ba_accessibility

Scripts to conduct job accessibility analysis in Buenos Aires

To get started and run the analysis:

1. Install python 3.8 
2. Install pip3: sudo apt-get install python3-pip
3. Install virtualenv: sudo apt install virtualenv
4. Clone ba_accessibility repository: git clone https://github.com/urbansim/ba_accessibility.git
5. Navigate to the main directory: cd ba_accessibility
6. Create a clean python 3.8 environment: virtualenv -p python3.8 env
7. Activate the new environment: source env/bin/activate
8. Install repo requirements: pip3.8 install -r ba_accessibility/requirements.txt
9. Navigate to the internal ba_accessibility folder: cd ba_accessibility
10. Execute the run_analysis script passing appropriate parameters

The parameters of the run_analysis.py script are:

• ud: Update Demographics flag. Must be passed the first time the script is executed and any time the jobs or population data changes.

• ug: Update GTFS flag. Must be passed the first time the script is executed and any time the input GTFS files change.

• p: List of project ids to process separated by spaces. At least one project id must be passed

Example

The example below runs the initial accessibility analysis for project 1

python3.8 run_analysis.py -ud -ug -p 1 

Outputs

The model exports images in png format with baseline accessibility and accessibility change after the project is implemented, for 60 min and 90 min time thresholds. These images can be found in the ba_accessibility/ba_accessibility/results/output_maps path, together with a reuslts shapefile that can be used for postprocessing

Baseline Accessibility, 90 minutes

The model also exports the following results in csv format:

• baseline.csv: accessibility values per hexagonal zone for the baseline scenario
• project_[PROJECT_CODE].csv: accessibility values per hexagonal zone for the scenario with project
• results_all_projects.csv: summary of population weighted accessibility values for the area of influence

Hardware Requirements

The complete analysis for two projects runs in about 8 hours in a machine with 32GB of RAM.