| Home | Omniscope | Download | Support | ||
| About Visokio | Partners | What's new in 2.6 | Editions | Forums | Video tutorials & reference |
| Careers | Contact us | Demos & screenshots | Clients | Knowledge Base | Featured videos |
The Sidebar Tools menu includes an option to change filtering behaviour by activating the Halo function. This function changes filtering behaviour from the default to a modified form which preserves all records which share a specific value in a specified field in the IN universe, even if the value in the filters is not one of the selected values.
The Halo function is activated from the Sidebar Tools > Halo option, by changing the common value field (column) from (None) to one of the fields in the data set. Once the Halo function is activated, a yellow caution triangle appears on top of the Sidebar telling you that filtering behaviour is now in Halo mode.
The Omniscope Halo function is a powerful analytical tool already used extensively with retail point-of-sale (POS) data. The example below is based on this file [1]. The file contains POS data on 223,484 items purchased at a number of convenience store Units/POS positions over a month, in a wide range of categories. Items sharing the same 'UniqueTransaction' number (first column text field) were bought at the same time.
On the top of the Sidebar, the Halo function has been set to include all records sharing the same value in the 'Unique Transaction' field. This means that filtering for a particular item will carry with it all the items that were bought at the same time. Notice the yellow caution triangle that has appeared on top of the Sidebar.

In this example file, the Halo function is used to select all items which were bought together with a particular promotional item (i.e. they have the same unique transaction number). In the example file, we filter the ProdName column to select the 'meal deal' promotion. With the Halo function on, filtering on 'ProdName='meal deal' returns not only the 'meal deal' transactions, but also all other 'haloed' items that share the same unique transaction numbers.
Try downloading the example file [1] and repeating the analysis for specific days of the week, time slots or outlets.
Links:
[1] http://www.visokio.com/files/Resources/OUGuide/372_HaloFunction/372HaloAnalysisExample.iok