XOvis

The objective of this project is to gain insights into how XOs are used in Nepalese classrooms. In order to learn about XO usage, first data is collected from their Journal backups on the schoolserver. Subsequently, the data is processed and imported into a database and finally visualized using fancy interactive charts. The application that implements the visualization of the data is called XOvis, a Couch App built using Kanso framework. See dataflow.svg for a schematic representation of the entire workflow.

Installation

On RHEL/Fedora machine

Install with a bash script

• To install the application, simply run install_xovis.sh

    ./scripts/install_xovis.sh
  

Install manually

• Install project dependencies

    yum install python-pip git couchdb nodejs npm"
  

• Install Kanso framework

    npm install -g kanso
  

• Clone this repository in the destination of your choice

    git clone https://github.com/martasd/xovis.git
  

• Install the application into a new database

    cd xovis
    kanso createdb http://localhost:5984/xovis
    kanso push http://localhost:5984/xovis
  

  where xovis is the name of the new database.

Alternative install (without requiring NodeJS)

• If you prefer to avoid having to install Node JS, you can also install from a database dump:

    ./scripts/install_xovis_dbdump.sh
  

  Note: Downloads of database dump file.

Load existing deployment data into the database

• Insert XO Journal backup data into the same database using a Python script

    pip install -r requirements.txt
    ./process_stats/process_journal_stats.py dbinsert xovis --deployment <deployment-name>
  

  Note:

  The script can also output statistical data to a file instead of inserting it into a database. To produce all statistical data from the Journal, one row per Journal record, call:

    process_journal_stats.py all
  

  To extract statistical data about the use of activities on the system, use:

    process_journal_stats.py activity
  

  To learn about all options of the script, see:

    process_journal_stats.py --help
  

• To manage Couch databases using a browser dashboard, go to

    http://localhost:5984/_utils
  

Visualize

• Open up a browser and go to

    http://localhost:5984/<deployment-name>/_design/xovis-couchapp/index.html
  

Enjoy the beautiful view!

Acknowledgments

process_journal_stats.py script is based on olpc-journal-processor script Leotis' Buchanan and get-journal-stats by Raul Gutierrez Segales.