Cardographer Jupyter explorations

Some explorations with Jupyter for analysing Cardographer usage data.

Getting started with Jupyter

Python & Jupyter

If you already use python then hopefully you know how you want to use python virtual environments, etc., so I'll leave that to you. If you don't already use python then...

It may be that Anaconda is the best place to start as it includes pandas and numpy, which should be useful for data analysis - see pandas install notes.

In theory Anaconda should be easier, but will need more space (1GB?) Anaconda installation. Alternatively something using miniconda will be smaller but more complicated to set up.

Install miniconda. Then, in a terminal,

conda create -n jupyter
source activate jupyter
conda install pandas

Install jupyter:

conda install -c conda-forge notebook
conda install -c conda-forge nb_conda_kernels
conda install -c conda-forge jupyter_contrib_nbextensions
# modules we are specifically using...
conda install -c conda-forge openpyxl matplotlib librosa
conda install pip

Run jupyter

Note, each time you open a new terminal you will need to switch to the 'jupyter' environment before you can do the other bits...

source activate jupyter

Then run Jupyter:

jupyter notebook

A browser should open showing a view of your home directory. Navigate to this directory.

Open, e.g. mixedreality_analysis1.

Creating a new python notebook

From the New dropdown select Python 3 (ipykernel). Rename the document (note, this is the filename).

The Data

There is some example data in data/. The initial example uses the "Detail" export, but not the "split", so all frames in one session appear in the same column.

The first column is card ID.

The other column names are (currently) Cardographer web platform session ID. Appending that string to the URL [https://cardographer.cs.nott.ac.uk/platform/user/session/](https://cardographer.cs.nott.ac.uk/platform/user/session/ may give you information about the session (if you have permission).

Working the data

See mixedreality_analysis1.

So far this loads the data as a Pandas dataframe, parses the JSON values and flattens the detail export array to give a list of frame titles and box names that the card is used in.

There are then a set of basic analyses which count how many groups use each card in the six main boxes on the board.