Advanced Spatial Analysis

Description:

Level: Professional (P)

Spatial data analysis allows us to go beyond traditional investigation of data, uncovering relationships and patterns based on where things are. This adds an extra layer of understanding of our data, helping us make inferences and predictions.
 
Spatial data analysis has a wide range of applications across many different sectors. For example, improving access to healthcare, and increasing efficiency in transport networks. Spatial data analysis has grown considerably in popularity due to increased access to spatial data and spatial analysis tools. This online course will give participants the skills to load, explore, visualise, and model spatial data within R.

Learning Outcomes

By the end of the course participants will:

• Be able to load, process, and explore different types of spatial data in R.

• Produce a range of visualisations to help understand and communicate spatial data.

• Analyse and interpret spatial data, including measures of autocorrelation.

• Understand the importance of accounting for spatial dependency when modelling spatial data.

• Use appropriate spatial modelling approaches to uncover relationships and patterns in spatial data.
   

Topics Covered

Day 1: A recap of spatial data, the different types available, and showing how to load and tidy data in R. Ensure everyone is comfortable with this and visualising data to explore underlying patterns. Use this to introduce the concept of autocorrelation/spatial dependency.
 
Day 2: More detailed explanation of geostatistical data, including how to quantify autocorrelation with a histogram. Introduce Gaussian random fields and how they can be used to interpolate data.
 
Day 3: Interpolation/kriging with geostatistical data. Geostatistical models that are available. How to apply, interpret, and communicate geostatistical models to explore this data.
 
Day 4: More detailed explanation of areal data and common dependency structures (e.g. neighbourhood or weighted matrices). Show how to extract dependency structures and explore autocorrelation in the data using Moran’s I. Introduce the concept of spatial modelling using random effects.
 
Day 5: Spatial modelling of areal data using conditional autocorrelation and generalised additive models. How to apply, interpret and diagnose these models. Issues with areal data to be aware of.
 

Target Audience

This course is open to anyone that would like to utilise spatial data in their analysis. This could include PhD students, academics, government or other public sector analysts, etc.

Knowledge Assumed

Participants are expected to be comfortable with loading and tidying data using R, preferably using the Tidyverse packages. They are expected to be able to load, tidy and visualise spatial data, although there will be a short recap of this on the first day of the course. They must have an understanding of regression models. Understanding of random effects modelling is also preferred.
 
Attendees must have access to a laptop or computer for the entirety of the course. They must have R (at least version 4) and RStudio (at least version 2024.01) installed on their machine, and be able to install packages from the online R repository (CRAN). In particular, they should have installed the tidyverse and sf packages, and ensure they can be loaded into RStudio.

Cost:

Prices from £786.00 to £1,090.80 (inc. VAT)

Website and registration:

Region:

Greater London

Keywords:

Quantitative Data Handling and Data Analysis, Spatial modelling, R,

Related publications and presentations from our eprints archive:

Quantitative Data Handling and Data Analysis

Back to the training database