Indexing with a point extent returns zero dimensional array about dimensionaldata.jl HOT 4 CLOSED

I think you can wrap it with Touches(extent) to get the one pixel? Selecting with an extent with no area would only return an exact point from a Points raster.

from dimensionaldata.jl.

Wrapping it into Touches works for Rasters but not for YAXArray. With a YAXArray DimensionalData gives a range that is of by one and thereby I get an empty range. I suspect that this is coming from the fact, that The dimensions in the YAXArray are all point axes.
the upper one is the dimensions of the YAXArray and below are the dimensions of the Raster:

julia> dims(rxy)
X Sampled{Float64} -28.4029998779297:0.0550011811599821:18.18300056457514 ForwardOrdered Regular Points,
Y Sampled{Float64} -23.4029998779297:0.0550012159753107:21.863000869751005 ForwardOrdered Regular Points,
Ti Sampled{DateTime} DateTime[1996-01-16T12:00:00, …, 1996-12-16T12:00:00] ForwardOrdered Irregular Points

julia> dims(rasxy)
X Sampled{Float64} Float64[-28.4029998779297, -28.347998696769718, …, 18.12799938341516, 18.18300056457514] ForwardOrdered Explicit Intervals{Center},
Y Sampled{Float64} Float64[-23.4029998779297, -23.347998661954392, …, 21.807999653775695, 21.863000869751005] ForwardOrdered Explicit Intervals{Center},
Ti Sampled{DateTime} DateTime[1996-01-16T12:00:00, …, 1996-12-16T12:00:00] ForwardOrdered Explicit Intervals{Center}

Does this mean, that we would have to ensure that the dimensions are Intervals in YAXArrays?

from dimensionaldata.jl.

Im not even sure what it does for points or why its different! It could be a bug, I probably assumed Touches was only for Intervals.

But it does sound like what you have is interval pixels if you want a point to fall somewhere inside a pixel.

You can detect if a netcdf or tif is intervals, and most of the time it usually is.

from dimensionaldata.jl.

Think I will close this, its really just YAXArrays using Points when it has Intervals

from dimensionaldata.jl.