Introduction

EyeOpenR Insight offers a wide range of tools for filtering and sub-setting your data (see also Knowledge Base articles on Groups, plus AI Tools [Filter Assistant]). In EO Insight there are two broad categories of filters, those that are applied to the entire Board, which are referred to as ‘Global’ filters, and those that are only applied to a single widget, which are referred to ‘Widget’ filters. Below you will find explanations of each of these types of filters and how they can be applied and tracked in EO Insight.

Global Filters

Global filters filter the data accessed by all widgets on your board, other than widgets that have a ‘widget’ filter enabled (see ‘Widget’ filters below). Therefore when you apply a global filter, all widgets on the board (other than those with widget filters enabled) will update their output in response to the changes in the newly filtered dataset.

NOTE: AI Widgets created by the AI Widget Generator are dependent on the data available at the time they were generated and do not therefore update when global filters are adjusted. Should you wish to update the AI Widget you can create a new AI Widget with the AI Widget Generator and if required delete the existing AI widget (see AI Tools Knowledge Base article for further information). 

RECOMMENDATION: As all widgets on your board update each time a global filter is adjusted, for large datasets or boards containing many widgets we recommend applying global filters before adding widgets to your board. This will avoid unnecessary time spent recalculating your results. 

How to apply simple filters to the entire Board

Step 1: Click on the ‘Filters’ logo in the sidebar to reveal the ‘Filters’ sidebar.

Step 2: Click on the ‘Filters’ tab at the top of the ‘Filters’ sidebar to view the ‘Global Filter’ menus

Step 3: Click on the drop down for the property you wish to filter i.e. Attribute, Panelist, Product, Session, Sequence or Replica

Step 4: Within the given property menu you will find a drop down list containing all values in your dataset currently available to filter. Select any combination of these values in the list that you wish to remain in the dataset available to the board.

Step 5: Click the dark blue ‘Filter…’ button to apply the filter to your board

How to combine constraints on Attributes to filter the entire Board

The ‘Attribute Value Filters’ functionality in EO Insight allows you to combine constraints on multiple attributes/questions to create a subset of your data. Below are the steps you can take to apply such filters to your entire board.

Step 1: Click on the ‘Filters’ logo in the sidebar to reveal the ‘Filters’ sidebar.

Step 2: Click on the ‘Filters’ tab at the top of the ‘Filters’ sidebar to view the ‘Global Filter’ menus

Step 3: Click on the drop down for the ‘Global Filters for Attribute Values’

Step 4: Within the drop down labelled ‘Choose Attribute(s) to Filter’ you will find a list containing all attributes/questions in your dataset currently available for use in a ‘Value Filter’. Select any combination of these attributes in the list to which you would like to apply constraints to subset your data.

Step 5: Click the dark blue ‘Confirm Attribute(s)’ button to confirm your selection

Step 6: Within the ‘Global Filters for Attribute Values’ panel you should now see options that allow you to define the constraints you wish to apply to your selected attributes (8).

Step 7: Use the AND/OR toggle switch to determine whether the constraints your define here should be combined with an AND or an OR operator 

Step 8: For each of your selected attributes you can define which properties they must possess to remain in the dataset. For example, in the below image these settings will ensure only rows in the dataset in which Region = GBR ‘AND’ Age is in the range 0 - 50 remain in the dataset accessible to the board.

Step 9: Once you have defined the constraints you wish to apply to all variables click the ‘Apply’ button to apply this filter to your dataset. Your board will then update accordingly, with any widgets present on the board updating their output based on this newly filtered dataset.

Where to find information about which Global Filters are applied

When Global filters are applied to your Board you find out which Global filters are applied in the ‘Overview’ tab of the ‘Filters’ sidebar. How to access and interpret this information is explained below.

Step 1: Click on the ‘Filters’ logo in the sidebar to reveal the ‘Filters’ sidebar.

