Organize your Dataflow columns and attributes
Discover how you can organize data with the help of the advanced features in Dataflow in Productsup.
Introduction
Dataflow offers several advanced features that help you adjust this page for comfortable work. You can access these features from the upper ribbon above the Dataflow columns:
The features let you perform the following actions. Select the links to go directly to the description of the necessary feature:
Channel selection — This drop-down menu lets you select a destination export channel for your product data.
Preview — This button unfolds a column on the left-hand side of the Dataflow columns and lets you see the values of the imported attributes.
Filter Columns — This button unfolds a section with search fields and lets you filter by specific attributes in the import, intermediate, and export columns.
Sort by — This drop-down menu lets you specify the sorting order for attributes in the import, intermediate, and export columns.
Hide/Show — This drop-down menu lets you set the Dataflow view, hiding or revealing specific columns and attributes. You can also select a custom view pre-defined in Data View.
Automap connections — This drop-down menu lets you choose among the options for automapping the attributes between the columns.
Delete — This drop-down menu lets you remove rule boxes, custom attributes, or connections between attributes in bulk.
Versions — This drop-down menu keeps all versions of your Dataflow and lets you retrieve a necessary version in case of a problem.
Filter Boxes — This search field lets you look up certain rule boxes applied to your attributes while hiding those that don't match the search keywords.
Rule Box Conditions — This feature lets you create and set up conditions to determine the cases when the platform should apply rule boxes to an attribute. See Apply rule boxes based on conditions.
The following sections describe each feature in detail.
Select a channel
The Channel selection button lets you select a destination export channel for your product data. You can see only the channels you have activated in Exports. See Map your product data for more information about channels.
To select a channel:
Go to Dataflow from the site's main menu.
Select the Channel selection button on the upper ribbon.
Select the export channel to which you want to export your data.
Preview import data
The Preview button gives you a quick preview of the values of the imported data and helps you map your attributes more accurately.
To activate Preview:
Go to Dataflow from the site's main menu.
Select the Preview button on the upper ribbon. An additional section with the attribute's values opens.
Scroll through the products using arrows to see different values in the attributes.
If your Import column contains many attributes destined for different export channels, it can be challenging to map attributes. In this case, you can filter by an attribute's value to understand what each attribute means.
To filter by attribute's value:
Activate the Preview mode. See Step 2 of the previous section.
Select the magnifier icon next to the necessary value.
In the Attribute value search pop-up window, enter a value for an attribute in the Search value field.
Select the Search button.
To hide the Preview section in the Dataflow:
Select the Preview button on the upper ribbon again.
Filter by column attributes
The Filter Columns button lets you find a specific attribute in the Import, Intermediate, or Export channel columns.
To filter by a specific attribute in a column:
Go to Dataflow from the site's main menu.
Select the Filter Columns button on the upper ribbon.
In the opened section, enter the attribute you want to filter by in the search field of the specific column.
Table 1. Example of the Filter Column feature work:Before
After
Tip
You can set up filters for several columns at a time.
Sort by column attributes
The Sort by button lets you define the order of attributes to appear in Dataflow. It doesn't affect the order of the attributes in the export. It only customizes your view in Dataflow.
To access sorting options for attributes:
Go to Dataflow from the site's main menu.
Select the Sort by button on the upper ribbon.
Choose your option from the drop-down list:
Connection - Connected attributes appear first.
Column Name - Attributes appear in alphabetical order, from A to Z.
Import Order - Attributes appear in the Import column in the order of their import.
Import Order and Connection - Attributes appear in the import column in the order of their import and the connected attributes come first.
Export Order - Attributes appear in the export column in the order they appear in the export file.
Raw Import & Export Order - Attributes appear in the import and Export columns in the order they appear in the import and export files.
Hide or show column attributes
The Hide/Show button lets you customize the Dataflow view by hiding irrelevant information to be able to focus on specific columns or attributes:
To hide or reveal attribute data:
Go to Dataflow from the site's main menu.
Select the Hide/Show button on the upper ribbon.
Select a type of data that you want to hide or show. Choose one of these options to have the following effect:
Hide Unconnected Columns - Hides all unmapped column attributes.
Hide Connected Columns - Hides all mapped column attributes.
Hide Import - Hides the entire import column.
Hide Export - Hides the entire selected export column.
Hide Warning Connections - Hides any red warning connections generated after importing data from another site.
Note
A warning connection is a connection between two columns in Dataflow, highlighted in red.
A warning connection appears when there is an active connection between the attributes in the import and intermediate columns, but the connected import attribute doesn't exist in the import feed. You can also see a warning connection if the rule box applied to this connection doesn't have the required column entry.
Hide Attributes without Values - Hides all attributes that don't have values.
Display Attributes from Datasources - Lets you see only the attributes from the selected data sources.
Column View - Lets you select a custom Dataflow view previously created in Data View.
Attribute Width - Lets you regulate the width of the columns to let you fully see the attribute names.
Map attributes with Automapper
You can connect attributes between stages one by one, or you can use the Automapper feature and save time. Automapper connects attributes in seconds using the exact name matching, dictionary, or artificial intelligence suggestions.
You can map attributes from import to intermediate and from intermediate to export.
