NOTE: Some UI elements in the Dataset Management Experience have been updated since 12/7
Georeference columns can be configured from a dataset draft. If this is for an existing dataset, click 'Edit' from the action bar to get started.
From the data preview page, open the Add Georeference tab from the sidebar on the left.
The geocoding toll will accept 3 different formats for creating a georeference column.
Latitude and Longitude values from two columns in the dataset.
Address information that is split up into street address, city, state and zipcode.
Combined values like latitude-longitude or full street address.
NOTE: After selecting the method to generate a georeferenced column and filling in the required field, Run Geocoder can be selected to preview the geocoded values.
When done, click on Add Column,
Remapping your georeference column
If for any reason the mapping of your georeference column is not stored in the revision (this can occur if there is a change in your source schema or if you are editing your dataset for the first time with the new publishing tool), you will have the option to remap your columns.
When this happens, you will first see a Column Mismatch warning like below:
From here you can select Other Fix Options > Configure My Location Column. This will open the Georeference modal as described above, where you can set up your location column based on lat/long, address, or a combined location column!
Adding a georeference column using Transforms
Georeference columns can also be created using the data transformer tool. To do so, use the Add Column button from the sidebar to get started. From the new column, open the column drop down menu and select Data Transforms.
From the transforms editor, you can specify which column values will be used to create the new georeference column. For this example, Latitude and Longitude data was used:
Reference text: to_point('POINT ('||`longitude`||' '||`latitude`||')')
The to_point() function converts text values to create a georeferenced point. In this dataset the longitude and latitude values are split into 2 columns, so in this transform the values are combined using the || transform to combine multiple strings.
The make_point() function can be used to create georeferenced points as well if the column values are already stored in a number data type. The example below converts text format lat/long to numbers.
Reference text: make_point(to_number(`latitude`),to_number(`longitude`))