Step 2: Click on the ‘Overview’ tab at the top of the ‘Filters’ sidebar to view the ‘Overview’ information

Step 3: The ‘Active Data Metrics’ panel shows which properties of your board have been impacted by the current filters. If a metric panel is coloured ‘RED’ this means that it’s value has changed relative to the original dataset (e.g. in the screenshot below both the number of panelists and constraints on attributes has changed due to the global filters currently applied). If a panel remains ‘Blue’ this means it has not been affected by the current filters and its value remains the same as for the original dataset. To see all unique values for a given variable that are currently available to your board (after filters are applied) hover over or click the metric box and a list of unique values will be displayed.

Step 4: In the ‘Overview’ panel below the ‘Active Data Metrics’ you can find lists of the values excluded by the currently active filters. When ‘Detailed Attribute Filters’ are applied this will be registered in this section, should you wish to know more about the detailed attribute filters you can go to the ‘Active Detailed Global Filters’ panel of the ‘Filters’ tab to see which constraints are currently applied to which attributes globally. 

How to reset Global Filters

If you wish to remove your Global filters and restore the original dataset follow the steps below.

Step 1: Click on the ‘Filters’ logo in the sidebar to reveal the ‘Filters’ sidebar.

Step 2: Click on the ‘Overview’ tab at the top of the ‘Filters’ sidebar to view the ‘Overview’ information & options

Step 3: Click on the ‘Remove all Global Filters & Groups’ section title to reveal the options

Step 4: Click the ‘Restore’ button to restore the original dataset with no filters or groups. Your board will update in response to the newly restored dataset.

Widget Filters

Widget filters allow you to adjust, at the level of an individual widget, the dataset the widget uses to generate its output. Therefore you can decide for each widget on your board exactly what information (subset of your dataset) it can access and therefore use to generate its results.

Within each widget you can choose to simply adjust which variables it sends to its given analysis method (i.e. using the ‘Data for Analysis’ function) or enable the ‘Widget Filters & Groups’ which makes the widget independent of any ‘Global Groups and Filters’ and therefore you can adjust the subset of your original dataset you wish the widget to access or define groups that are only applied to the given widget.

RECOMMENDATION: For large datasets or boards containing many widgets we recommend applying comprehensive global filters to ensure only minimal data is available to the entire Board before adding widgets to your Board in order to avoid unnecessary lengthy calculations. Individual widgets that require access to larger portions of your dataset can have widget level filters applied to perform analyses on this larger dataset.

How to choose which variables are sent for Analysis

The ‘Data for Analysis’ function within each widget allows you to control which data is sent to the widget’s analysis function, therefore further subsetting data following any widget or global filters. In other words the ‘Data for Analysis’ function allows the user to choose which sub-set of attributes is sent to the Widget analysis method, without impacting the set of attributes available to the widget for other purposes. For example:

• Subset the attributes in the dataset currently output from the global filters & clusters that are then sent for analysis by the widget method. If by contrast you activate the ‘Widget Filters & Clusters’ then the dataset used by the widget is a clean version of the original dataset (i.e. no influence of global filters & clusters).

• Subset the attributes in the dataset currently output from the widget level filters & clusters that are then sent for analysis by the widget method. This way you can still have a widget level filter on ‘gender’ for example but do not need to send the ‘gender’ variable for analysis.

The ‘Data for Analysis’ function can be applied by following the below steps:

Step 1: Click on the drop down button in the top right corner of the widget to reveal the within widget menu

Step 2: Click on the ‘Filters & Groups’ option to access the within Widget Filter and Grouping settings

Step 3: The widget should automatically select the ‘Data for Analysis’ tab when entering the within Widget Filter and Grouping settings

Step 4: Choose from the drop down list the variables you wish to send for analysis

Step 5: Click on the ‘Submit for Analysis’ option to initiate the widgets analysis function with this new subset of your data

How to apply simple filters to a single Widget

Step 1: Click on the drop down button in the top right corner of the widget to reveal the within widget menu

Step 2: Click on the ‘Filters & Groups’ option to access the within Widget Filter and Grouping settings

Step 3: Select the ‘Activate Widget Filters & Groups’ tab from the menu on the left side of the widget

Step 4: Toggle the switch to ON to enable the widget level Filters & Group options. This will release the widget from any Global filters and groups so that it has access to the original dataset.

