OpenStreetMap Stats Generator

On the Fly Commandline Stats Generator for OpenStreetMap User Contributions

I tweet stats Every day/week/month for Global/Region and #hotosm hashtag at https://twitter.com/stats_osm and store them here

Monitored stats are available under stats, Currently Bot monitors OSM as whole , hotosm hashtag , Nepal Country : weekly,monthly and yearly stats are stored in github and twitter

Installation

Follow windows_installation for Windows Users .

• Install osmium lib on your machine

pip install osmium

• Install osmsg

pip install osmsg

Usage:

osmsg [-h] [--start_date START_DATE] [--end_date END_DATE]
             [--username USERNAME] [--password PASSWORD]
             [--timezone {Nepal,UTC}] [--name NAME]
             [--country COUNTRY [COUNTRY ...]] [--tags TAGS [TAGS ...]]
             [--hashtags HASHTAGS [HASHTAGS ...]]
             [--length LENGTH [LENGTH ...]] [--force] [--meta] [--rows ROWS]
             [--users USERS [USERS ...]] [--workers WORKERS]
             [--url URL [URL ...]] [--last_week] [--last_day] [--last_month]
             [--last_year] [--last_hour] [--days DAYS] [--charts] [--summary]
             [--exact_lookup] [--changeset] [--all_tags] [--temp]
             [--format {csv,json,excel,image,text} [{csv,json,excel,image,text} ...]]
             [--read_from_metadata READ_FROM_METADATA] [--update]

Options:

  -h, --help            show this help message and exit
  --start_date START_DATE
                        Start date in the format YYYY-MM-DD HH:M:Sz eg:
                        2023-01-28 17:43:09+05:45
  --end_date END_DATE   End date in the format YYYY-MM-DD HH:M:Sz eg:
                        2023-01-28 17:43:09+05:45
  --username USERNAME   Your OSM Username : Only required for Geofabrik
                        Internal Changefiles
  --password PASSWORD   Your OSM Password : Only required for Geofabrik
                        Internal Changefiles
  --timezone {Nepal,UTC}
                        Your Timezone : Currently Supported Nepal, Default :
                        UTC
  --name NAME           Output stat file name
  --country COUNTRY [COUNTRY ...]
                        List of country name to extract (get id from
                        data/countries), It will use geofabrik countries
                        updates so it will require OSM USERNAME. Only
                        Available for Daily Updates
  --tags TAGS [TAGS ...]
                        Additional stats to collect : List of tags key
  --hashtags HASHTAGS [HASHTAGS ...]
                        Hashtags Statistics to Collect : List of hashtags ,
                        Limited until daily stats for now , Only lookups if
                        hashtag is contained on the string , not a exact
                        string lookup on beta
  --length LENGTH [LENGTH ...]
                        Calculate length of osm features , Only Supported for
                        way created features , Pass list of tags key to
                        calculate eg : --length highway waterway , Unit is in
                        Meters
  --force               Force for the Hashtag Replication fetch if it is
                        greater than a day interval
  --meta                Generates stats_metadata.json including sequence info
                        , start_data end_date , Will be useful when running
                        daily/weekly/monthly by service/cron
  --rows ROWS           No. of top rows to extract , to extract top 100 , pass
                        100
  --users USERS [USERS ...]
                        List of user names to look for , You can use it to
                        only produce stats for listed users or pass it with
                        hashtags , it will act as and filter. Case sensitive
                        use ' ' to enter names with space in between
  --workers WORKERS     No. of Parallel workers to assign : Default is no of
                        cpu available , Be aware to use this max no of workers
                        may cause overuse of resources
  --url URL [URL ...]   Your public list of OSM Change Replication URL ,
                        'minute,hour,day' option by default will translate to
                        planet replciation url. You can supply multiple urls
                        for geofabrik country updates , Url should not have
                        trailing / at the end
  --last_week           Extract stats for last week
  --last_day            Extract Stats for last day
  --last_month          Extract Stats for last Month
  --last_year           Extract stats for last year
  --last_hour           Extract stats for Last hour
  --days DAYS           N nof of last days to extract , for eg if 3 is
                        supplied script will generate stats for last 3 days
  --charts              Exports Summary Charts along with stats
  --summary             Produces Summary.md file with summary of Run and also
                        a summary.csv which will have summary of stats per day
  --exact_lookup        Exact lookup for hashtags to match exact hashtag
                        supllied , without this hashtag search will search for
                        the existence of text on hashtags and comments
  --changeset           Include hashtag and country informations on the stats.
                        It forces script to process changeset replciation ,
                        Careful to use this since changeset replication is
                        minutely according to your internet speed and cpu
                        cores
  --all_tags            Extract statistics of all of the unique tags and its
                        count
  --temp                Deletes downloaded osm files from machine after
                        processing is done , if you want to run osmsg on same
                        files again keep this option turn off
  --format {csv,json,excel,image,text} [{csv,json,excel,image,text} ...]
                        Stats output format
  --read_from_metadata READ_FROM_METADATA
                        Location of metadata to pick start date from previous
                        run's end_date , Generally used if you want to run bot
                        on regular interval using cron/service
  --update              Update the old dataset produced by osmsg , Very
                        Experimental : There should be your name stats.csv and
                        summary.csv in place where command is run

