Creating Simple

GIS Applications

with Python

Sam Giebner

Jesse Adams

Slides

http://slides.com/northpointgis/simple_python_applications_ws

Resources

https://github.com/northpointgis/Simple-Python-Applications-Workshop

Workshop Materials

Outline

Introduction - Tools & Tkinter

    Exercise #1

Classes - Code organization

    Exercise #2

Hooks - Capture user interaction

    Exercise #3

Modules - Connecting the dots

    Exercise #4

What is your experience with Python?

1. A python is a type of snake
2. Dabbled
3. Aspiring apprentice
4. Regular user
5. Snake Charmer

Why create a Python GUI?

Request

Coworker

Client

Goals:

• Save time
• Eliminate/reduce error
• Fill gaps in existing tools
• Amplify everyday and non-GIS users

Sharing Script/Tools

Traditional Options:

• Standalone script
• Script in an ArcGIS Toolbox
• ArcGIS Python Add-in

Great, but...

Why would I create a standalone data processing application?

• Lightweight
• Efficient
• Stable
• "On the fly" data validation
• Fully customizable user experience
• "Reactive" error handling
• Not reliant on ESRI products
• *Cross-platform
• *No Installation

Tkinter

What is Tkinter?

• most commonly used GUI toolkit for Python 
• object-oriented layer on top of Tcl/Tk
  • Tcl: Tool Command Languege
  • Tk: standard Tcl GUI toolkit 

basically, it's a collection of Python modules

Tkinter App: 3 Main Parts

1. Tk instance

2. Widgets

3. Tkinter event loop

Tk Instance

    from Tkinter import *

    root = Tk()

18 Core Widget Classes

• Button

• Canvas

• Checkbutton

• Entry

• Frame

• Label

• Listbox

• Menu

• Menubutton

• Message

• Radiobutton

• Scale

• Scrollbar

• Text

• Toplevel

• LabelFrame

• PanedWindow

• Spinbox

    # Create widget instance     
    new_button = Button(tk_instance)

    # Style widget
    new_button['text'] = 'Click Here'
    new_button['command'] = do_something()

    # Place widget
    new_button.grid(row=0, column=2)

Widget Workflow

Initiate Tkinter event loop

    root.mainloop()

Exercise 1

Classes

What are they?

• A fundamental building block of Python
• Logical grouping of data and functions
• Blueprint for createing objects
• Include bound functions called methods

    class NewClass():

        def __init__(self):
            pass

        def method_one(self):
            print("I'm a method!")

__init__ & self

self: refers to the object instance

    __init__(self):
        pass

__init__: the method that initilzes the instance

Application Organization

Interface (view) class

• generates interface
• handles user interaction
• connects to processing modules

Processing (controller) class

• processes data provided by user (via view)
• returns product to interface

Exercise 2

Hooks

Hooks or Callbacks

Capture, tie in user interaction

def callback():
    print "clicked!"

Button(text="click me", command=callback)

def callback(number):
    print "button", number

Button(text="one",   command=lambda: callback(1)).pack()
Button(text="two",   command=lambda: callback(2)).pack()
Button(text="three", command=lambda: callback(3)).pack()

Using a single callback for multiple widgets; lamda allows for passing args to callback

Exercise 3

Modules

What is a Module?

• A Python object that you can bind and reference
  • Contain:
    • Functions
    • Classes
    • Variables
    • Runnable code

Package

__init__.py

Module.py

Subpackage

__init__.py

Module.py

What is a Package?

• A collection of related modules

Imports

Global import:

    from data_processing.csv2shp_esri import CSV2SHP_ESRI

    import arcpy

Local import:

• __init__.py
• adding directories to the python path sys.append

Exercise 4

Distributing Your App

Considerations

• Execution

• Dependencies

Execution

Run the python (.py) script

Run batch (shell) script that calls python script

• ​​Useful for sending system parameters to app (such as python version)

Use a packaging system like PyInstaller to create a single-file executable

• Beware that these setups might fail to include system-wide installs like arcpy

Dependencies

Make sure all application dependencies are installed in the root directory of the primary python script

• This may not be possible with libraries such as arcpy, and the end-user should be aware of application requirements such as this

Avoid hard-coding path names

Folder structure should remain intact

Packaging systems can help with some of these issues