Working with Feature Data
using ArcPy
David Wynne, Ghislain Prince
Abstract
• Working with Feature Data using ArcPy
• In this workshop, see highlights and demonstrations of accessing and editing
feature data with arcpy and the powerful data access module.
• Major topics include accessing data types in folders and geodatabases,
starting edit sessions, fast cursor support, using NumPy arrays to extend
analysis, and managing geodatabase objects (domains, subtypes, versions,
and replicas).
• Demos: http://esriurl.com/7537
Agenda
Cursors Listing data NumPy
Edit session Walk
support
Cursors
Cursors
• Cursors provide record-by-record access
- Basic necessity for many workflows
SearchCursor Read-only access
UpdateCursor Update or delete rows
• Two implementations:
InsertCursor Insert rows
- arcpy.da cursors (introduced at 10.1)
- ‘Classic’ cursors (date back to 9.0)
• Which one? Unless you have legacy code you don’t want to update, should
default to using arcpy.da cursors
- Much faster
- Classic cursors can be a Bottleneck for some Python workflows
Cursors
• arcpy.da cursors use lists and tuples
- Row values are accessed by index
• Different from ‘classic’ cursors
- Work with row objects
- Row values are handled using setValue,
getValue properties
with statements
• arcpy.da Cursors support with statements
• With statement
- Provide clarity
- Other benefits: such as allowing edits on multiple tables in the same workspace
Fields and tokens
• For best performance, use only those
fields you need
• Tokens can be also be used
- Get only what you need : asking for full
geometry is expensive
Demo
Cursor demo
Editor class
Editor class
• Supports edit sessions and edit operations
• Edits are temporary until saved and permanently applied
• Can quit an edit session without saving changes
Editor
• When do you need to use?
- To edit feature classes that participate in a…
- Topology
- Geometric network
- Versioned datasets in ArcSDE geodatabases
- Some objects and feature classes with class extensions
Editor using a with statement
• Editor supports with statements
- Handle appropriate start, stop and abort calls for you
Open an edit
session and
start an edit
operation
Exception—operation is aborted, and
edit session is closed without saving
No exceptions—stop the operation
and save and close the edit session
Editor class
• Editor class also includes methods for working with edit sessions and
operations
Editor methods
startEditing ({with_undo},
Starts an edit session.
{multiuser_mode})
stopEditing(save_changes) Stops an edit session.
startOperation() Starts an edit operation.
stopOperation() Stops an edit operation.
abortOperation() Aborts an edit operation.
Undo an edit operation (roll
undoOperation()
back modifications).
redoOperation() Redoes an edit operation.
Editor class
Demo
Editor demo
Iterating over data
List functions and Walk
Iterating over data
• More list functions to support workflows
arcpy.da list functions
ListDomains Lists the attribute domains belonging to a
geodatabase.
ListFieldConflictFilters Lists the fields in a versioned feature class
that have field conflict filters applied.
ListReplicas Lists the replicas in a workspace.
ListSubtypes Return a dictionary of the subtypes for a
table or feature class.
ListVersions List the versions in a workspace.
Walk
• Traverse a directory structure to find ArcGIS data types
• Returns a tuple of three: path, path names, and filenames
• Similar pattern to Python’s os.walk
• Comparison:
- Walk: http://esriurl.com/5931 vs. the hard way: http://esriurl.com/5932
Demo
Iterating over data
NumPy
NumPy support
• NumPy is a 3rd party Python library for scientific computing
- A powerful array object
- Sophisticated analysis capabilities
• arcpy.da supports converting tables and feature classes to/from NumPy
• RasterToNumPyArray / NumPyArrayToRaster
- Added at 10.0 to support converting rasters to and from numpy arrays
NumPy functions
• arcpy.da provides additional support for tables and feature classes
Function
FeatureClassToNumPyArray Convert a feature class to an
array
TableToNumPyArray Convert a table to an array
NumPyArrayToFeatureClass Convert an array to a Feature
Class
NumPyArrayToTable Convert an array to a Table
ExtendTable Join an array to a Table
Export to NumPy
• Can convert tables and feature classes into numpy arrays for further analysis
Import from NumPy
• Take the product of your work in numpy and export it back to ArcGIS
• Need to output polygons, lines, multipoints? http://esriurl.com/5862
Demo
Exporting to NumPy
Questions?
• Demos: http://esriurl.com/7537
• http://www.esri.com/events/devsummit/session-rater
