Geospatial Health Data: Modeling and Visualization with R-INLA and Shiny Book Club
Welcome
Book club meetings
Pace
1
Geospatial health
1.1
Why?
1.1.1
About the book
1.2
How?
1.2.1
Check/use R4ds Slack
1.2.2
Using Git and Github
1.2.3
Following the flow
1.3
Geospatial health
1.3.1
Disease mapping
1.3.2
Communication of results
1.4
Resources
1.5
Meeting Videos
1.5.1
Cohort 1
2
Spatial data and R packages for mapping
2.1
Spatial data and R packages for mapping
2.1.1
Types of spatial data
2.2
Coordinate reference systems
2.2.1
Geographic coordinate systems
2.2.2
Projected coordinate systems
2.2.3
Setting Coordinate Reference Systems in R
2.3
Shapefiles
2.4
Making maps with R
2.4.1
ggplot2
2.4.2
leaflet
2.4.3
mapview
2.4.4
tmap
2.5
Meeting Videos
2.5.1
Cohort 1
3
Bayesian inference and INLA
3.1
Bayesian inference
3.1.1
A small analogy with R: garden of forking data
3.2
Integrated nested Laplace approximation
3.2.1
Resources:
3.3
Meeting Videos
3.3.1
Cohort 1
4
The R-INLA package
4.1
Linear predictor (LP)
4.2
Inla()
4.3
Priors specification
4.4
Example: mortality rates following surgery
4.4.1
Model
4.4.2
resuts!
4.5
Control variables to compute approximations
4.6
Meeting Videos
4.6.1
Cohort 1
5
Areal data
5.1
Introduction
5.2
Spatial neighborhood matrices
5.3
Standardized incidence ratio
5.4
Spatial small area disease risk estimation
5.4.1
Spatial modeling of lung cancer in Pennsylvania
5.5
Spatio-temporal small area disease risk estimation
5.6
Conclusions
5.7
Meeting Videos
5.7.1
Cohort 1
6
Spatial modeling of areal data. Lip cancer in Scotland
6.1
Data set: wrangling
6.1.1
Map of Scotland counties
6.1.2
Data
6.2
Mapping SIRs
6.3
Modeling
6.3.1
Our model:
6.3.2
neighborhood matrix
6.3.3
Using INLA
6.4
Mapping relative risks
6.5
Exceedance probabilities
6.6
Meeting Videos
6.6.1
Cohort 1
7
Spatio-temporal modeling of areal data. Lung cancer in Ohio
7.1
Data
7.2
Expected cases
7.3
SIRs
7.4
Mapping
7.4.1
Time plots of SIRs
7.5
Modeling
7.5.1
Neighborhood matrix
7.5.2
Inference using INLA
7.6
Mapping relative risks
7.7
Meeting Videos
7.7.1
Cohort 1
8
Geostatistical data
8.1
Geostatistical data
8.1.1
Gaussian Random Fields (GRF)
8.1.2
Stationarity
8.1.3
Usefull covariance functions
8.2
Stochastic partial differntial equation approach
8.3
Spatial modeling of rainfall in Paraná, Brazil
8.3.1
Data
8.3.2
Model
8.4
Meeting Videos
8.4.1
Cohort 1
9
Spatial modeling of geostatistical data. Malaria in The Gambia
9.1
Data & data preparation
9.1.1
data
9.1.2
Prevalence
9.1.3
Transforming coordinates
9.1.4
Mapping prevalence
9.1.5
Environmental covariates
9.2
Modeling
9.2.1
Mesh construction
9.2.2
Building the SPDE model on the mesh
9.2.3
Index set
9.2.4
Projection matrix
9.2.5
Prediction data
9.2.6
Stack with data for estimation and prediction
9.2.7
Model formula
9.3
Mapping malaria prevalence
9.4
Mapping exceedance probabilities
9.5
Meeting Videos
9.5.1
Cohort 1
10
Spatio-temporal modeling of geostatistical data. Air pollution in Spain
10.1
Map
10.1.1
Getting Spain
10.1.2
Filter to only get main territory
10.2
Data!
10.2.1
Getting it
10.2.2
Cleaning it
10.3
Modeling!
10.3.1
Mesh construction
10.3.2
SPDE model on the mesh with penalised complexity
10.3.3
Index
10.3.4
projection matrix:
10.3.5
Grid:
10.3.6
Ap + stack
10.3.7
Model formula:
10.4
Mapping
10.5
More on space time model:
10.6
Meeting Videos
10.6.1
Cohort 1
11
Introduction to R Markdown
11.1
Introduction
11.2
Steps in creating an Rmarkdown documents
11.3
Structure of R Markdown
11.4
Text Formating in R Markdown
11.5
R Code Chunks
11.6
Figures
11.7
Tables
11.8
Examples on how to compose a note in R Markdown
11.9
Meeting Videos
11.9.1
Cohort 1
12
Building a dashboard to visualize spatial data with flexdashboard
12.1
The R package flexdashboard
12.2
R Markdown
12.3
Layout
12.4
Dashboard components
12.5
A dashboard to visualize global air pollution
12.6
Data
12.7
Table using DT
12.8
Map using leaflet
12.9
Histogram using ggplot2
12.10
Demonstration on how to build a flexDashboard in R
12.11
Meeting Videos
12.11.1
Cohort 1
13
Introduction to Shiny
13.1
Introduction
13.2
Structure of a Shiny app
13.3
Inputs
13.4
Outputs
13.5
Inputs, outputs and reactivity
13.6
Examples of Shiny apps
13.7
HTML content
13.8
Layouts
13.9
Sharing Shiny apps
13.10
Meeting Videos
13.10.1
Cohort 1
14
Interactive dashboards with flexdashboard and Shiny
14.1
Packages:
14.1.1
We start making a
flexdashboard
:
14.2
Build an awesome flexdadhboard
14.2.1
Download the data
14.2.2
Build the map
14.3
Add some features with Shiny
14.3.1
Add a table
14.3.2
Make the histogram
14.4
Meeting Videos
14.4.1
Cohort 1
15
Building a Shiny app to upload and visualize spatio-temporal data
15.1
SLIDE 1
15.2
Meeting Videos
15.2.1
Cohort 1
16
Disease surveillance with SpatialEpiApp
16.1
SLIDE 1
16.2
Meeting Videos
16.2.1
Cohort 1
Published with bookdown
Geospatial Health Data: Modeling and Visualization with R-INLA and Shiny Book Club
5.6
Conclusions
Things to consider:
Misaligned Data Problem (MIDP)
Modifiable Areal Unit Problem (MAUP)
ecological inference problem:
aggregation bias
specification bias