- Small changes to parameters can have major effects
- Code Review tools do not support used file formats
- Knowledge sharing is blocked by lack of tools
Data Science
Difficult To Track Changes
In the field of data science, even the smallest changes in model parameters can have a major impact. Tracking which changes lead to the final result is essential to ensure a consistent quality. Especially when working in a team, you don’t want to be surprised by changes that your colleagues have made.
Code reviews can help to keep track and share knowledge within the team, but most tools can only handle text files. Comparing the contents of more complex file formats that combine code with data or results, such as Jupyter Notebooks, is not supported.
Review Your Changes
Jupyter Notebook Support
Compare Jupyter Notebooks in code reviews! See which lines of codes were changed and how this affects the computed graphs and tables. Comment on those changes and request fixes. Review Jupyter Notebooks just like regular source code files.
Fewer Bugs
By supporting data science specific file formats we give you the chance to catch bugs early. You no longer have to wonder why your results changed after someone else edited the file. MergeBoard generates a precise change visualization as part of a merge request or for any existing git commit.
Faster Reviews
Every growing data science project will eventually have regular source code files that need to be reviewed. MergeBoard's unique change visualization will then enable an efficient review. By adding visual annotations to the code changes, developers can understand them faster and have less to remember.
Interested?
We are currently in the process of finalizing our Jupyter notebook support. You can already try out viewing .ipynb files using our public MergeBoard version, but displaying the changes between two file versions isn’t yet released.
We are looking for industry partners who are interested in trying out these features and provide us with feedback. Schedule a demo call if this sounds interesting for you.
