Dataseed works with a number of different file types such as csv, tsv, xls and xlxs. You can also import from various sources:
|Your computer||Google Drive|
Data comes in all shapes and sizes, but unfortunately these aren’t all Dataseed friendly. We’re here to help you get your data into shape - just follow the easy guidelines outlined below.
Let’s start with the basics.
To use Dataseed’s full capabilities it’s best to work with raw data (also known as primary data), i.e. data collected from a source and which has not been subjected to processing or any other manipulation. The first thing to do is clean up aggregated and descriptive data in your dataset.
Make sure there is more than one column, there are no blank columns or rows and no duplicate column headers in your dataset. These are usually simple to fix, but may require using a tool like Data Wrangler or OpenRefine if the problem is more systematic.
Remove any introductory and other unnecessary text from your data. Headers must be limited to a single row. If your dataset is small enough to visually inspect you may want to fill in any blank cells and missing column headers.
The data also needs to be in a standard format where each row is a single record. This means that datasets with cross tabulations or pivot tables need to be normalised before importing it into Dataseed.
Dataseed classifies data as into six data types: text, date, numeric (whole), numeric (decimal) and geo.
Dates formats (all dates > 1900)
|day, dd month yyyy||month dd, yyyy|
|dd month yyyy||mm/yyyy|
|yyyy (where yyyy is between 1500 and current year)|
- World ISO3166 (alpha2, alpha3, numeric and name)
- UK (ons, gss, unitid, and name)
- England & Wales (ons, gss, unitid, and name)
For more information on about how to create geo charts, please our article on creating maps / geo charts.