Anshul Singhvi

`lazy=true` seems unnecessarily lazy?

12

Currently, `lazy = true` defaults to always creating a FileArray, which means that any read requires re-opening the dataset, and calling GDAL again if using e.g a VRT. Consider this...

open all rasters in `extract` too

1

same thing as the last PR. Since this is inside Rasters it's probably OK to use `open` this way? Assuming that all output arrays are pre-allocated (not views).

https://oscarperpinan.github.io/rastervis/#vectorplot has some awesome visuals, we should: - [ ] make sure we can do all of them in Rasters - [ ] have demos to show that off!

Add tests for unitful `cellarea`

2

Does what it says on the tin. Unfortunately it looks like Unitful can't `convert` to Float64 (see https://github.com/PainterQubits/Unitful.jl/issues/742) so going to Proj is where this fails. It also occurs to...

This is not immediately necessary but would be cool to have - it would simply be special treatment for the z-dimension, as we already have for x and y. Could...

integrate zonal with OnlineStats

5

https://github.com/joshday/OnlineStats.jl `deepcopy` works on any online stat, so we can deepcopy the passed in stats and then `fit!` into them as part of the zonal. This can be in a...

`rasterize` fails when passed single geometry and `to=dims(raster)`

2

MWE: ```julia using Rasters, GeoInterface geom = GeoInterface.Polygon([GeoInterface.LinearRing([(0.0, 0.0), (0.0, 10.0), (10.0, 10.0), (10.0, 0.0), (0.0, 0.0)])]) raster = Raster(ones(11, 11), (X(1:0.1:2), Y(1:0.1:2))) rasterize(geom; to = raster, fill = 1,...

Strange behaviour where `crop` is cropping too much

3

```julia using Rasters, RasterDataSources, ArchGDAL using TiledIteration # easy splitting into tiles dem = Raster(WorldClim{Climate}, :tmin, month = 1) ras = read(view(dem, 1:10, 1:10)) ext = Extent(X=(lookup(ras, X)[2], lookup(ras, X)[7]),...

**Describe the bug** If I try to triangulate this using `triangulate(tuple.(lon, lat))`, then it just keeps running. With `Makie.tricontourf(lon, lat, data)` it terminates with a stack overflow within the triangulation....

I coded up a brief implementation of the [AbstractTrees.jl](https://juliacollections.github.io/AbstractTrees.jl/stable/) interface here: ```julia using AbstractTrees import SpatialIndexing as SI AbstractTrees.children(node::SI.Branch) = SI.children(node) AbstractTrees.nodevalue(node::SI.Branch) = SI.mbr(node) AbstractTrees.children(node::T) where {T