Hi Drew
Below is the basic script I put together. It's pretty straightforward.
Just be warned that, if run as is, it truncates the dataset and deletes everything! Don't get caught out if that's not what you want to do!
You would need to adjust the part that reads the source data and manipulates it into the json object as you need to construct an object with the correct fieldnames etc. I was just updating plain tables, so no geometry involved here but could be worked in I guess.
from arcgis.gis import GIS
from getpass import getpass #to accept passwords in an interactive fashion
import pandas as pd
import time
import datetime
username = input()
password = getpass()
gis = GIS("https://www.arcgis.com", username, password)
item_to_update = gis.content.get('item id from agol')
for index, table in enumerate(item_to_update.tables):
if table.properties.name == 'name of the table':
table_to_update = table
#confirmation we have the correct table...
print(str(table_to_update.properties.name))
# WARNING!! This deletes EVERYTHING in the table!
table_to_update.delete_features(where="objectid > 0")
# A couple of little helper functions:
def getDate(datestring):
if datestring is None:
return None
datestring = str(datestring).strip()
if datestring == 'nan' or datestring == 'NaT':
return None
try:
return datetime.datetime.strptime(datestring, "%Y-%m-%d %H:%M:%S")
except:
pass
return None
def getString(s):
if s is None:
return ''
if isinstance(s, str) is False:
s = str(s).strip()
if s == 'nan':
return ''
return s
sourceDataFrame = pd.read_excel(r'c:/temp/spreadsheet_data.xlsx')
sourceDataFrame.head()
features_to_be_added = []
notifier = 0
#this bit needs to be customised for the data source and data it contains...
for index, row in sourceDataFrame.iterrows():
notifier += 1
if notifier == 1000:
print('Number processed: ' + str(index + 1))
notifier = 0
try:
nextFeature = {"attributes":{}}
nextFeature["attributes"]["field1"] = getString(row['Field1'])
nextFeature["attributes"]["field2"] = getString(row['Field2'])
nextFeature["attributes"]["field3"] = getDate(row['Field3'])
features_to_be_added.append(nextFeature)
except ():
print(">>>>>>>>>>>>>>>>>>Error at index: " + str(index) )
len(features_to_be_added) #confirmation of number of features...
start = time.clock()
print('Number of features to upload: ' + str(len(features_to_be_added)))
chunksize = 2000
chunks = [features_to_be_added for x in range(0, len(features_to_be_added), chunksize)]
for index, chunk in enumerate(chunks):
print('chunk #' + str(index)+ ' with ' + str(len(chunk)) +' features started at {0:.2f} seconds'.format(time.clock() - start))
results = table_to_update.edit_features(adds = chunk)
print('Completed upload of {0} features in {1:.2f} seconds'.format(str(len(features_to_be_added)), time.clock() - start) )
