Sections in this category

Transforming a Column: Binning Numerical Columns

  • Updated

Sisu automatically applies binning for numerical columns so the values are grouped into categories that can be used as a Sisu factor. By default, these categories are in quartiles:  binned on the 0-25th percentile, 26-50th percentile, 51-75th percentile, and 76th to 100th percentile.

This quartile binning default setting can be changed when desired.

 

Related Articles:



Changing the Binning Percentage for a Column

  1. Go to the Projects module, and select the Project and Analysis that contains the columns you want to manage.
  2. Click Manage Columns, and select a numeric data type column in the list to display its settings.

    Select_Numeric_Column_for_Binning.png

  3. Select the Bin type desired.

    Select_Bin_Type.png

    The default bin type is “Automatically bin”, which uses the quartile binning described at the beginning of this article. There are two additional options, described below.

    Width of bins

    Change the width of the bin (i.e., the range of values in a single bin).

    Bin_Type_Width.png

    In the above example, the width (range) of each bin will be 20, which means the ranges will be 0-20, 21-40, 41-50, etc. If you enter 500, the bin sizes will be 0-500, 501-1000, etc.). 

    Number of bins

    Change the number of bins (i.e., how many bins there are in total).

    Bin_Type_Number.png

    By default, Sisu breaks the data into quartiles, meaning there are 4 bins. In the example above, there will be 10 bins instead of the default 4. 


  4. Click Save and re-run your Analysis, if desired.