Convert YAXArrays

This section describes how to convert variables from types of other Julia packages into YAXArrays and vice versa.

::: warning

YAXArrays is designed to work with large datasets that are way larger than the memory. However, most types are designed to work in memory. Those conversions are only possible if the entire dataset fits into memory. In addition, metadata might be lost during conversion.

:::

Convert Base.Array

Convert Base.Array to YAXArray:

using YAXArrays

m = rand(5,10)
a = YAXArray(m)

╭──────────────────────────╮
│ 5×10 YAXArray{Float64,2} │
├──────────────────────────┴──────────────────────────────────── dims ┐
  ↓ Dim_1 Sampled{Int64} Base.OneTo(5) ForwardOrdered Regular Points,
  → Dim_2 Sampled{Int64} Base.OneTo(10) ForwardOrdered Regular Points
├─────────────────────────────────────────────────────────── metadata ┤
  Dict{String, Any}()
├────────────────────────────────────────────────────────── file size ┤ 
  file size: 400.0 bytes
└─────────────────────────────────────────────────────────────────────┘

Convert YAXArray to Base.Array:

m2 = collect(a.data)

5×10 Matrix{Float64}:
 0.385566   0.893873  0.682451  0.808871  …  0.0619354  0.124175  0.515777
 0.443632   0.353103  0.564634  0.704278     0.368022   0.145434  0.293256
 0.575769   0.134637  0.839297  0.265568     0.92303    0.563123  0.886301
 0.316638   0.233516  0.332875  0.474806     0.276348   0.631347  0.677238
 0.0149217  0.229851  0.914996  0.135088     0.94013    0.346661  0.977482

Convert Raster

A Raster as defined in Rasters.jl has a same supertype of a YAXArray, i.e. AbstractDimArray, allowing easy conversion between those types:

using Rasters

lon, lat = X(25:1:30), Y(25:1:30)
time = Ti(2000:2024)
ras = Raster(rand(lon, lat, time))
a = YAXArray(dims(ras), ras.data)

╭────────────────────────────╮
│ 6×6×25 YAXArray{Float64,3} │
├────────────────────────────┴────────────────────────── dims ┐
  ↓ X  Sampled{Int64} 25:1:30 ForwardOrdered Regular Points,
  → Y  Sampled{Int64} 25:1:30 ForwardOrdered Regular Points,
  ↗ Ti Sampled{Int64} 2000:2024 ForwardOrdered Regular Points
├─────────────────────────────────────────────────── metadata ┤
  Dict{String, Any}()
├────────────────────────────────────────────────── file size ┤ 
  file size: 7.03 KB
└─────────────────────────────────────────────────────────────┘

ras2 = Raster(a)

╭──────────────────────────╮
│ 6×6×25 Raster{Float64,3} │
├──────────────────────────┴──────────────────────────── dims ┐
  ↓ X  Sampled{Int64} 25:1:30 ForwardOrdered Regular Points,
  → Y  Sampled{Int64} 25:1:30 ForwardOrdered Regular Points,
  ↗ Ti Sampled{Int64} 2000:2024 ForwardOrdered Regular Points
├─────────────────────────────────────────────────── metadata ┤
  Dict{String, Any}()
├───────────────────────────────────────────────────── raster ┤
  extent: Extent(X = (25, 30), Y = (25, 30), Ti = (2000, 2024))

└─────────────────────────────────────────────────────────────┘
[:, :, 1]
  ↓ →  25          26           27         28         29         30
 25     0.961001    0.00362772   0.31403    0.710498   0.283036   0.640441
 26     0.995719    0.315687     0.723564   0.915848   0.499723   0.464818
 27     0.0412041   0.47859      0.653071   0.766245   0.839447   0.379263
 28     0.291882    0.46657      0.21924    0.119464   0.789704   0.889357
 29     0.711002    0.971304     0.49964    0.998782   0.307888   0.125684
 30     0.382774    0.92273      0.68083    0.473735   0.557147   0.507159

Convert DimArray

A DimArray as defined in DimensionalData.jl has a same supertype of a YAXArray, i.e. AbstractDimArray, allowing easy conversion between those types.

Convert DimArray to YAXArray:

using DimensionalData
using YAXArrayBase

dim_arr = rand(X(1:5), Y(10.0:15.0), metadata = Dict{String, Any}())
a = yaxconvert(YAXArray, dim_arr)

╭─────────────────────────╮
│ 5×6 YAXArray{Float64,2} │
├─────────────────────────┴────────────────────────────────── dims ┐
  ↓ X Sampled{Int64} 1:5 ForwardOrdered Regular Points,
  → Y Sampled{Float64} 10.0:1.0:15.0 ForwardOrdered Regular Points
├──────────────────────────────────────────────────────── metadata ┤
  Dict{String, Any}()
├─────────────────────────────────────────────────────── file size ┤ 
  file size: 240.0 bytes
└──────────────────────────────────────────────────────────────────┘

Convert YAXArray to DimArray:

dim_arr2 = yaxconvert(DimArray, a)

╭─────────────────────────╮
│ 5×6 DimArray{Float64,2} │
├─────────────────────────┴────────────────────────────────── dims ┐
  ↓ X Sampled{Int64} 1:5 ForwardOrdered Regular Points,
  → Y Sampled{Float64} 10.0:1.0:15.0 ForwardOrdered Regular Points
├──────────────────────────────────────────────────────── metadata ┤
  Dict{String, Any}()
└──────────────────────────────────────────────────────────────────┘
 ↓ →  10.0        11.0        12.0       13.0       14.0       15.0
 1     0.360273    0.0664705   0.429435   0.476739   0.36173    0.084243
 2     0.775973    0.119542    0.290053   0.803549   0.436001   0.463916
 3     0.0134931   0.796019    0.956738   0.497975   0.194278   0.610775
 4     0.557964    0.60109     0.195567   0.938961   0.947251   0.149577
 5     0.0338467   0.520137    0.849301   0.415606   0.600694   0.952859

::: info

At the moment there is no support to save a DimArray directly into disk as a NetCDF or a Zarr file.

:::