Greater London Authority

This isn’t about a single data file but an entire dataset. The dataset is the UK’s Greater London Authority spend data:

The good news is that, as per UK government policy, the GLA provides regular data on spending over £250, and, furthermore, the data is in simple CSV files.

That’s where the good news ends. Here are primary issues:

  • Data is in many separate files (65 as per summer 2013) with no consistent naming scheme so downloading them requires scraping (or tedious click and save)
  • Amongst the 65+ CSV files one find approximately 25+ different structures (structure means things like: number of blank or non-data lines at top of files, blank columns, varying headers). This therefore requires elaborate coding to consolidate the data into a single form. You can see a full summary of all the different file structures in this data wrangle

Additional specific errors:

  • Almost every file does not start with actual data headers but with random metadata (see file to left as an example)
  • (July 2013) Period 8 2012/13 is not a CSV file but is in fact an HTML file showing a 403 Access Denied from someone’s login session. (It suggests that each CSV file is produced by manual download of a file from an internal system (probably SAP-based to judge by some file headings). See cached HTML version
    • (March 2013) Bad file for Period 8 2012/13 (13 October - 10 November). The file is not named in the usual way “Mayor’s%20250%20Report%20-%202012-13%20-%20P8%20%20-%20Final.csv” and appears to be an Excel file that was not converted to CSV!
  • Amounts are formatted with “,” making them appear as strings to computers.
  • Dates vary substantially in format from “16 Mar 2011” in this file to “21.01.2010” in January 2010 data
  • Use of (978) to indicate negative amounts rather than -978
  • Repeated data in 2012-13-P4 file (HTML view on datapipes)



The original web page now 404s:

And all the raw spending files have moved without redirects so they break … e.g. this no longer works:

The new web page only list files back to 2013 and it is not clear where the earlier files are located.