WSCBS_Assignment4b_brane_Visualization

This implementation is the visualization part of Titanic Project (Machine Learning from Disaster).

Usage

This is the fourth package in our pipeline of project(Assignment 4b). We uses visualization package: seaborn and matplotlib to generate plots.

Here we provide with 5 visualization plots:

1. Survived Passenger Gender Distribution (gender())
2. Survived Passenger P-Class Distribution (pclass())
3. Survived Passenger Ticket fee Distribution (Ticket())
4. Survived Passenger Title Distribution (Title())
5. Correlation Heatmap including new features (Correlation())

Each package, as a part in the brane pipeline, can be run separately to produce the corresponding results (processed data, ML models, visualization)

Requirements

• Complete installation of Brane (manual1, manual2)
• Brane Dependencies (also in manual1, manual2)

Setup

By source code (Git repository)

1. Download the source code by git clone

   $ git clone https://github.com/TISNN/brane-visualization.git
   $ cd brane-visualization

2. Build brane package by .yml file

   $ brane build container.yml

3. Check availablity

   $ brane list

By brane package method

1. import brane package

   $ brane import TISNN/brane-visualization

2. Check availablity

   $ brane list

If you see visual package with version==8.0.0, it was successfully built.

Run

$ brane --debug test --data ./data visual

1. Choosing the plot you'd like to generate
2. Enter "EDA" as source string
3. It runs correctly with output "Successful! Figure saved to <>"
4. The figure will be save to ./data folder in your local file system.

Tests for package

Automated builds and running (CI/CD)

This repository is equipped with a GitHub Action workflow.

Every time we push the code to this repository, it will automatically run the tests using branescript. The build status of the project can be viewed on the Actions page.

• The brane is the executable compiled binary file, used for automated testing.
• The test.txt is the branescript used for automated testing.

Unit pytest

You also can test for a single function by python.

Parameters can be changed in file: pytest.py

$ python pytest.py