Open psorianom opened 3 years ago
Adding an EDA category would be a very good solution, I am ok with it. I believe that the pandas profilings can still be useful if someone wants to explore the dataset. Do you want me to add this category to the dgml.py code ? And to the app filters
Yes thanks @giuliasantarsieri !
When using the automatic mode of DGML, sometimes there are datasets that are not usable for machine learning (they do not pass the filters or there is some other problem with them). We still compute its pandas-profile. The problem is that they do not have a
task
so they do not appear on the DGML website. This calls for the question, what to do with these datasets ? Should we add a new task category such as "EDA" (Exploratory Data Analysis) ?