It is a Simple python script processes osm files live and produces stats on the fly

1. It can Generate Stats on Country Level for countries based on geofabrik urls . Countries are available in here
2. It can also take any other server replication changefile to extract stats (Tested with Geofabrik and Planet Replication)
3. Can Generate hashtag statistics
4. Supports multiple output formats , Visualize Automatic Stats Being Genarated here
5. It can create summary charts automatically along with stats , Visualize them here
6. Can generate stats for any custom timestamp provided

{"name":"username","uid":uid,"changesets":1,"nodes.create":1071,"nodes.modify":2100,"nodes.delete":0,"ways.create":146,"ways.modify":69,"ways.delete":0,"relations.create":0,"relations.modify":1,"relations.delete":0,"building.create":138,"building.modify":11,"building.delete":0,"highway.create":5,"highway.modify":49,"highway.delete":0,"waterway.create":0,"waterway.modify":4,"waterway.delete":0,"amenity.create":0,"amenity.modify":3,"amenity.delete":0,"landuse.create":3,"landuse.modify":1,"landuse.delete":0,"natural.create":0,"natural.modify":3,"natural.delete":0,"total_map_changes":3387}

Start Right Away :

• Extract Stat of last hour and visualize stats/charts

  By default replication is minute url.

osmsg --last_hour

• With Hashtags information

osmsg --last_hour --changeset

• Last week data with summary & all tags info and using day replication

osmsg --last_week --url day --summary --all_tags

More Example Commands

• To extract stats of last_hour for whole_world Using Planet Replication with specific tags

osmsg --url "https://planet.openstreetmap.org/replication/minute" --format csv --tags building highway waterway amenity --name stats --all_tags --last_hour

In order to extract for specific country just add --country with country name as in data/countries_un.csv for eg : For Nepal : --country Nepal

• To extract stats for last day whole world with all the tags and specified stat :

  You can either pass url "https://planet.openstreetmap.org/replication/day" itself or if you want to use planet default replciation url you can simply pass as : minute , day & hour , script will get url itself

osmsg  --url day --format csv --last_day --name stats --all_tags

If you wish to have tags with specific count for key you can include them as --tags "building" "highway" & add --country to extract specific country , if you use geofabrik country updates you don't need to use --country option

• To extract specific country with Geofabrik URL (extracts stats for 15 days custom date range)

export OSM_USERNAME="yourusername"
export OSM_PASSWORD="yourpassword"

osmsg  --url "http://download.geofabrik.de/asia/nepal-updates" --format csv --start_date "2023-01-15 00:00:00+00:00" --end_date "2023-01-30 00:00:00+00:00" --name all_tags_stats --all_tags

• To get stat of country for year with geofabrik replication:

