Brute force circle sectors with map projections or Triangle
Illustration of the difference between a discrete (flat tile) grid and a continuous triangulated grid.
give values as value
Turn sf m-poly layer into segments, first with tidyr gather/spread join, then with silicate SC. The performance is compelling.
Turn sf layer into segments, first with tidyr gather/spread join, then with silicate SC. The performance is compelling. The first needs to clean up row_per_path.
The silicate TRI and SC models applied to a 3D multipatch shapefile of NYC building footprints.
Polygon constrained triangulation via ear clipping
Plotting contours of the volcano data set with p5
Using V8 to drive the js Mapbox earcut library for polygon ear-clipping triangulation
segments, edges, nodes and arcs in silicate - this doesn't explain much but it's a happy preview of where things can go
simple space-time KDE with stkde
New in ggplot > 2.2.1
silicate finally has a sensible basis for decomposing neigbouring polygonal features to arc-node form
Topology of shared boundaries between neighbouring polygons requirings "conforming elements" where all edges are atomic
a work in progress outline of the basics of raster in R
rgdal now has HDF5 amongst others for the Windows binary build on CRAN, making it more consistent with general facilities across platform
A relief map of North Carolina counties from a few lines of R code. In-development package https://github.com/hypertidy/anglr
Argument for a general common form, a normalization of complex hierarchical structures such as planar shapes
consistent polygon/raster lossless reprojection via mesh plot
gam seems to return AIC in several different ways, unclear what each means
This does my head in because it juggles relational ids of segments and the structure id of the paths they must take to join together in turtle form. It's just a record to show a basic working example, to be tidied up
Draft geo-faceted summary of sea ice changes
library(tidync) f <- tail(raadtools::sstfiles()$fullname, 1) library(dplyr) tab <- tidync(f) %>% hyper_tibble(lon = lon > 50 & lon < 100, lat = lat > -75 & lat < -45) %>% dplyr::select(anom, err, ice, sst, lon, lat) library(tidyr) tablong <- tab %>% gather(variable, measure, -lon, -lat) %>% dplyr::filter(!is.na(measure)) library(ggplot2) ggplot(tablong, aes(x = lon, y = lat, fill = measure)) + geom_raster() + facet_wrap(~variable)
Reprojecting after decomposition to smaller pieces to avoid wrapping issues
Sample output from a 4D ocean model, using tidync to pass tbl_cube data to ggplot2
3D convex hull etc
Decomposition to triangles to apply continuous measures via finite elements
filtrate is a function to define the hyper slab index in a NetCDF
tidync early experiment
Raster extract is amazing for extraction of grid values with cell indexes, this seems under-used in R and it's partly because it's tricky to juggle the outputs of the index-building functions.
Basic georeferencing of the standard grid used for passive microwave sea ice products. As per https://gis.stackexchange.com/questions/238344/georeferencing-netcdf-of-unknown-coordinate-system-using-lan-lon-netcdf
filling between lines is a dark art
Working with ROMS in R, angstroms package
Mapping broad-scale marine stuff
neighbours by join, for all nc counties
neighbours by join
The tidyr::nest tables perform version of normalization, to reduce duplication in data. This document illustrates an "inside-out" version of the nest process to produce two tables.
Simple features topology with ggraph
This document shows a migration of exsting sp code to the 0.2-7 (CRAN now has 0.2-8). The dev version of sf requires significant changes to get the same result.
Albers Equal Area Conic with Antarctica from R, no cartographic catastrophe.
plotting ROMS data with R in a few different ways
Polygon decomposed to small triangles, merged with global topography, plotted with plotly
Build mesh from maps for plotly
A long description of the lack of topology in spatial data structures, and development tools to convert to mesh/network data structures while keeping the database-object framework common to GIS.
A timeline of the bushfire extent on 7 February 1967
Incomplete survey of building and plotting polygons with holes