Step 5: Click on the ‘Add Widget Filter’ tab in the left hand menu to view the ‘Add Widget Filter’ options

Step 6: Click on the drop down for the property you wish to filter i.e. Attribute, Panelist, Product, Session, Sequence or Replica

Step 7: Within the given property menu you will find a drop down list containing all values in your dataset currently available to filter. Select any combination of these values in the list that you wish to remain in the dataset available to the widget.

Step 8: Click the dark blue ‘Filter…’ button to apply the filter to your widget

How to combine constraints on Attributes to filter a single Widget

The ‘Widget Attribute Content Filters’ functionality allows you to combine constraints on multiple attributes/questions to create a subset of your dataset for a given widget. Below are the steps you can take to apply such filters to a single widget.

As explained previously and shown again in the screenshot below you first need to ensure that the ‘Widget Filters & Groups’ option is enabled for the given widget:

If the ‘Widget Filters & Groups’ option is enabled follow the below steps:

Step 1: Click on the ‘Widget Attribute Value Filters’ tab in the left hand menu to view the ‘Widget Attribute Value Filters’ options

Step 2: Within the drop down labelled ‘Choose Attributes to Filter’ you will find a list containing all attributes/questions in your dataset currently available for use in a ‘Value Filter’. Select any combination of these attributes in the list to which you would like to apply constraints to subset your data.

Step 3: Click the dark blue ‘Confirm Attribute(s)’ button to confirm your selection

Step 4: Click on the ‘Widget Attribute Value Filters’ tab in the left hand menu to view the ‘Widget Attribute Value Filters’ options This will reveal the section in which you can define the constraints you wish to apply to your selected attributes.

Step 5: Use the AND/OR toggle switch to determine whether the constraints your define here should be combined with an AND or an OR operator 

Step 6: For each of your selected attributes you can define which properties they must possess to remain in the dataset. For example, in the below image these settings will ensure only rows in the dataset in which Region = GBR ‘AND’ Age is in the range 0 - 50 remain in the dataset accessible to the widget.

Step 7: Once you have defined the constraints you wish to apply to all variables click the ‘Apply’ button to apply this filter to your dataset. Your widget will then update its output based on this newly filtered dataset.

Where to find information about which Widget Filters are applied

When Widget filters are applied you can find out which filters are applied on the ‘Active Widget Data Metrics’ tab. How to access and interpret this information is explained below.

Step 1: Click on the ‘Active Widget Data Metrics’ option in the left side menu to view the ‘Active Widget Data Metrics’ tab.

Step 2: The ‘Widget Data Metrics’ panel shows which properties of the dataset accessible to the widget have been impacted by the current widget filters. If a metric panel is coloured ‘RED’ this means that its value has changed relative to the original dataset (e.g. in the screenshot below both the number of panelists and constraints on attributes has changed due to the widget filters currently applied). If a panel remains ‘Blue’ this means it has not been affected by the current filters and its value/number of observations remains the same as for the original dataset. To see a list of all unique values currently available to the widget for a given variable hover over or click on the corresponding metric box.

Step 3: In the panel below the ‘Active Data Metrics’ you can find lists of the values excluded by the currently active filters. When ‘Detailed Attribute Filters’ are applied this will be registered in this section, should you wish to know more about the detailed attribute filters you can go to the ‘Active Detailed Widget Filters’ tab to see which constraints are currently applied to which attributes within the widget. 

How to reset Widget Filters

If you wish to remove your Widget filters and restore the widgets dependency on the Global dataset (i.e. incl Global Filters & Groups) follow the steps below.

Step 1: Click on the ‘Reset Widget’ option in the left side menu

Step 2: Toggle the ‘Widget Filters & Groups’ switch to OFF. The widget will update its output in response to the newly restored dataset.