Geo Datasets for GB, NI, CI and the Isle of Man.

PostcodePal™

Database Generator

Generate and download a postcode areas, districts or sectors database based on centroids.

Centroids

Our centroid databases are available for download in a variety of different centroid types, which include true centroids, average centroids and our slightly more controversial density based centroids which are much more accurate. Please scroll to the bottom of this page for more details.



 DB #1    Postcode Areas Centroids Database  
  Info   A postcode areas database dynamically generated by our server using our latest datasets and utilising true or average centroids.
Format CSV / SQL / XML / JSON
Price FREE
Records 124
URL 1 Generate True Centroids Database & Download
URL 2 Generate Average Centroids Database & Download
 
 DB #2    Postcode Districts Centroids Database  
  Info   A postcode districts database generated by our server using our latest datasets and utilising true, average or density centroids.
Format CSV / SQL / XML / JSON
Price FREE
Records 2,981
URL 1 Generate True Centroids Database & Download
URL 2 Generate Average Centroids Database & Download
URL 3 Generate Density Centroids Database & Download
 
 DB #3    Postcode Sectors Centroids Database  
  Info   A postcode sectors database generated by our server using our latest datasets and utilising true, average or density centroids.
Format CSV / SQL
Price FREE
Records 11,151
URL 1 Generate True Centroids Database & Download
URL 2 Generate Average Centroids Database & Download
URL 3 Generate Density Centroids Database & Download
 

Centroid Types

If you're unsure which type of centroid database best suits your needs then please take your time to read the pros and cons, before downloading. The most accurate centroid type is "density".

True Centroids

A true centroid is calculated by taking every single unique coordinate from all of the live postcode units in an area, district or sector and applying a mathematical formula in order to find the true centre point. A true centroid is the fairest as all coordinates have an equal say over the position of the final centroid and the calculation itself is completely non-discriminatory. All geographical locations are equated for, regardless of whether they are residential, commercial or po boxes. Excluding one from the other does not make a centroid database any more accurate in its entirety and exclusion is also unfair and discriminatory. The "bigger" picture is always the "most" accurate.

Average Centroids

An average centroid is calculated by taking every single unique coordinate from all of the live postcode units in an area, district or sector and applying a mathematical formula in order to find the average centre point. An average centroid is very similar to a true centroid except that clustered areas carry slightly more weight in determing the position of the final centroid coordinate. The advantages and disadvantages between true and average centroid calculations are equal in mathematical volume, so neither has any advantage or disadvantage over the other.

Density Centroids

A density based centroid is based on a complex mathematical formula developed exclusively by us. It's by far the most accurate centroid there is for calculating distances between user input data, but it's also discriminatory. In order to calculate this centroid we take every unique coordinate in a postcode area, district or sector and measure its density in relation to every other coordinate. The best way to visualise this is to imagine dropping hundreds or sometimes thousands of stones into a pond and analysing the size of the circles they generate and creating points at every intersection, as these circles overlap. These points are the secondary centroids that we then use to calculate the position of the primary centroid, which becomes the official centre of density. It is discriminatory because lower density locations are bullied by higher density locations in the mathematical war over the position of the final centroid. It's the most accurate type of centroid for user input data, because the users represent density. They are more likely to be located in a higher density area than a lower density one. This means that the greater majority of the population will get a more accurate calculation in terms of distance, but a smaller minority will get less accurate calculations, because the area in which they live is less populated.

The centre of populated density is more accurate for most and less accurate for least. District EX19 is a fine example of this. A random scatter of postcodes on the left and the town of Winkleigh on the right. A true or average centroid would be positioned in the middle of a park at great distance from the populated areas on either side. The random scatter do not carry much weight in terms of populated density because they are scattered and do not relate to each other very well, which indicate smaller towns or villages. The ripples in the pond do not even meet the splash created by Winkleigh so are completely disregarded. The distances involved are just too great. This means that no secondary centroid is created in which to base the position of the primary centroid. Had bigger splashes been made and intersecting overlaps found then the story would be quite different. As it would have been if there were multiple intersecting overlaps from the top and bottom. The whole point of the centre of populated density is to ensure that it's centralised to encompass the maximum amount of coordinates. The question you need to be asking yourself is do you require a centroid in the middle of a park or one that is centralised in relation to the greater mass of population whom are statistically most likely to be in that location?

Legal Notice: Even though this centroid is more accurate for the greater majority of the population, your usage and deployment of this centroid type may actively discriminate against a minority of the population based upon where they live. We cannot be held legally responsible for your deployment of these centroids. For complete peace of mind we recommend that you use non-discriminatory true or average centroids in your calculations or preferably switch to a more accurate unit level calculation. A must if you wish to run a successful "find my nearest" solution.

Important Notes

District and Sector based centroid calculations "can" be accurate to within 2-3 miles, but they can also be inaccurate by upto 30-70 miles. Distance calculations within the same district or sector will ALWAYS be 0 miles. As a general rule the "greater" the distance the more "accurate" the calculation. These databases are not recommended for "find my nearest" solutions, but are more suitable for performing calculations over larger distances and for determing rough geography.