Panama Papers Dataset 2016
Structured data about Panama papers collected from the official ICIJ website
Update: The full dataset that has been released by ICIJ recently is available here
How were the data collected ?
I wrote a detailed explanation here
Dataset structure
├── de.csv
├── en.csv
├── es.csv
├── fr.csv
├── pt.csv
└── viz-data
├── 014fe3bc.json
├── 02664021.json
├── 034793d9.json
├── 08cc5165.json
├── 0a1122ef.json
├── 0a3cf4fd.json
├── 0dae973f.json
...
...
...
├── fb706ef9.json
├── fb7f4be3.json
└── ff9bbc4d.json
└── csv-viz-data
├── 014fe3bc-nodes.csv
├── 014fe3bc-edges.csv
├── 02664021-nodes.csv
...
...
...
├── fb7f4be3-edges.csv
├── ff9bbc4d-nodes.csv
└── ff9bbc4d-edges.csv
└── iPython NB
├── Panama Project.ipynb
Explore the data
the main data provided by ICIJ can be found in CSV file, that is available in 4 languages
Main files
The CSV file is broken down to the following columns
story-title
story-subtitle
story-image-if-linked-person
story-narrative
story-category-code
story-priority
story-region-codes
story-country-1-code
story-country-1-name
story-country-2-code
story-country-2-name
story-document-1-title
data-person-1-name
data-person-1-description-if-politician
data-person-1-relationship,
data-person-1-image
data-person-1-viz-publish
data-person-1-comment
Visualization data
ICIJ report provided a set of visualizations that you can browser through this link
Data content for these visualizations is a graph structure, each graph is encoded to a JSON file, named after the data-person-1-viz-publish attribute found in the main CSV file.
Visualization nodes and edges csv data
The JSON data is normalized and the nodes and edges data is given in different csvs.
iPython Notebook
An iPyhton notebook to load and normalize the json data and play around with it.
Disclaimer
I have nothing to do with ICIJ, This is just a small project i started to see if i can get the published data, and do some cool stuff with it.
Credit
All the data are coming from the official ICIJ website
Contribute
This is a data dump, CSV and JSON files are not organized and cleaned, feel free to send a PR if you want to do that ;)