Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Cancel
- Home
- :
- All Communities
- :
- User Groups
- :
- Python Snippets
- :
- Documents
- :
- Get Statistics From All Fields in an Attribute Tab...
Options
- Subscribe to RSS Feed
- Mark as New
- Mark as Read
- Bookmark
- Subscribe
- Printer Friendly Page
- Report Inappropriate Content
Get Statistics From All Fields in an Attribute Table
Subscribe
788
0
02-13-2018 06:53 AM
Get Statistics From All Fields in an Attribute Table
Recently a question made me think of how to calculate statistics of all fields (all attributes) for each record of an attribute table. See thread here:how to add up all the fields in an attribute table
The OP actually wanted to get the summary statistics of a field (not sum all fields of a record), but I thought I would share a snippet of how you can accomplish this. See the code below:
def main():
import arcpy
# input featureclass and output field
fc = r'C:\GeoNet\SumAllFields\data.gdb\test_pol'
fld_result = 'AvgDate'
# settings to play with
wild_card = '*DATE' #'FLD*'
stats_type = 'mean' # min, max, sum, mean, count, std
fld_types = ['Date', 'Double', 'Integer', 'Single', 'SmallInteger']
include_system_fields = False
ignore_nodata = True
# get field names
fld_names = GetFieldNames(fc, wild_card, fld_types, fld_result, include_system_fields)
# add the field if it does not exist
if len(arcpy.ListFields(fc, fld_result)) == 0:
arcpy.AddField_management(fc, fld_result, "DOUBLE")
# update cursor
flds = [fld_result]
flds.extend(fld_names)
with arcpy.da.UpdateCursor(fc, flds) as curs:
for row in curs:
lst = list(row)[1:]
row[0] = GetStats(lst, stats_type, ignore_nodata)
curs.updateRow(row)
def GetFieldNames(fc, wild_card, fld_types, fld_result, include_system_fields):
# exclude result field
exclude_fields = [fld_result]
# create initial list of fields to include
fld_names = sorted([fld.name for fld_type in fld_types for fld in arcpy.ListFields(fc, wild_card, fld_type)])
# exclude system fields
if include_system_fields == False:
try:
exclude_fields.append(arcpy.Describe(fc).AreaFieldName)
except:
pass
try:
exclude_fields.append(arcpy.Describe(fc).LengthFieldName)
except:
pass
for exclude_field in exclude_fields:
if exclude_field in fld_names:
fld_names.remove(exclude_field)
return fld_names
def GetStats(lst, stats_type, ignore_nodata):
import numpy
stats = None
if ignore_nodata:
lst = [a for a in lst if not a is None]
try:
if len(lst) > 0:
if stats_type == 'sum':
stats = sum(lst)
elif stats_type == 'min':
stats = min(lst)
elif stats_type == 'max':
stats = max(lst)
elif stats_type == 'mean':
if type(lst[0]) is datetime:
mean = (numpy.array(lst, dtype='datetime64')
.view('i8')
.mean()
.astype('datetime64'))
ts = (mean - numpy.datetime64('1970-01-01T00:00:00Z')) / numpy.timedelta64(1, 's')
stats = datetime.utcfromtimestamp(ts)
else:
stats = sum(lst) / float(len(lst))
elif stats_type == 'count':
stats = len(lst)
elif stats_type == 'std':
arr = numpy.array(lst)
stats = numpy.std(arr, axis=0)
else:
print 'Error: Unknown stats_type'
except Exception as e:
print 'Error:', e
print ' - lst:', lst
print ' - ignore_nodata:', ignore_nodata
finally:
return stats
if __name__ == '__main__':
main()
To play with this code, change:
- on line 5, the path to the featureclass or table
- on line 6 the output field name that will hold the result
- on line 9 you can specify a wild card for the fields to include: The example "*DATE" will consider all fields that end with the text DATE. You can also leave it as "*" to include all fields
- on line 10 you define the statistics type. Valid values are: min, max, sum, mean, count and std
- on line 11 you define a list of field types to include in the operation
- on line 12 you can toggle between include double fields like shape length and area or to exclude them
- on line 13 the ignore_nodata var allows you to ignore NULL/None values, they will be removed from the list of values used to determine the statistics.
Have fun...
