Description
This course is designed to introduce you to the world of professional project management with Python. Real life projects are complex and require a structured approach to be successful. This course will teach you how to organize your project, develop it in a team, version code and data at the same time, and keep it tidy and reproducible to go in production smoothly and produce value for the business. The course is designed to oil and optimize your current Data Science practices and team.
Topics include
- Introduction to a professional project management
- Professional project organization for code structuring, commenting, and documentation.
- Git for version control, collaboration, and effective code management.
- Git-Flow methodology for a structured development process.
- Web API to retrieve data from the internet and to exploit public datasets.
- Scientific and reproducible Reports (Using LaTeX formulas)
- Interactive Reports with Jupyter notebook
- Scientific presentations with Jupyter notebook
- Use Docker to create applications and to deploy them
- Introduction to FastAPI for creating APIs
- Model Deployment
- Automate the build and deployment process with Gitlab CI/CD
- Git basics and Git flow
- issues & merge requests - Gitlab
- Package registry - Gitlab
- Container registry - Gitlab
- Sprint planning - Gitlab
- DVC: Data Version Control to track data changes and to reproduce experiments
- Reproducible model training
- Store big and common datasets on GDrive or S3 storages with DVC
- DVC pipelines, metrics, experiments to automate the data process
- Reproducible data process with DVC: version code and data at the same time (for data analysis, ML model training, …).
- Poetry scripts and new CLI commands
- studio.iterative.ai
- Confluence and Collaborative documentations
- Git/Jira Boards
What you will be able to do
- Organize your project in a professional way
- Cooperate with your team in an efficient way
- Version code and data at the same time
- Create the right presentation or report for every situation
- Use Gitlab for issues, merge requests, package registry, container registry, and sprint planning
- Use DVC to track data changes and to reproduce experiments
- Create and automate the build of reports
- Store big and common datasets on GDrive or S3 storages with DVC
- Automate the data process with DVC pipelines, metrics, and experiments
Duration
3 days
Prerequisites
Ability to use Python for data analysis.
Audience
This is an advanced course for professionals and team leaders who need to find the best way to manage data science projects. It is also suggested for those who has experience with IT and programming and want to improve their skills in data science.