Last year, Microsoft announced the Team Data Science Process (TDSP), an agile, iterative, data science methodology to and a set of practices for collaborative data science. TDSP helps organizations structure their data science projects by providing a standardized set of Git repositories, document templates and utilities that are relevant at different stages of their development lifecycle. Microsoft recently released TDSP Data Science Utilities version 0.11 with several new features and enhancements.
- To better support the data science community, we are now releasing IDEAR in Jupyter Notebooks (Python 2.7). Data scientists who prefer Python can now explore and visualize data using similar functionality as what IDEAR had earlier provided in R. Users can upload the IDEAR Jupyter Notebook to a Jupyter Notebook server, configure the working directory in the Jupyter Notebook, and start investigating data sets.
- For data scientists who prefer Visual Studio with RTVS as their data science IDE, we now have the option to run IDEAR in R in Visual Studio. You can do so simply by changing the option “Shiny pages browser” to External in Visual Studio.
- Data scientists usually write code to extract datetime components such as year, month, weekday, week of year or hour, and can then use them as extra variables for further analysis and modeling. This new feature of IDEAR in R extracts these datetime components automatically and adds them directly to the original dataset, with column names ending with _autogen_year, _autogen_month, etc. IDEAR in R works with this enhanced dataset, allowing data scientists to visualize and obtain insights on how these date and time components impact the target variable.
Read more about this release here.