Release Notes

New Functionality

  • APIs - Full JSON API endpoints for all capabilities, which allow complete integration with downstream reporting and visualization tools
  • Data Profile - Display of Outliers, Histograms, Descriptive Statistics, and Data Types
  • Auto Sampling - Automatic detection of a statistically relevant sample size for large tables
  • Optionally Skip Blank Values - In some use cases, blank values have no meaning, therefore the user is allowed to ignore these non-values during analysis

Enhancements to Existing Functionality

  • Collections - The user can now build groups of tables called Collections and view all Scans inside a Collection
  • Performance Improvements - Scans can utilize multiple CPU processors and the construction of databases have been optimized for higher I/O speeds

Compatible Browsers

  • Chrome 51.0+
  • FireFox 47.0+
  • Safari 9.1+
  • Internet Explorer 10, 11

New Functionality

  •   Several pre-scan features have been added:
    • Users are now able to preview the data prior to performing a scan.
    • While previewing the data, Users are able to instruct EmcienScan to ignore specific columns in the data.  Ignored columns are not part of the results.
    • Users have the ability to select the data file encoding and delimiter or will auto-select
  • The ability to "one-click" create multiple scans by selecting multiple tables and/or views at once from a single data source.  
    • A name can be assigned to multiple scans for use as a prefix to each scan in the collection
    • The configured sample size is applied to each scan
  • To give more detail on what is taking place during the data scan, a scan log page has been added 
    • The log is available for viewing as soon as the scan is initiated 

Enhancements to Existing Functionality

  • Now able to analyze columns with 1000+ unique values. In the previous version these columns were skipped during the scanning, and displayed as "Difficult to Predict"
  • Enhancements that provide more granular details about column data:  
    • Column Data Types have been extended.  Columns are now classified as one of the following:
      • Categorical
      • Integer
      • Real Values
      • Mixed
      • Mostly Integer
      • Mostly Real Values
      • Mostly Categorical
      • Empty
    • Tags have been implemented to identify:
      • Columns that perfectly predict another (either uni-directional or bi-directional locksteps)  PD
      • Columns that are or are part of a lockstep group RC
      • Columns that contain unique or nearly unique data UV, MU

Defects Resolved

  • EmcienScan fails when data does not contain predictive columns.
  • Users are not automatically logged out after the inactivity period expires when application left open on Home Page

Compatible Browsers

  • Chrome -  47.0.2526.106  and   48.0.2564.109
  • FireFox - versions 43.0.4  and  44.0.2
  • Safari - versions 8.0.8   and   9.0.2
  • IE 11

New Functionality

  • EmcienScan can now connect to a variety of databases/data warehouse sources via ODBC
  • EmcienScan will separate the weakly correlated columns for a dataset from the columns with correlations
  • EmcienScan will alert the user when there are duplicate or redundant columns by using the lock badge

Enhancements to Existing Functionality

  • There is no limit on the number of rows a data source can have for a scan
  • The scanning animation will show several steps (extraction, scanning, joining)
  • .csv downloads are available for numerical EmcienScan output
  • Sample data is organized by predictability

Defects Resolved

  • Failures are more elegant with better error messages

Compatible Browsers

  • Chrome -  46.0.2490.80
  • FireFox - version 42.0
  • Safari - version 8.0.8
  • IE 11