Re: [Talk-GB] Quarterly project - taginfo tracker

Hi Steve, Thanks for your email. Hope you don’t mind but I’ve cc’d Talk-GB in case this is relevant to anyone else. Unfortunately there is a bit of a lag because the Geofabrik extract for Great Britain is a bit out of date by the time it’s downloaded. The extract that is available right now

Re: [Talk-GB] Quarterly project - taginfo tracker

On 09/11/2016 07:36, Greg wrote: I’ve now implemented Steve’s request for more clickable links in the overview map for each district at . In addition to opening the web page for each FHRS establishment, you can now open the OSM page for each relevant

Re: [Talk-GB] Quarterly project - taginfo tracker

Hi all, I’ve now implemented Steve’s request for more clickable links in the overview map for each district at . In addition to opening the web page for each FHRS establishment, you can now open the OSM page for each relevant node/way or load the

Re: [Talk-GB] Quarterly project - taginfo tracker

On 1 November 2016 at 10:14, Greg wrote: > It looks like the FSA or local authority has changed the FHRS IDs for a > large number of establishments since they were inputted into OSM. For > example Dyce Academy: > > http://www.openstreetmap.org/way/237994869 > OSM fhrs:id:

Re: [Talk-GB] Quarterly project - taginfo tracker

Looking at the graphs, I wonder what happened to Aberdeen? Seems to have 350+ fhrs:ids that weren't expected. http://gregrs.dev.openstreetmap.org/fhrs/district-369.html On Sun, 30 Oct 2016 at 14:25 Dave F wrote: > Much clearer, thanks > > DaveF > > > On 30/10/2016

Re: [Talk-GB] Quarterly project - taginfo tracker

Much clearer, thanks DaveF On 30/10/2016 14:00, Greg wrote: Good point, Dave. The first graph previously included only OSM nodes/ways which had been fully matched with fhrs:id and postcode but I have now amended the first graph and the ‘overall %’ graph to show nodes/ways with a

Re: [Talk-GB] Quarterly project - taginfo tracker

Good point, Dave. The first graph previously included only OSM nodes/ways which had been fully matched with fhrs:id and postcode but I have now amended the first graph and the ‘overall %’ graph to show nodes/ways with a mismatched/missing postcode too. If anyone would like to improve the

Re: [Talk-GB] Quarterly project - taginfo tracker

Hi Greg The first graph. Does that represent the addition of just the fhrs:id tag or with a postcode as well? Cheers DaveF On 29/10/2016 12:24, Greg wrote: Hi all, I’ve created some graphs to show FHRS completeness and progress so that we can track the impact of the Quarterly Project. I

Re: [Talk-GB] Quarterly project - taginfo tracker

Hi all, I’ve created some graphs to show FHRS completeness and progress so that we can track the impact of the Quarterly Project. I hope to update the graphs weekly. Please do have a look here [http://gregrs.dev.openstreetmap.org/fhrs-stats/graphs.html] and let me know if you have any

Re: [Talk-GB] Quarterly project - taginfo tracker

__ From: Greg <gre...@fastmail.co.uk> Sent: 18 October 2016 20:28:56 To: Rob Nickerson Cc: Talk-GB Subject: Re: [Talk-GB] Quarterly project - taginfo tracker Hi Rob, I've been working on some visual analysis of the CSV file data at [http://gregrs.dev.openstreetmap.org/fhrs-stats/

Re: [Talk-GB] Quarterly project - taginfo tracker

Hi Rob, I’ve been working on some visual analysis of the CSV file data at [http://gregrs.dev.openstreetmap.org/fhrs-stats/] using R but it’s not quite ready for public viewing yet… I hope to show the most complete districts in terms of fhrs:id and postcode, the districts with the most errors,

Re: [Talk-GB] Quarterly project - taginfo tracker

On 17/10/2016 21:47, Greg wrote: I think the most useful link would probably be one that loads the OSM node/way in JOSM using the remote control function in the same way that the suggested matches map links do. Obviously you’d have to put the address details into JOSM manually, but you could

