Update with latest modules

JuliaEarth · Sep 20, 2023 · 00049f1 · 00049f1
1 parent 0964392
commit 00049f1
Show file tree

Hide file tree

Showing 10 changed files with 41 additions and 109 deletions.
diff --git a/Project.toml b/Project.toml
@@ -9,11 +9,11 @@ Chain = "8be319e6-bccf-4806-a6f7-6fae938471bc"
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
 DensityRatioEstimation = "ab46fb84-d57c-11e9-2f65-6f72e4a7229f"
 Distances = "b4f34e82-e78d-54a5-968a-f98e89d6e8f7"
-GeoClustering = "7472b188-6dde-460e-bd07-96c4bc049f7e"
 GeoStatsBase = "323cb8eb-fbf6-51c0-afd0-f8fba70507b2"
+GeoStatsModels = "ad987403-13c5-47b5-afee-0a48f6ac4f12"
 GeoStatsSolvers = "50e95529-e670-4fa6-84ad-e28f686cc091"
+GeoStatsTransforms = "725d9659-360f-4996-9c94-5f19c7e4a8a6"
 GeoTables = "e502b557-6362-48c1-8219-d30d308dcdb0"
-KrigingEstimators = "d293930c-a38c-56c5-8ebb-12008647b47a"
 LossFunctions = "30fc2ffe-d236-52d8-8643-a9d8f7c094a7"
 Meshes = "eacbb407-ea5a-433e-ab97-5258b1ca43fa"
 PointPatterns = "e61b41b6-3414-4803-863f-2b69057479eb"
@@ -37,11 +37,11 @@ CategoricalArrays = "0.10"
 Chain = "0.5"
 DensityRatioEstimation = "1.0"
 Distances = "0.10"
-GeoClustering = "0.5"
-GeoStatsBase = "0.35"
-GeoStatsSolvers = "0.6"
-GeoTables = "1.6"
-KrigingEstimators = "0.12"
+GeoStatsBase = "0.36"
+GeoStatsModels = "0.1"
+GeoStatsSolvers = "0.7"
+GeoStatsTransforms = "0.1"
+GeoTables = "1.7"
 LossFunctions = "0.11"
 Makie = "0.19"
 Meshes = "0.35"
@@ -52,5 +52,5 @@ ScientificTypes = "3.0"
 TableTransforms = "1.9"
 Tables = "1.10"
 Unitful = "1.15"
-Variography = "0.18"
+Variography = "0.19"
 julia = "1.9"
diff --git a/docs/src/clustering.md b/docs/src/clustering.md
@@ -20,33 +20,16 @@ Consider the following data as an example:
 viz(Ω.geometry, color = Ω.Z)
 ```
 
-We can cluster the data with traditional clustering models from
-[MLJ.jl](https://github.com/alan-turing-institute/MLJ.jl):
-
-```@example clustering
-using MLJ
-
-kmeans = MLJ.@load KMeans pkg=Clustering
-
-𝒞 = cluster(Ω, kmeans(k=50))
-
-viz(𝒞.geometry, color = 𝒞.cluster)
-```
-
-but there is no guarantee that the clusters will consist of contiguous
-volumes in space. Alternatively, we can use the following geostatistical
-models:
-
 ## GHC
 
 ```@docs
 GHC
 ```
 
 ```@example clustering
-𝒞 = cluster(Ω, GHC(20, 1.0))
+𝒞 = Ω |> GHC(20, 1.0)
 
-viz(𝒞.geometry, color = 𝒞.cluster)
+viz(𝒞.geometry, color = 𝒞.CLUSTER)
 ```
 
 ## GSC
@@ -56,9 +39,9 @@ GSC
 ```
 
 ```@example clustering
-𝒞 = cluster(Ω, GSC(50, 2.0))
+𝒞 = Ω |> GSC(50, 2.0)
 
-viz(𝒞.geometry, color = 𝒞.cluster)
+viz(𝒞.geometry, color = 𝒞.CLUSTER)
 ```
 
 ## SLIC
@@ -68,7 +51,7 @@ SLIC
 ```
 
 ```@example clustering
-𝒞 = cluster(Ω, SLIC(50, 0.01))
+𝒞 = Ω |> SLIC(50, 0.01)
 
