The graphs below represent statistics relating to the OpenStreetMap GB Quarterly Project 2016 Q4, which involved using Food Hygiene Rating Scheme open data to enrich the map. Links between the OSM and FHRS data were created using the OSM fhrs:id tag, and for a match to be considered successful, the addr:postcode tag had to be added to the OSM entity.

The data used for the graphs below was gathered using the FHRS/OSM comparison tool and relates to the time period between 10th October 2016 and 1st January 2017. Some figures for the districts of Dumfries, Haringey and Kingston upon Thames have been approximated because of sudden changes in the FHRS data during the project.


Overall progress

The percentage of FHRS establishments successfully matched (i.e. including a matched postcode) has been improved by a factor of 2.97 from 1.7% to 5.04%. The percentage of FHRS establishments successfully matched (but with a mismatched or missing postcode) has been improved by a factor of 2.63 from 2.14% to 5.63%.

As a result of the project, 238 districts out of 377 now have at least one successfully matched FHRS establishment as opposed to 166 districts.

Top 10 districts by progress

The graph above shows the 10 districts that have made the most progress during the project in terms of the number of matches made between FHRS establishments and OSM entities. The top district was Bristol, with 1775 new matches.

The graph above shows the 10 districts that have made the most progress during the project in terms of the increase in the percentage of FHRS establishments matched in OSM. (This crudely takes into account the fact that districts with more establishments probably also have more mappers.) The top district was Gravesham, with an increase of 89.6% of FHRS establishments matched in OSM.

The graph above shows how progress was made during the project for the districts with the highest number of new FHRS matches.

The graph above shows how progress was made during the project for the districts with the highest increase in the percentage of FHRS establishments matched in OSM.

Top 10 districts for completeness

At the end of the project, the top district in terms of the percentage of FHRS establishments matched is Gravesham with 97.4% of FHRS establishments matched by fhrs:id and 94.5% of FHRS establishments matched in terms of both fhrs:id and addr:postcode.

At the end of the project, the top district in terms of the percentage of relevant OSM entities (i.e. those likely to be in the FHRS database) which have a non-mismatched addr:postcode tag (i.e. one that doesn’t disagree with the FHRS one if the entity has been matched) is Wellingborough with 86.5% of relevant OSM entities having a non-mismatched postcode.

Districts with the most errors

The graph above shows the 10 districts with the highest number of mismatched fhrs:id tags at the end of the project. Mismatches are generally caused by the FHRS IDs of establishments being changed since they were originally added, by errors in copying, or by adding multiple fhrs:ids delimited by semicolons, which is currently unsupported by the FHRS/OSM comparison tool.

The graph above shows the 10 districts with the highest number of missing or mismatched addr:postcode tags at the end of the project.

District completeness histograms

315 out of 377 districts still have only 0–5% of FHRS establishments matched but there are some districts with a much higher percentage of matches.

The most common percentage of relevant OSM entities (i.e. those likely to be in the FHRS database) which have a non-mismatched addr:postcode tag (i.e. one that doesn’t disagree with the FHRS one if the entity has been matched) is around 16%.


Created by Gregrs using R and RStudio. Code freely available on GitHub.