Re: [Talk-GB] Quarterly project - taginfo tracker

Hi Steve, I’m glad you’re finding the tool useful and thanks for your suggestion. I’ve added a feature request in the issue tracker here [https://github.com/gregrs-uk/python-fhrs-osm/issues/20]. I probably won’t have time to look at this over the coming week, but I may have time at some point

Re: [Talk-GB] Quarterly project - taginfo tracker

Thanks for providing this tool, Greg. I've found it very useful so far for adding address information based on the possible matches identified on the second map, where most of the dots have a link both to OSM and to the FHRS entry. However, I've now reached the stage where I need to look more

Re: [Talk-GB] Quarterly project - taginfo tracker

These district statistics CSV files will now be copied each day to http://gregrs.dev.openstreetmap.org/fhrs-stats/ , so there will be a file per day in case anyone would like to track progress. Thanks, Greg > On 9 Oct 2016, at 21:00, Greg

Re: [Talk-GB] Quarterly project - taginfo tracker

The python-fhrs-osm tool will now create a CSV file containing statistics for each district each day, which should allow detailed tracking. It will be located at http://gregrs.dev.openstreetmap.org/fhrs/stats--mm-dd.csv once

Re: [Talk-GB] Quarterly project - taginfo tracker

In case it's useful for tracking, you can get the full list of tags used by my comparison tool at the link below. The tool downloads nodes/ways with those tags and any nodes/ways with an fhrs:id tag set. https://github.com/gregrs-uk/python-fhrs-osm/blob/master/filter-osm.sh#L5-L13 Thanks, Greg

Re: [Talk-GB] Quarterly project - taginfo tracker

On 04/10/2016 16:15, Dan S wrote: Hi, "Horses for courses" is certainly correct - but let me comment: please don't rely on website=* as a way for people to get hold of further info. Even if the establishment continues to exist, many small places have websites that vanish, or are stolen, or are

Re: [Talk-GB] Quarterly project - taginfo tracker

Hi, "Horses for courses" is certainly correct - but let me comment: please don't rely on website=* as a way for people to get hold of further info. Even if the establishment continues to exist, many small places have websites that vanish, or are stolen, or are full of incorrect info. I've

Re: [Talk-GB] Quarterly project - taginfo tracker

External links of all kinds can die; but for me the issue is that external addresses make life difficult for data consumers. J On 4 October 2016 at 16:05, Dave F wrote: > Horses for courses. I map what I'm interested in, which maybe considered > selfish by some,

Re: [Talk-GB] Quarterly project - taginfo tracker

Horses for courses. I map what I'm interested in, which maybe considered selfish by some, but it still adds relevant data. Websites provide current data on establishment. Surely, by your logic, churn will also mean the names of establishment will disappear? Comparing a newly issued fhrs

Re: [Talk-GB] Quarterly project - taginfo tracker

I'd really strongly disagree here. Getting address data into OSM is important; the shop/amenity stuff is the sugar which coats that pill. In 5 years time (see my figures) many of the fhrs:id will have disappeared from the website. Jerry On 3 October 2016 at 13:13, Dave F

Re: [Talk-GB] Quarterly project - taginfo tracker

For me, fhrs:id is obviously needed. ATM I'm only adding 'website' as an additional tag. From that an end user can ascertain info such as an establishments address & the services it provides. Adding full addresses is something I might get around to doing later id there not anything more

Re: [Talk-GB] Quarterly project - taginfo tracker

I think all these measurements are a great idea but in the first instance for the taginfo script I think fhrs:id is sufficient. At least we'll know that is a direct response to our QP. Postcode data I think is a little more tricky than a raw count of objects tagged with addr=postcode. For

Re: [Talk-GB] Quarterly project - taginfo tracker

On 2 October 2016 at 15:38, Rob Nickerson wrote: > Which tags would you like me to set up a tag-info script for? We can then > track these throughout the quarter. Off the top of my head, I'd have thought it would be good to know about number of instances of fhrs:id=*