Warning
Before using Automapper, we recommend saving your current version of Dataflow so that you can roll back to it in case of any issue. See Manage versions in Dataflow.
To use Automapper:
Go to Dataflow from the site's main menu.
Select Automap connections in the upper ribbon.
Choose an option from the drop-down menu:
Import to intermediate – Automaps the attributes between the import and intermediate stages.
Intermediate to export (all intermediate attributes) – Automaps all intermediate attributes between the intermediate and export stages.
Intermediate to export (only attributes connected to import) – Selects only the attributes mapped to import and automaps them between the intermediate and export stages.
Map all import attributes to export – Automaps the attributes from the import to export stage. If the names of the attributes don't match, the platform creates custom attributes at the relevant stage. The attribute order doesn't change in your export file, although the platform arranges the attributes alphabetically in your Dataflow view.
If you select one of the Automapper options, the AUTOMAPPER pop-up window opens to let you specify your mapping possibilities.
In step 1, select the automapping method:
Direct matches – The platform maps attributes that are exactly the same in both stages. The matching is case-insensitive. For example, ProductTitle may match to producttitle.
The drop-down has the following options:
Yes means that the direct matching is active. It is the default selection.
No means that the platform ignores the direct matches and uses other automapping methods.
Dictionary matches – The platform maps the attributes using our extensive internal dictionary.
The drop-down has the following options:
Yes means that the dictionary matching is active. It is the default selection.
When set to Yes, the Confidence scale becomes active and lets you change the matching accuracy. Set lower confidence to get more matches that are less accurate or set higher confidence to get fewer but more exact matches. By default, Confidence is 80%. You can see the results in the next step.
No means that the platform ignores the dictionary matches and uses other automapping methods.
Predicted matches – The platform maps the attributes with the help of an artificial intelligence model. When selected, this method matches only the attributes that remained unmapped after the direct or dictionary matching.
The drop-down has the following options:
Yes means that the predicted matching is active.
If you select Yes, the Confidence scale becomes active and lets you change the matching accuracy. Set lower confidence to get more matches that are less accurate or set higher confidence to get fewer but more exact matches. By default, Confidence is 80%. You can see the results in the next step.
No means that the platform ignores prediction matches and uses other automapping methods. It is the default selection.
In step 2, you can see all mapping suggestions from the previously selected automapping methods. By default, the platform marks all the suggestions as selected, but you can deselect those that you think are irrelevant. When you are ready, select Apply and finish to confirm the mapping.
Delete attributes, rule boxes, or connections
The Delete button lets you bulk delete rule boxes, custom attributes, or connections between attributes.
To delete multiple attributes, rule boxes, or attribute connections:
Go to Dataflow from the site's main menu.
Select the Delete button on the upper ribbon and select the option you want to apply from the drop-down menu.
There are several different options in the Delete drop-down menu, and the following table explains what these options do:
Section | Menu item | Explanation |
|---|---|---|
Connections and Boxes | Import-Intermediate | Select this option to delete all Import-Intermediate attribute connections. |
Intermediate-Export | This option deletes all Intermediate-Export attribute connections. | |
All | This option lets you delete all attribute connections across all columns. | |
Rule boxes | Import-Intermediate | Delete all rule boxes for Import-Intermediate attributes by selecting this option. |
Intermediate-Export | Select this option to delete all rule boxes for Intermediate-Export attributes. | |
All | Delete all rule boxes for all attributes across all columns by selecting this option. | |
Connections | Connections missing Import input | This option lets you delete all non-functioning connections, also called warning connections, generated after importing data from an additional source. |
Custom Columns | Remove Unconnected (Intermediate) | Select this option to delete all unconnected Intermediate attributes created as custom attributes. |
Remove Unconnected (Export) | Select this option to delete all unconnected export channel attributes created as custom attributes. |
Manage versions in Dataflow
As you edit and adjust elements in your Dataflow, it is helpful to save versions of your Dataflow before testing something or making major changes so that you can roll back to earlier versions in case of a problem.
Dataflow automatically saves versions before you commit significant changes, such as deleting all column connections. But you can also save versions manually.
You can see the list of versions by selecting the button on the upper ribbon.
To save a new version of Dataflow:
Go to Dataflow from the site's main menu.
Select the Versions button.
Select + Save Dataflow as New Version.
In the pop-up menu, enter the version's name and leave a comment (optional).
Select Save. The Versions menu now includes your just saved version.
There are several icons to the right of each saved version.
These highlighted icons let you perform the following actions:
View a comment.
Edit a version name or comment.
Delete a version.
To retrieve a version of Dataflow:
Select the Versions button.
Select the version you wish to retrieve.
Select Yes if you want to save your current Dataflow version before loading a previous version.
Note
You can not edit an older version - you can only save a new one. Loading a previous Dataflow version overwrites the current version. If you continue without saving, you will lose all changes for all export channels in the current site since the last saved version.
Filter rule boxes
The Filter Boxes search field lets you look up certain rule boxes applied to your data while hiding those that don't match the search keywords.
To filter by rule boxes:
Go to Dataflow from the site's main menu.
Enter the name of the rule box you want to search for or a keyword associated with it to the Filter Boxes search field.
The platform hides all non-matching rule boxes from Dataflow, showing only the rule boxes that match your search term.