Skip to contents

Getting Started with riskworkflowr

Source: vignettes/getting-started.Rmd
getting-started.Rmd

Overview

riskworkflowr provides reproducible workflows for spatial risk analysis, including:

  • spatial assignment
  • event aggregation
  • risk metrics
  • SMR analysis
  • cartographic outputs

The package is designed to support consistent workflows across:

  • administrative boundaries
  • custom areal units
  • H3 hexagonal grids

Why riskworkflowr?

Many spatial risk workflows require analysts to repeatedly combine:

  • spatial joins
  • event aggregation
  • comparative risk metrics
  • choropleth mapping
  • reproducible analytical workflows

riskworkflowr aims to provide a consistent framework for these commonly repeated tasks while remaining compatible with the broader R spatial ecosystem.

Core workflow 

point events
→ spatial assignment
→ aggregation/counts
→ risk metrics
→ mapping

A core design principle of the package is:

"units are units"

The same analytical workflow should operate consistently across:

  • administrative boundaries
  • custom polygons
  • H3 hexagonal grids
  • other areal unit systems

Install package 

pak::pak("GeoRiskExplorer/riskworkflowr")

Load package

Create example data 

data <- data.frame(
  event_count = c(5, 10, 15),
  population = c(1000, 2000, 3000)
)

Calculate rates 

risk_calc_rate(
  data = data,
  count_col = "event_count",
  denominator_col = "population"
)
##   event_count population rate_per_10000
## 1           5       1000             50
## 2          10       2000             50
## 3          15       3000             50

Calculate SMR 

risk_calc_smr(
  data = data,
  observed_col = "event_count",
  denominator_col = "population"
)
##   event_count population expected_count smr smr_lower smr_upper
## 1           5       1000              5   1 0.3246973  2.333666
## 2          10       2000             10   1 0.4795389  1.839036
## 3          15       3000             15   1 0.5596924  1.649348
##             smr_ci_flag
## 1 not_clearly_different
## 2 not_clearly_different
## 3 not_clearly_different

Methodological scope

The package primarily focuses on practical and reproducible workflows for exploratory spatial risk analysis and communication.

The implemented methods should not be interpreted as replacing more advanced epidemiological, spatial statistical, or inferential modelling approaches where such methods are appropriate and supported by the available data.

Important assumptions

This package implements practical and reproducible spatial risk workflows commonly used in operational and applied analytical contexts.

Alternative epidemiological and spatial statistical approaches may be more appropriate depending on:

  • data availability
  • denominator quality
  • covariate availability
  • inferential objectives
  • spatial scale
  • analytical assumptions

Next steps

Additional vignettes provide detail on:

  • spatial assignment
  • aggregation workflows
  • SMR analysis
  • H3 workflows
  • mapping
  • methodological considerations