Release Notes v4.1

New

Elastic platform

  • There is an elastic platform of Hypercube that enable the use of Hypercube by a large number of users.
  • This is based on containers technologies with Docker & Amazon ECS.

Discretization

  • Two new kinds of discretization are available:
    • one by equal width,
    • one can be handcrafted.
  • There is a new page to manage it:
discretization

discretization

  • Those custom discretizations can be used for Xray, Learning and Prediction.

Demo projects

  • You can create a Titanic project demo that create for you analysis for Titanic dataset.

Chat

  • You can now chat with HyperCube team.
chat

chat

Localization

  • You can choose the language in setting page.

Improvements

Dataset grid view

  • You can now view easily large dataset on 'View Data' page
dataVIew

dataVIew

Xray

  • You can now:
    • click 'load more' button when there are a lots of variables
    • sort variables
xray

xray

UI

  • You can now see all your data visualizations in a new page

  • Visual enhancements of the key performance indicators page
  • You can edit the key indicators colors for target and description key

Fixed Bugs

  • User:
    • User connection with LDAP, groups correction
  • Dataset:
    • Automatic column separator detection for upload dataset
    • Import Dataset keeps now Datetime Metadata Metatype
  • Datagrid:
    • Visual corrections of view data page
  • Correlation:
    • In Correlation, filter variables by tags is now possible
    • The mutual information has been momentarily disabled
  • Visualization:
    • "Group by" visualization in BarChart is not mandatory
  • XRay:
    • In Xray Target has to be in dataset
  • NLP:
    • It's now impossible to launch a text analysis visualization after an error or when no dataset is selected
  • Ruleset:
    • Ruleset minimization had wrong cumulative coverage and ruleset visualization correction
    • Cannot minimize a ruleset by average value
    • Learning couldn't be deleted when in progress
  • Model:
    • Exporting scores from model
    • Models' chart lines in view model
  • UI:
    • Display Dark-Blue theme