Datasets are mostly about visualizing and exploring aggregated data. For users with the correct permissions, however, it can also be helpful to view the details of individual records. That’s what’s covered in this section.

There are also special circumstances when you’ll want to override dataset values that you’ve imported, or create entirely new records that don’t exist in your data source. It’s almost always better to change the data at the source, so overriding values in Impact should only be used when that’s not possible.

The Records tab

On the Datasets Explore tab, we’ve applied an “Age is less than 18” filter and there are 331 matching results.

With the correct permissions, I can click to the Records tab to see the raw data.

Exporting records

To download the records on the Records tab, click the Send To button and choose “CSV” or “Excel”.

CSV is the best choice for large datasets. CSV files are built on the server rather than in your browser, so even very large datasets export without trouble, and they include every record that matches your filters. Small files download right away. For bigger ones, Spider Impact sends you an alert when the file is ready. CSV files contain plain values without formatting, which makes them easy to import again.

Excel exports are built in your browser, so they work best for smaller amounts of data. If an Excel export is large, Spider Impact warns you and recommends CSV instead, though you can still continue with Excel. If an export is too large for Excel, Spider Impact asks you to use CSV.

Editing record values

By default, records’ values can’t be changed in datasets. When editing a field, however, you can allow values to be overridden, either on the Datasets Records tab or with a form.

When one or more fields are editable and you have the correct permissions, you can click on a record on the Records tab.

This shows a dialog where you can override record values. Datasets will remember the values you enter, even after new data is imported into your dataset.

Note that you can only make a field editable when you have a primary key defined for your dataset. That’s because the software needs to be able to uniquely identify the record that is being changed so that the same change will be made after you import data again later.

Adding new records

When editing a dataset, you can allow entirely new records to be added.

When enabled, there is an Add button on the Records tab.

Just like editing records, added records will be preserved even after future data is imported into the dataset.

Imports never overwrite records you've added by hand, so an import can't use a Record ID that already belongs to one of them. When you upload a spreadsheet to a dataset, either on its Edit tab or as a new revision of a file in the Files section, Spider Impact checks the file before uploading it. If any Record IDs match records that were added manually, the upload stops and lists them so you can fix the file and upload it again.

Some matches can only be found while the dataset builds, for example when the Record ID field is a rollup or geographic field. Those rows are skipped, the rest of the file imports normally, and the dataset's Edit tab lists the Record IDs that were skipped.