davidjudilla / bus-drivers-arc Goto Github PK

The Crash Analysis Toolbox is a collection of python based ArcMap scripts, and tools for use in ArcGIS for Desktop versions 10.2.x and above. These tools offer a more streamlined approach to analyzing traffic crash data, compared to ArcGIS’ model builder. The tools are also modular, meaning that python tools can be removed or added to each project without much effort.

The primary user of the Hotspot Analysis Tool Suite is Dr. Ghazan Khan, an assistant professor of civil engineering, specializing in transportation engineering, at California State University, Sacramento. Dr. Khan has expert-level knowledge of the ArcGIS platform, and he is a subject matter expert in the crash analysis field. Dr. Khan lectures about ArcGIS and road safety, and his civil engineering students may utilize the developed tools in their research (particularly graduate students). All users of the Hotspot Analysis Tool Suite are expected to be proficient in ArcGIS.

The Crash Analysis toolbox has 4 major tools: Crash Radius Density, Crash Network Density, Network K Analysis, and Cross K Function. Each one of these tools acts takes an existing crash data (see "scratch/collision data/Collisions.csv" for test data) and takes in various inputs (defined by each tool) and outputs analysis by each type of tool.

Each of the tools are located within their own python script (as well as helper python scripts) in the "toolbox" folder. Also, Each of the tools support have documentation built into the xml metadata which show in the ArcMap User Interface "Item Descriptions". To see them, go to the ArcCatalog, then right click on a python tool and select “Item Description”. A new window will open, describing the tool. These "self-documenting" item descriptions are maintained in the corresponding tools .xml file in the "toolbox" folder. These files can be edited in two ways, through the xml file or through the ArcCatalog UI (See Section 4 - Tools in the "As Built Design - Crash Analysis Toolbox").

• The Crash Analysis toolbox has 7 major tools in the "toolbox" folder (refer to the User Manual or consult with Dr. Khan for details on each tool as they have background information of each tool as well as expected output):

  Crash Radius Density

  • Summary
    • Creates a bounding circle with a user defined radius around crashes (as coordinates), allowing a user to view how many crashes cluster within the user defined radius of a given crash point.
  • Usage
    • An existing feature dataset must exist in order to use the CrashRadiusDensity tool.
    • The tool can then be run by inputing a feature dataset as well as the radius for each of the crash points.

  Crash Network Density

  • Summary
    • The CrashNetworkDensity tool creates a network using an existing dataset of crashes (as coordinates) as well as the Open Street Map Network (view Syntax).
  • Usage
    • An existing Feature Dataset must exist to use this tool.
    • Once a feature dataset exists, the CrashNetworkDensity tool can be used on this dataset to create a network with the Open Street Map Network.

  Create Random Points on a Network Dataset

  • Summary
    • Creates a specified number of points randomly on a network dataset as a feature class.
  • Usage
    • See parameters for Required and Optional Fields.
    • Outputs a series of random points on a network dataset.

  Global K Function

  • Summary
    • The Global K Function tool allows the user to find crash clustering and dispersion at different distance bands.
    • The “Global” analysis type calculates the global statistic to identify the presence or absence of a systematic process, the magnitude, and the extent of spatial patterns in crash data.
  • Usage
    • See parameters for description for Required and Optional inputs.
    • Outputs a table of raw data which is used to generate the Network K Function.

  Cross K Function

  • Summary
    • The Cross K Function tool that allows the user to find crash clustering and dispersion at different distance bands using Dr. Ghazan Khan’s modified version of the Ripley’s K Function. For a given set of crash points, this function determines deviations from spatial homogeneity at different distances along a road network.
    • Cross-K analysis, is comprised of local statistics at individual crash locations to identify individual crashes as part of a statistically significant clustered (hotspot), dispersed (coldspot), or random point pattern. The methods under the crash cluster analysis category were developed based on the principles of the K-Function method in network space.
  • Usage
    • See parameters for description for Required and Optional inputs.
    • Outputs a table and a database of raw data which is used to generate the Network K Function.

  Random ODCM Permutations

  • Summary
    • Generates OD Cost Matrices with oberved data and a complementrary set of random point permutations.
    • The tool can take a long time to run based on the number of permutations the tool is set to run for.
  • Usage
    • Before using, a requirement is that the user needs to know which type of analysis to do Global K-Function vs. Cross K- Function Analysis.
    • See parameters for description of Required and Optional inputs.

• ArcGIS Desktop 10.2+
• Python 2.7
• Some tools require additional packages
  • The ArcGIS Editor for OSM (OpenStreetMap) - ArcGIS extension required for the Crash Network Density and Network K Analysis Tools, specifically to be able use open-source network data of streets within the toolbox. For installation, follow the link below (urls subject to change):
    
    • 10.2.x - http://www.arcgis.com/home/item.html?id=16970017f81349548d0a9eead0ebba39
    • 10.3.x - http://www.arcgis.com/home/item.html?id=75716d933f1c40a784243198e0dc11a1

• Crash Analysis Toolbox - User Manual: User Manual

• First Understand Dr. Khan's problem from a theoretical standpoint (take ArcGIS out of the equation)
• Learn Basics of ArcGIS (Check out Youtube Videos on) - What is the ArcGIS System, how is it used - What are the various data types in ArcGIS (geodatabases, feature classes, network layers) - How to add geographical data from a csv/excel file onto a map in ArcGIS - What are tools and toolboxes in ArcGIS - Understand what OpenStreetMap API is
  - Program a simple tool in ArcMap
• Become a user for the existing tools: - Learn how to use the tools, Read the User Manual for the tools, CONSULT WITH DR. KHAN on why he uses the tools if anything becomes unclear. - Learn Git and how to setup a repository with the existing code.
• Study the existing code base and incrementally add features needed!

• Each tool is contained into the Crash Analysis Toolbox
  • Crash Analysis Toolbox.pyt - the python code that defines the tool box
  • Crash Analysis Toolbox.Tool.pyt.xml - the xml file that contains the metadata + item descriptions of the toolbox.
• Each tool is defined in their own <>.pyt for the classes of the each tool, the <>.xml for their metadata + item descriptions. Some of the tools have helper classes.

• Project Sponsor:
  Ghazan Khan, Ph.D
  Assistant Professor, Transportation Engineering
  Department of Civil Engineering
  CALIFORNIA STATE UNIVERSITY, SACRAMENTO
  6000 J Street, Sacramento, CA 95819-6029
  

• Development Team:
  "Bus Drivers" - CSUS Computer Science Senior Project Team
  CSC 190/191 - Fall 2015/Spring 2016
  Ben Botto
  Kian Faroughi
  Kenneth Spence
  Victor Zepeda
  Austin Purcell
  Kevin Choe