Skip to content

Home

Jump to bottom
Sandro Sousa edited this page Jun 9, 2017 · 11 revisions

Using the Segreg plugin

Welcome to the Segreg wiki!

Important: It is highly recommended to check the change log for new versions, it may contain important bug fixes and improvements. If so, please update the plugin on QGIS Plugin Manager and make sure to always use the latest version: https://github.com/sandrofsousa/Segreg/releases

By the end of this short tutorial you will be ready to use the plugin and explore its measures.

Before installing, make sure you have allowed experimental plugins in QGIS settings, this can be done by clicking on Show also experimental plugins on the following path:

Plugins → Manage and install plugins → Settings

After installed, the plugin will be available at Vector menu.

Step 1. Select Attributes - Required

After importing a valid shapefile (with a projected CRS), select the attributes (groups) and data ID field as desired clicking on Confirm Selection, this process will populate internal variables to use this fields values to compute the measures.
Select the attributes and confirm before moving to the next tab, it will only be available after conforming the groups. A message in QGIS canvas will be displayed informing that the groups were selected with success.

Step 2. Compute population intensity - Optional

Run the population intensity for a spatial version of the results. Select the weight method and bandwidth in meters for neighborhood weighting. This process computes a distance matrix, so this can take a while dependending on the number of tracts in your data. For a shapefile containing 30k tracts it usually takes 2 minutes.

If the population intensity is computed (clicking on Run intensity), these results will automatically be used as input for calculating the measures. In case a non-spatial version is desired, just ignore this step (leaving it empty) and go to the next tab.

A message QGIS canvas will be prompted showing the shape of the matrix and success for computing the population intensity. The shape means the number of lines and columns of your data, for instance (460, 3) in the following picture.

Step 3. Select Measures

Select the measures desired and click Compute measures. All of them can be selected by clicking Select all.

The complexity of dependencies between local and global variables is handled internally, select the measures according to how you want to see the results.

A success message will be displayed at the end of the process.

Step 4. Select output file

Select the name and local directory for saving the results on the ... prompt. The global values will be saved on a different file in the same directory with the name ending by _global.txt.

By checking Add result to canvas, a new shapefile using the same name will be created including data from the input layer and the results from measures.

Differently from the CSV files, the new layer is created on the memory only, so make sure you save it before closing QGIS.

The content of the layer is the same from the CSV file, it can be checked by selecting Open Attribute Table on QGIS.

Output file with results (csv)

The output is compounded of two files, one containing the local results and a second one with the global measures. The number of columns will vary according to the measures selected for compute. The first columns are made of ID, x/y coordinates and the selected groups, the other columns can be:

  • intens_0: The populational intensity for group 0 (selected groups)
  • iso_00: Isolation of group 0 to itself
  • exp_01: exposure of group 0 to group 1
  • dissimil: Dissimilarity index for each locality (e.g. census tract)
  • entropy: Entropy index for each locality
  • indexh: H Index for each locality

Global results will be displayed on a similar way (according to what was selected for compute), but with each measure as a single value in the following layout:

For Isolation/Exposure, results are displayed as a matrix with all groups combined:

0, 0 = isolation index of group 0
0, 1 = exposure index of group 0 to 1

The shape of the matrix is based on the number of groups, for instance 3 groups will result in a matrix of 3 x 3 values as in the previous picture:

[ [G0,G0    G0,G1    G0,G2]
  [G1,G0    G1,G1    G1,G2]
  [G2,G0    G2,G1    G2,G2] ]

Additional info

Report bugs to https://github.com/sandrofsousa/Segreg/issues

Make sure to be running always the latest version, you can download the latest stable version directly from QGIS. For new features and beta versions Fork the Github repository or download the zip file and add it manually to your local plugin's path. Contributions are also welcome!

Disclaimer: This plugin is experimental and can be unstable, the results are not guaranteed.

References

  • Feitosa, F. F., Camara, G., Monteiro, A. M. V., Koschitzki, T., & Silva, M. P. (2007). Global and local spatial indices of urban segregation. International Journal of Geographical Information Science, 21(3), 299-323.

  • Iceland, J. (2004). The multigroup entropy index (also known as Theil’s H or the information theory index). US Census Bureau. Retrieved July, 31, 2006.

Clone this wiki locally