Currently Developing!
This module aims to acquire feature relathionship while EDA with graph representation.
You can check sample notebook
- Each node (dots) refers to each feature.
- larger node size means that feature has more statistically meaningful relationship.
- Each edge (lines) indicates statistically meaningful (p < 0.05) relationships.
- Each of the red and blue colors is positive and negative relationships.
- higher saturation (more bold color) means stronger relationships
- Correlation-based Graph (without direction)
- pandas
- numpy
- networkx
- matplotlib
- Directed Graph with Regression
- Bipartite Graph for Group Difference