After assigning columns to categories, you may need to map the values in the original column (such as man, woman) to a value in the Global Schema (in this case Male and Female).
The mappings tool in the Bunker's web UI can help you fix a large category of mismatches between your data and the Global Schema. It can also help you clean up messy data, by converting several different original values into one value.
Before starting, to understand the kind of problems mappings can help solve, let's look at an example. Here's a screenshot of a few rows of data which have been assigned to the Gender category.
First, let's understand what this screenshot is showing you.
- The first column, Gender, contains your original data.
- The other two columns, Binary and Other and Extended Gender, are representations of your data. A representation is a way of presenting your data in a specific form, making it easier to match it against your collaborator's dataset.
- A single piece of data can have more than one representation. In this example, the Binary and Other representation nets down gender into three options: Male, Female and Other. The Extended Gender representation contains a much wider range of options, reflecting the many different gender identities the individuals represented in your dataset may have.
Now, we can see there are a couple of problems with the way your Bunker has interpreted your data.
- Where your data says male, everything is fine. Both the representations correctly contain the value Male.
- Where your data says f, we can safely assume the individual concerned is female. However, your Bunker hasn't been able to interpret this data without your help. So, both the representation columns contain a red warning symbol.
- Finally, where your data says Bi-gender, your Bunker has successfully mapped this to the Extended Gender representation - but hasn't known how to interpret it for the Binary and Other representation.
These problems can be solved using mappings.
Setting up mappings
To set up mappings for a category, click the settings icon above your original data column, and select Mappings from the drop-down.
If you don't see the Mappings option, this means the category you've selected doesn't support mappings. Mappings are only suitable for categories which, like Gender, contain values selected from a pre-defined list of options.
The Mappings dialog will appear. It looks like this:
To find your way around the Mappings dialog, remember the following.
- The dialog has a tab for each representation. So in this example, there are two tabs, for each of the Binary and Other and Extended Gender representations.
- On the left-hand side of the dialog, you'll see a list of the values which appear in your data (most common first).
- On the right-hand side, you'll see the allowable values for the selected representation.
To set up a mapping, just turn on the appropriate checkboxes. For example, select the f box in your original data and the Female box as the category representation..
You can select several checkboxes on the left-hand side, but you can only select one checkbox on the right-hand side. Every value in your original data maps to one value in the representation.
Once you've selected the appropriate checkboxes, click the Map button to add the mapping to your configuration. Carry on doing this until all your values are mapped.
At any time, you can click the bar labelled Mapped to review the mappings you've set up.
Once you're happy (remember to check the tab for each representation!), click Save to close the dialog and save your configuration.
Completing our example, all our data is now mapped to an appropriate value for both representations, and the red warning messages have disappeared.
If you can't fix your data using mappings alone, the transformation tools might help. If you have successfully mapped your dataset, the next step is to normalise and publish your dataset.