The goal of this project is to estimate the education infraestructure gap and recommend optimal locations for new schools in Florianopolis and Pará based on capacity, demand, coverage, and location of current schools. To this we will use UrbanPy that leverages public and open-source data for modelling demographics and accesibility to points of interest.
- Demographics and School data collection
- Data cleaning and aggregation
- Exploratory Data Analysis
- Location Modelling
- Results Visualization
-
Exploratory Data Analysis
Interim Report - 11/12/23
The main goal of this project is to propose a methodology to define the location of new schools in Brazil. This goal have been divided in three sub-goals:
- Estimate the Schools’ Capacity in Brazil.
- Estimate the Schools’ Accessibility in Brazil.
- Determine the optimal location of new schools according to capacity and accessibility.
For this purpose we are using as the following data sources:
- National Census Data from 2010*
- Schools Census Data from 2022
- Education Indicators from 2022
- Meta’s High Resolution Population Density Maps from 2019
*This currently is the most up-to-date census data available for Brazil. Once the 2020 Census Data is available this data can be easily updated through the library GeoBr.
The data pre-processing stage is done with UrbanPy, a library developed by the IDB to process high resolution geospatial data, this process consists in (a) data cleaning, (b) selecting relevant variables from each data source, (b) aggregating all variables in H3 Hexagons of resolution 8, a uniform spatial unit that will allow to apply both fast and replicable geographical analysis algorithms to our data.
To measure School Capacity four variables have been calculated for each school at education levels and sublevels.
- Average number of students per professor
- Average number of students per classroom
- Average number of students per course
- Proportion of professors by effort level
These variables are designed to represent where the schools are at maximum capacity and we will potentially need more schools and where existing schools can augment the number of students and/or professors to improve their service.
On the other hand, School Accessibility is represented by seven variables at the H3 Hexagon level:
- Population by school-age group
- Average income per capita (Reales)
- Travel time to the nearest school by foot
- Travel time to the nearest school by car
- Number of schools at <1km distance
- Number of schools <15 minutes travel time
These variables are designed to represent the areas in the city where there is a low accessibility to school and to verify which population demographics are most affected considering age and income. For example: “30% of the population between 3 to 5 years is located in areas that don’t have schools in a radius of 1km and the travel time duration to the nearest school is 1 hour or more.”
Using the hotspot analysis technique, we could identify areas where School Capacity and Accessibility have statistically low or high values. This allows decision makers to focus on the coldspots (low-low) to promote the location of new schools. Another case is identifying areas with high Capacity and low Accessibility on which improving the current accessibility could mean more that students will be able to attend the schools and receive education.
This section outlines the functionality and workflow of a School Location Tool designed to facilitate strategic decision-making in educational planning. The tool employs a user-friendly interface to guide users through the process, resulting in valuable insights for optimizing school placements within specified regions.
1.1 Upon accessing the tool, a modal card greets the user with a welcome message and an explanation of the project goals.
1.2 The user initiates the process by clicking the "Get Started" button.
1.3 Following this action, the modal card closes, and the user is directed to the main dashboard.
2.1 The dashboard displays a map featuring the microregions of Pará.
2.2 Users can define their Area of Interest (AoI) by clicking on a specific microregion.
2.3 Upon selection, the map automatically centers on the chosen microregion.
2.4 Hexagons and Schools data pertinent to the selected microregion are visually presented.
3.1 Users have the option to delve into the variables related to Accessibility and Capacity.
3.2 Selection and weighting of variables for the SCI (Spatial Composite Index) Accessibility and SCI Capacity are user-controlled.
3.3 Upon finalizing variable selection, a background process calculates the Schools Capacity and Accessibility Indexes for the chosen microregion.
4.1 An automated process generates a Hotspot Analysis based on the Schools Capacity and Accessibility Indexes for the selected microregion.
4.2 The map is dynamically updated to reflect the results of the Hotspot Analysis.
4.3 Users are empowered to download comprehensive results in a zip file, including:
- A CSV file containing the Hexagon dataset with all variables and indexes within the selected microregion.
- A CSV file containing the Schools dataset within the selected microregion.
- A Pandas profile report offering in-depth insights into the Hexagon dataset.
The School Location Tool provides a systematic and efficient approach to aid educational planners in making informed decisions for optimal school placement within specified regions.