Focus-Glue-Context Fisheye Transformations for Spatial Visualization • mapycusmaximus

Focus–Glue–Context (FGC) Fisheye Transformations for R

mapycusmaximus provides geometric transformations for spatial and planar data inspired by focus–context visualization principles.
The package implements the Focus–Glue–Context (FGC) model, which enlarges a region of interest (focus), smoothly compresses its surroundings (glue), and preserves the outer region (context).
These transformations facilitate detailed local inspection while maintaining overall spatial structure.

Overview

Fisheye transformations are a well-established approach to visualizing dense data without losing context (Furnas 1986; Sarkar & Brown 1992).
mapycusmaximus extends these principles to modern spatial data workflows in R, integrating seamlessly with the sf ecosystem and ggplot2 visualization grammar.

Main Features

Focus–Glue–Context transformation: smooth magnification, transition, and preservation of spatial zones.
Full integration with sf: works directly on geometries and automatically manages CRS transformation.
Flexible centers: numeric, geographic (EPSG:4326), normalized, or centroid-based.
Customizable parameters: control zoom, squeeze, and optional angular twist (revolution).
Visualization utilities: compare original and transformed coordinates with plot_fisheye_fgc().
Efficient implementation: vectorized operations and safe polygon re-closure.

Installation

Install the stable version from CRAN:

install.packages("mapycusmaximus")

or install the development version from GitHub:

# install.packages("devtools")
devtools::install_github("Alex-Nguyen-VN/mapycusmaximus")

Quick Start

Basic coordinate transformation

library(mapycusmaximus)

grid <- create_test_grid(range = c(-1, 1), spacing = 0.1)

transformed <- fisheye_fgc(
  coords = grid,
  r_in = 0.34, r_out = 0.5,
  zoom_factor = 1.5,
  squeeze_factor = 0.3
)

plot_fisheye_fgc(grid, transformed, r_in = 0.34, r_out = 0.5)

Spatial data integration

library(sf)
library(ggplot2)

poly <- st_sfc(st_polygon(list(rbind(
  c(0,0), c(1,0), c(1,1), c(0,1), c(0,0)
))), crs = 3857)

fisheye_poly <- sf_fisheye(
  poly,
  r_in = 0.3, r_out = 0.6,
  zoom_factor = 2.0,
  squeeze_factor = 0.25
)

ggplot() +
  geom_sf(data = poly, fill = NA, colour = "grey50") +
  geom_sf(data = fisheye_poly, fill = NA, colour = "red")

The Focus–Glue–Context Model

The transformation divides space into three radial zones:

Zone	Definition	Effect
Focus	`r ≤ r_in`	Enlarged by `zoom_factor`
Glue	`r_in < r ≤ r_out`	Smooth compression controlled by `squeeze_factor`
Context	`r > r_out`	Retains original geometry

Parameter	Description	Default	Range
`r_in`	Focus radius	0.34	> 0
`r_out`	Outer radius	0.50	> `r_in`
`zoom_factor`	Magnification factor	1.5	> 1
`squeeze_factor`	Glue compression	0.3	(0, 1]
`method`	Strategy (`"expand"` / `"outward"`)	`"expand"`	string
`revolution`	Angular twist (radians)	0	any

Coordinate System Handling

sf_fisheye() automatically manages coordinate reference systems:

Geographic data: projected to a suitable UTM/MGA zone (e.g. EPSG:7855 for Victoria, AU).
Projected data: used directly.
Manual override: specify target_crs.

Center specification

Numeric: center = c(x, y) (map units)
Geographic: center_crs = "EPSG:4326"
Normalized: center = c(nx, ny), normalized_center = TRUE
Geometry input: center = sf_object (centroid is used)

Advanced Usage

Using an `sf` object as the center

melb_poly <- suburbs[suburbs$name == "Melbourne", ]
fisheye_vic <- sf_fisheye(vic, center = melb_poly)

Normalized center input

fisheye_vic <- sf_fisheye(vic, center = c(0.2, -0.1), normalized_center = TRUE)

Geographic center

fisheye_cbd <- sf_fisheye(
  vic,
  center = c(144.9631, -37.8136),
  center_crs = "EPSG:4326",
  r_in = 0.2, r_out = 0.5
)

Custom coordinate transformation

shift_and_scale <- function(coords, scale = 2, shift_x = 100) {
  cbind(coords[,1] * scale + shift_x, coords[,2] * scale)
}

transformed <- st_transform_custom(
  sf_obj = vic,
  transform_fun = shift_and_scale,
  args = list(scale = 1.5, shift_x = 500)
)

Supported Geometries

mapycusmaximus supports the following sf geometry types:

POINT,
LINESTRING,
POLYGON, MULTIPOLYGON
Geometry collections

Performance Notes

Set preserve_aspect = FALSE to allow stretching and improve performance.
For large datasets, apply the transformation in spatial chunks.
Stronger compression (squeeze_factor → small) increases computational cost.

References

Furnas, G. W. (1986). Generalized Fisheye Views. Proceedings CHI’86, 16–23.
Sarkar, M., & Brown, M. H. (1992). Graphical Fisheye Views of Graphs. Proceedings CHI’92, 83–91.
Laa, U., Cook, D., & Lee, S. (2020). Burning Sage: Reversing the Curse of Dimensionality in Visualization. arXiv:2009.10979.

Package Dependencies

Imports

sf (≥ 1.0.0)
ggplot2 (≥ 3.0.0)
dplyr

Suggests

rmapshaper, tmap, testthat, knitr, rmarkdown

Citation

citation("mapycusmaximus")

Contributing

Contributions are encouraged. To contribute:

Fork the repository.
Create a feature branch:
```
git checkout -b feature/your-feature
```
Commit and push changes.
Submit a pull request with a concise summary.

Please see CONTRIBUTING.md for details.

License

Licensed under the MIT License.
See LICENSE for the complete text.

Acknowledgements

We acknowledge the developers of sf, ggplot2, and the R spatial community for providing the computational and theoretical foundation enabling this work.
Conceptual inspiration draws from research on focus–context visualization and the burning-sage transformation for enhance exploratory clustering and classification problems through reversing the curse of dimensionality.

mapycusmaximus