  Example of nepal for year of 2022

  Processing geofabrik country based osm change files are faster as they will have changes only for country and smaller in size , or you can pass --country parameter and pass id from data/countries.csv

osmsg --start_date "2022-01-01 00:00:00+00:00" --url "http://download.geofabrik.de/asia/nepal-updates" --username 'your osm username' --password 'user osm password' --tags 'building' 'highway' 'waterway' 'amenity'  --format csv

• To extract on-going mapathons statistics

  Example of extract last 8 days of data for Turkey and Syria using hashtag smforst with geofabrik as source --summary will allow to divide and provide stats as summary sepearated by daily , You can use this to get both summary csv and user contribution csv !

osmsg --country turkey syria --username "OSM USERNAME" --password "OSM PASSWORD" --hashtags smforst --length highway --force --days 6 --tags building highway amenity waterway --all_tags --summary

• Extract mapathon stats for hashtag only

  You can use --length to get length of created features , supported for way features along with count

osmsg --url minute --hashtags mymapathonhashtag1 hashtag2 --length highway --force --tags building highway amenity waterway --all_tags --summary --start_date "2023-01-15 00:00:00+00:00" --end_date "2023-01-15 11:59:00+00:00"

Or you can simply say --last_day like this , can use hour replication to avoid processing too many files

osmsg --url hour --hashtags mymapathonhashtag1 hashtag2 --length highway --force --tags building highway amenity waterway --all_tags --summary --last_day

Check more commands examples inside stats/ stats_metadata.json

Now start generating stats with above sample commands

TIPS & Tricks of using OSMSG:

OSMSG uses/supports sources , --url provided on argument will be used for osm changefiles and for changeset it is default of planet replication/minute ( used for hashtag and country )

1. To process weekly / monthly or yearly stats , Using minute replication might take forever on small machines , You can use daily replication files /replication/day to process faster and avoid network download issues for small minute fiels

2. If you are generating stats for rightnow / past 10min / past 1 hour using specific timeframe , stick with minutely (Sometimes to reflect changes you made on stats , Planet may take few minutes)

3. For hashtag stats , planet changeset minute replication is by default for now (Found only this reliable source , couldn't find daily/monthly replication files) , Generating daily/weekly is feasible , To generate more than that you can supply multiple countries geofabrik urls so that script can go through only country changefiles as stated on example above for turkey and syria , similarly you can pass your countries url and generate monthly yearly etc or you can run osmsg one time for a year it will download and store osm files , and next time when you run it would avoid downloading and it can generate stats quicker

4. For Country stats , if you use the --country option : You can also pass country name id supported from data/countries.csv and it supports multiple countries at a time if supplied , Name would be automatically translated to available geofabrik URL or You can pass your own URL eg :

osmsg --url "http://download.geofabrik.de/asia/nepal-updates" --username '${{ secrets.OSM_USERNAME }}' --password '${{ secrets.OSM_PASSWORD }}' --format csv --last_month --tags 'building' 'highway' 'waterway' 'amenity' --name last_month_stats  --all_tags

5. It is recommended use same dir for running osmsg because osmsg stores downloaded osm files to temp dir and it can be reused next time you run stats on same period , avoids downloading all again, for eg : I run some stats for #hashtag1 for last month , and i want to run again for #hashtag2 , osmsg will use same downloaded files and avoid downloading for second time if you use the same dir for running

Benchmarks :

Speed depends upon no of cores available on your CPU and your internet speed Generally on Normal i5 machine To process a year of data for country like Nepal it takes approx 3min .

OSM LOGIN info

OSM Username and Password are not Mandatory , They are only required if you want to use geofabrik internal changefiles , There are two options you can either pass your username password on command line or export those in your env variable and script will pick it from there so that you don't need to pass your username and password to command itself like this.

export OSM_USERNAME="yourusername"
export OSM_PASSWORD="yourpassword"

Now use command as it is without username and password , OSM Login is enabled from Geofabrik itself So OSMSG Doesn't uses any third party tool/methododology for the authentication

Contributing

Contributions are always welcome! Project is in active development phase , Improving documentation is always great place to start. Here's how you can get involved:

• File an issue if you find any bugs or have a feature request.
• Set up a development environment.
• Fork the repository.
• Make your changes and submit a pull request.