mohd-muzamil / irisdashboard

thesis dashboard usecase with IRIS dataset

Explorative dashboard to analyze multidimensional data visually. demo video

• Two views, i.e. Glyph View and Feature View, allow the user to interact with the dataset and select a specific data point which can then be compared with the remainder of the data points or lasso selected data points.
• The menu in the sidebar allows the user to choose a few options like type and size of glyphs, hide/show the glyph labels, and hide/show the background grid and option to apply overlap removal.
• The menu also has a search bar that allows users to search and select a specific datapoint using its ID.
• Users can also vary the k-parameter to the K Nearest Neighbours (KNN) algorithm. The cluster value derived using KNN is used to add a stroke colour over the glyph.
• The menu also allows users to choose between t-SNE and PCA dimensionality algorithms. The glyphs' positions over Glyphs-View are based on this DR algorithm.

  Steps(mac):

  Create a virtual environment in the base folder and install dependencies into venv.
  python3 -m venv venv
source ./venv/bin/activate
pip install -r requirements.txt
  To run the dashboard, execute below command from base folder path.
  python app.py