ArcGIS Online Item Usage export

Question

Hello everyone,

We are looking to export the usage of applications on our AGO site - not credit usage but actual views and associated analytics. It appears you can only view this data and not export it in any way. It would be great if we could somehow export the data via python. Attached image shows what we are after. Thanks!

Capture.PNG

MapService.png

luckachi_0-1758895257408.png

ThomasColson · Answer

The following, albeit a bit bulky, will read from a list of hosted feature services and create a csv for each with date and service calls per date. You can change the date range by changing the epoch date stamps. import csv import urllib import urllib2 import json import datetime # need to get a token tokenURL = 'https://www.arcgis.com/sharing/rest/generateToken' params = {'f': 'pjson', 'username': 'username', 'password': 'password', 'referer': 'https://www.arcgis.com'} req = urllib2.Request(tokenURL, urllib.urlencode(params)) response = urllib2.urlopen(req) data = json.load(response) token = data['token'] # read list of AGOL service names readCSV = list(csv.reader(open('service.csv'))) names = [] for row in readCSV: name = row[0] #generate url list of up to 7 api calls #Esri truncates usage stats into blocks of approx. 52 days #Making a single url call for entire year will fail #replace org with your org name and org_id with your org id url1 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1451624400000&endTime=1456808400000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token url2 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1456808400000&endTime=1461988800000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token url3 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1461988800000&endTime=1467172800000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token url4 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1467172800000&endTime=1472356800000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token url5 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1472356800000&endTime=1477540800000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token url6 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1477540800000&endTime=1482728400000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token url7 = 'https://org.maps.arcgis.com/sharing/rest/portals/org_id/usage?f=json&startTime=1482728400000&endTime=1483160400000&period=1d&vars=num&groupby=name&etype=svcusg&name='+(name)+'&stype=features&token='+token #Make 7 url calls for year. Will fail if 7 calls and feature was posted mid-year #so try blocks will keep trying less url calls try: print 'Reading the usage statistics for '+(name)+' for 2016....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url1).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url2).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url3).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url4).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url5).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url6).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except Exception: try: print 'Reading the usage statistics for '+(name)+' for the last 313 days of 2016....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url2).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url3).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url4).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url5).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url6).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except Exception: try: print 'Reading the usage statistics for '+(name)+' for the last 261 days of 2016....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url3).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url4).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url5).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url6).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except Exception: try: print 'Reading the usage statistics for '+(name)+' for the last 209 days of 216....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url4).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url5).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url6).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except Exception: try: print 'Reading the usage statistics for '+(name)+' for the last 157 days of 2016....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url5).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url6).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except Exception: try: print 'Reading the usage statistics for '+(name)+' for the last 105 days of 2016....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url6).read())['data'][0]['num'] data = data + json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except Exception: try: print 'Reading the usage statistics for '+(name)+' for the last 53 days of 2016....' filename = (name)+'.csv' data = json.loads(urllib.urlopen(url7).read())['data'][0]['num'] print 'Writing '+(filename)+'....' with open(filename, 'wb') as output: csv_writer = csv.writer(output) csv_writer.writerow(['date',(name)]) for date, count in data: csv_writer.writerow([datetime.datetime.fromtimestamp(int(date)/1e3), count]) except: print 'No usage statistics available too short time range' pass ‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

ThomasColson · Answer

I am convinced that this and this will do as you ask, but would require some python tinkering that is beyond my skill level. I too, am looking for the same thing. I have hundreds of hosted feature services, and we need a more automated/technical way of getting those into a spreadsheet so we can globally look at service usage and make management decisions on how or when we update data. Having to click the usage button on 400 different item description pages then screen-shooting the graph does not meet that need. Surprised with such a well-documented REST API, that Esri has not included this basic requirement.

IanShiland2 · Answer

Thanks I just voted that idea up. Not sure this thread will go anywhere in the AGO space but mods are welcome to move it.