Skip to content

How to develop in go‐consequences

Jump to bottom
Will Lehman edited this page Dec 19, 2023 · 1 revision

Step by Step description of how to estimate the direct and indirect loss using the go-consequences for flooding events using the National Structure Inventory.

Step 1: Download and Install Docker Desktop

  • Using the link, download and install the docker desktop on your computer Docker

Step 2: Download and Install Visual Studio Code

  • Using the link, download and install the Visual Studio code on your computer Visual Studio Code

Step 3: Download the Go-Consequence code from GitHub

  • Download the zip file of the go-consequences repository from GitHub from this link. Go-Consequence Package
  • Save in the directory and unzip

downloading source code

Step 4: Open Docker Desktop

  • Open the Docker Desktop application Note: You don't need to do anything else with docker. Vscode uses docker to host the container that is defined in go consequences. So it is a dependency; we are leveraging docker in vs code. Docker needs to be running, which it is now.

IMPORTANT NOTE: All operations below are now in Visual Studio Code

Step 5: Install the Remote Development extension

  • Click the Extensions tab on the left of VS code and search for "Remote Development."
  • Install the package

Remote Development Package Installation

Step 6: Open a remote window

  • Open a remote window using the "greater and lesser" arrow on the bottom left of the screen as shown in Figure below:

Launching the remote environment

  • A dialog box opens, and click the "open folder in a container" option

Opening Folder in Container

  • Navigate to the directory where the go-consequence master file is unzipped from Step 3.
  • There will be a prompt to install the Go language support and all tools. Install the tools
  • You will see the directory opened on the explorer panel in VS code application.

Step 8: Running a simple test to verify the setup is correct

  • Drop your grid in the sub-directory called data
  • The grid is the resultant depth raster from the HEC-RAS analysis
  • Depth raster should be projected to WGS 84 or EPSG 4326. A state plane projection will not work since National Structure Inventory (NSI) data is in EPSG 4326.
  • Go to the compute subdirectory and open the simulation_test.go file from the explorer panel.
  • Change line 38 to represent your depth grid file relative path
  • The path should be the remote container path (NOT your computer desktop path)

Defining a path to a depth grid

Step 9: Running the code and output results

  • Run the script using the Run button
  • The following dialog will be shown in the terminal if the program runs correctly.

Successful compute image

  • The NSI dataset is extracted from the cloud (There is no need to download NSI dataset)
  • The output is saved as a geo package (gpkg) file in the data folder with an extension ".gpkg."

Step 10: Reading the results

  • Read the results by importing the gpkg file to ArcGIS Pro or ArcMap.
  • Load the main.nsi_result under the gpkg. It will show all the NSI datasets on a map.
  • Column "depth" is the depth of water by structure
  • Structure damage and content damage are the loss by structure – Direct Loss

Step 11: Simulation of Indirect Loss

  • Use the function in line 47, "Test_StreamAbstract_FIPS_ECAM"
  • Change the file path in line 49 to the depth grid as in Step 8
  • Use the correct FIPS code in line 53 (highlighted in yello
  • StreamAbstractByFIPS_WithECAM("48201", dfr, nsp, w)
  • The reading results are similar to Direct Loss as in Steps 9 and 10.

Note: Ecam is a computable general equilibrium model, which has been shown to estimate losses below a 5% shock poorly. Effectively what ecam does at losses below 5% is optimize the economy because all economies generally are suboptimal. HAZUS is a better option for indirect loss below 5%.

Clone this wiki locally