====== pockyt analysis

This is a simple visual analysis of a Pocket dataset.

Links

• GitHub repository <https://github.com/alfredjkwack/pockyt-analysis>
• D3.js <https://d3js.org/>
• alfredjkwack/pockyt <https://github.com/alfredjkwack/pockyt>

About

Pocket <https://getpocket.com/>_ is an application for managing a reading list of articles from the Internet.

pockyt-analysis provides a simple analysis of the pocket data using the D3.js JavaScript library for bespoke data visualization . It relies on a fork of pockyt (a commandline client that interfaces the pocket API) to obtain the data as a json file.

Installation

Since this project is dependent on alfredjkwack/pockyt we will assume you have already installed it.

1. Clone this repository: :code:git clone https://github.com/alfredjkwack/pockyt-analysis.git
2. Generate a json file of your pocket data: :code:pockyt -json -o pocket-data.json
3. Launch a local web server and connect to it with a browser: :code:python -m SimpleHTTPServer

TODO

• Tag usage frequency graph
• Tag co-occurrence graph
• Tag cloud
• Tag pair arc diagram
• Tag usage frequency graph over time
• Tag pair heat map (what goes together and what does not)
• Filter entries by 1 or more tags
• Clean data set to remove Null entries (not sure what those are)
• [-] Highlight entries with no tags somehow
• Refactor code to be more modular
• Table : URL's with no title need some love to shorten
• Table : Sorting of the date column needs to be correct
• Table : Sorting of the columns should not lead to overflow
• Table : Sorting of the columns should not lead to column widths changing
• Table : Tags need a nicer color to match the theme
• Table : The table head has a little too much white space