View:

• Navigate in your browser (desktop only!)
• Click "View Graphs" to see what others have researched.

Create:

• Submit your own by searching/submitting. It must be an exact case-sensitive wikipedia page!

Access:

• Access the API with a few endpoints ./api/find/uuid ./api/find/title ./api/find/all-titles

• Performant Visualization
  • Can scale to 10k+ nodes and 100k+ edges
• Hands-free Wikipedia scraping
• Easy API

1. Clone the repository

git clone https://github.com/james-salafatinos/wExplore.git

2. Change the working directory

cd wExplore

3. Install dependencies

npm install
pip install pandas numpy fa2 networkx requests Cython matplotlib

NOTE you must have Microsoft Visual Studio Build Tools installed for fa2 https://visualstudio.microsoft.com/downloads/#build-tools-for-visual-studio-2019

4. Run the app

npm start

5. Visit

 http://localhost:8080/

🌟 You are all set!

1. Clone the repository via the Cloud Shell

git clone https://github.com/james-salafatinos/wExplore.git
cd wExplore

2. Create a .env file

• Give DB_URI and ENVIRONMENT variables

3. Deploy

gcloud app deploy

Google App Engine will first look at the app.yaml file and see that this is a custom runtime. It will then use the Dockerfile to create a container to deploy to a Google managed VM. The Docker file specifies the environment for this application to run (python [including C++ build tools] and node.js.

• OII Network Visualization
• [Sigma.js]
• [MongoDB]
• [Node.js]
• [Express.js]
• [GCP]

• Implement loading screen / loading wheel after new graph submit
• Implement initial "library" of submitted nodes for the user to browse (i.e., a list of network titles (John Church) a user can initially click into and browse instantly)
• Fix ability to refresh after a new graph submittal without worrying about re-submitting
• Handle the timeout error that may happen when way too many Wikipedia API requests are called
• Implement error handling / messaging to user (i.e., That was too large of a graph, sorry.)
• Implement ability to click on node and it opens a tab / a button that appears to go to wikipedia
• Implement doc linking to mitigate MongoDB 16mb limit
• Implement option to have user control some visual parameters

• MongoDB Doc size limit
• Mobile access and usability

• Can reduce file storage size by offloading the data.json construction (color_map, size_map, data.json) to the client. Then can store flat files of |source|target|force atlas embedding|.