ubuntu as I only needed Win for my granted ESRI license but hated to be “interrupted” by automatic updates, missing dll libraries and crashing flash container plugins on Firefox. So now I am building my personal computational slave once again based on Ubuntu and wanted to share with you my list of applications that I need for my daily work with geodata, matrices, graphics, cartographic outputs and statistical analysis.

The Hardware

I am running on a 3 year old Thinkpad X230:
  • 8GB RAM
  • i7-3520M CPU @ 2.90GHz × 4
  • 170GB SSD harddrive
  • 9 cell battery
  • external display? check!
I am using it mostly in a non-desktop way: in the train, on the airport/airplane, shared desks and so on.

The Usage Scenario

I am mostly working with geodata:
  • map design
  • ETL (extract, transform, load) tasks
  • geodata acquisition
  • geostatistical analysis
  • webmap production
My second major work is related to programming:
  • coding
  • GUI design
  • general design
  • hosting
  • database work
Looking at this short list you might get an idea about my software stack. So let’s check your thoughts:

The Application List

  • QGIS– for my major GIS work. Supports nearly anything with its Python backend
  • PostgreSQL with postGIS extension – data mart for my daily tasks and webmapping projects
  • RStudio – a GUI for the statistical, functional language R with syntax highlighting, package support, embedded pictures and visual log to do analysis on my spatial data
  • Filezilla – a FTP client to connect with ftp data providers and to the digital-geography.com servers
  • Inkscape – an editor for vector graphics. Useful in final level map production as well as marker/pin designs
  • Gimp – always good for image enhancement and maybe broader image related work
  • Libre Office – a good office package. For simple forms/graphics I use Libre Office Impress a lot!
  • Knime – for automated tasks when it comes to geocoding addresses from a file or reprojecting coordinates in a xls file
  • Sublime3 – I love a good editor which supports any coding style
  • Virtualbox – to host my Windows installations and my portable servers
  • Tilemill – to create my own tiles for stunning webmaps

custom tiles with tilemill

So this was probably more GIS/every day applications. As we are working on qgis2leaf and some other software I also need something for development:
  • QT designer – to create GUIs for QGIS plugins
  • GIT – our favourite collaboration tool at Geolicious
  • Sublime3 – once again: a great editor!!!
  • Deepin Terminal – an enhanced terminal for Ubuntu
So now it’s your turn! What is useful for your daily work? What did I miss?
  • F. Shah

    Pycharm is a great tool. I second Sublime3 to Pycharm otherwise your list is not different than mine. 😛

    • I also used pycharm a while ago, but I like sublime as it supports all other scripting languages as well 😉 Of course pycharm support html, css and JS but what about R, Matlab and Markdown?

      • F. Shah

        Agreed. Sublime has been my favourite editor since I have mainly worked in MATLAB and C++. I quit MATLAB in a sprit of FOSS and I ended up with Pycharm (Python) and QtCreator( C++). I Occasionally use it for Octave as well. This combination is boosting my speed by several times and I am happy with it but this might be different for others.

  • Chk.

    FileZilla bundles adware with its installer these days, I check out WinSCP. Especially if you are dealing with Linux/SSH accounts from time to time as well.

  • Ruben Remelgado

    For me idl is a necessity for multiple tasks, from image analysis to statistical analys and report generation. I find that R is useful for a lot of similar tasks anf it is definetely a lot easier to use. Since a lot offspatial analys tools are already implemented the coding effort is minimal. But it is far less powerfull for image processing at the moment. Idl has long years of work built into it for this purpose.
    Also, wthe increase in data size, hardware specifications tend to increase. In my previous job i was relying in a 8 core 32gb ram machine and roughly 1tb of to manage the data load.

  • And if you need to layout a brochure or blurb it’s hard to go past Scribus. It gets a lot of use on my OpenSUSE GIS workstation …

  • pimplbe

    How steep is the learning curve with KNIME ? It looks like a useful tool but reminds me of WEKA
    (what scares me ;>) with its Java backend and the graphical interface.

  • Hansinger

    i can recommend SADA for geostatistical analysis, especially indicator kriging… as far as i know this is the easiest way to do indicator kriging which is important for regionalization of non normal distributed data … they got a pretty nice support and an awesome documentation

  • As a long time linux user, apart of most of yours I use:

    – Terminator as my terminal emulator, lots of features
    – Python virtualenvironments to isolate projects and a bigger “datasci” environment
    – iPython for prototyping, simple scripting, webscrapping, and much more
    – Lots of small unix-style pieces to pipeline pieces like old grep and awk but also some new like jq, xml-json, csvlook, etc
    – Shutter to do and edit screenshots
    – Hamster to track my time

  • Jesús DC

    It´s nice to know how colegues work! very useful for me. Thanks +Riccardo!