-viz(𝒞.geometry, color = 𝒞.cluster)
+viz(𝒞.geometry, color = 𝒞.CLUSTER)
 ```
diff --git a/docs/src/index.md b/docs/src/index.md
@@ -97,7 +97,7 @@ coord = [(25.,25.), (50.,75.), (75.,50.)]
 𝒫 = EstimationProblem(𝒟, 𝒢, :Z)
 
 # choose a solver from the list of solvers
-𝒮 = Kriging(:Z => (variogram=GaussianVariogram(range=35.),))
+𝒮 = KrigingSolver(:Z => (variogram=GaussianVariogram(range=35.),))
 
 # solve the problem
 Ω = solve(𝒫, 𝒮)
@@ -131,13 +131,14 @@ The project is split into various packages:
 |:-------:|:------------|
 | [GeoStats.jl](https://github.com/JuliaEarth/GeoStats.jl) | Main package reexporting full stack of packages for geostatistics. |
 | [Meshes.jl](https://github.com/JuliaGeometry/Meshes.jl) | Computational geometry and advanced meshing algorithms. |
+| [GeoTables.jl](https://github.com/JuliaEarth/GeoTables.jl) | Geospatial tables compatible with the framework. |
 | [TableTransforms.jl](https://github.com/JuliaML/TableTransforms.jl) | Transforms and pipelines with tabular data. |
 | [Variography.jl](https://github.com/JuliaEarth/Variography.jl) | Variogram estimation and modeling, and related tools. |
-| [KrigingEstimators.jl](https://github.com/JuliaEarth/KrigingEstimators.jl) | Kriging estimation over general geospatial domains. |
+| [PointPatterns.jl](https://github.com/JuliaEarth/PointPatterns.jl) | Geospatial point pattern analysis and synthesis. |
 | [DensityRatioEstimation.jl](https://github.com/JuliaML/DensityRatioEstimation.jl) | High-performance density ratio estimation. |
 | [LossFunctions.jl](https://github.com/JuliaML/LossFunctions.jl) | Loss functions for (geo)statistical learning. |
-| [PointPatterns.jl](https://github.com/JuliaEarth/PointPatterns.jl) | Geospatial point pattern analysis and synthesis. |
-| [GeoClustering.jl](https://github.com/JuliaEarth/GeoClustering.jl) | Geostatistical clustering (a.k.a. domaining). |
+| [GeoStatsModels.jl](https://github.com/JuliaEarth/GeoStatsModels.jl) | Geostatistical models for geospatial interpolation. |
+| [GeoStatsTransforms.jl](https://github.com/JuliaEarth/GeoStatsTransforms.jl) | Geostatistical transforms for geospatial data. |
 | [GeoStatsSolvers.jl](https://github.com/JuliaEarth/GeoStatsSolvers.jl) | Built-in geostatistical solvers. |
 | [GeoStatsBase.jl](https://github.com/JuliaEarth/GeoStatsBase.jl) | Base package with core definitions. |
 

diff --git a/docs/src/kriging.md b/docs/src/kriging.md
@@ -2,8 +2,8 @@
 
 !!! note
 
-    This section describes the estimators used in the [Kriging](solvers/builtin.md) solver.
-    Most users don't want to use estimators directly because they lack features such as
+    This section describes the models used in the [KrigingSolver](solvers/builtin.md).
+    Most users don't want to use models directly because they lack features such as
     neighborhood search and change of support. If that is your case, please refer to the
     solver documentation.
 
@@ -22,45 +22,14 @@ This package implements the following Kriging variants:
 - Universal Kriging
 - External Drift Kriging
 
-All these variants follow the same interface: an estimator object is first created with a
-given set of parameters, it is then combined with the data to obtain predictions at new
-locations.
+All these variants follow the same interface: an object is first created with a given set
+of parameters, it is then combined with the data to obtain predictions at new geometries.
 
 The `fit` function takes care of building the Kriging system and factorizing the LHS with
-an appropriate decomposition (e.g. Cholesky, LU):
+an appropriate decomposition (e.g. Cholesky, LU), and the `predict` (or `predictprob`)
+function performs the estimation for a given variable and geometry.
 
-```@docs
-fit(::KrigingEstimator, ::Any)
-```
-
-and the `predict` function performs the estimation for a given variable and location:
-
-```@docs
-predict(::KrigingEstimators.FittedKriging, ::Any, ::Any)
-predictprob(::KrigingEstimators.FittedKriging, ::Any, ::Any)
-```
-
-Alternative constructors are provided for convenience that will immediately fit the Kriging
-parameters to the data. In this case, the data is passed as the first argument.
-
-For advanced users, the Kriging weights and Lagrange multipliers at a given location can be accessed
-with the `weights` method. This method returns a `KrigingWeights` object containing a field `λ` for
-the weights and a field `ν` for the Lagrange multipliers:
-
-```@docs
-weights
-```
-
-These weights can be combined with a vector of observations using the `combine` function:
-
-```@docs
-KrigingEstimators.combine
-```
-
-This functionality can be useful in real-time applications when the locations of the observations are
-fixed and an online stream of data is available.
-
-Finally, all estimators work with general Hilbert spaces, meaning that one can interpolate any
+All variants work with general Hilbert spaces, meaning that one can interpolate any
 data type that implements scalar multiplication, vector addition and inner product.
 
 ## Simple Kriging
@@ -100,7 +69,7 @@ or in matricial form ``\C\l = \c``. We subtract the given mean from the observat
 ```
 
 ```@docs
-SimpleKriging
+GeoStatsModels.SimpleKriging
 ```
 
 ## Ordinary Kriging
@@ -134,7 +103,7 @@ location ``\x_0`` are given by:
 ```
 
 ```@docs
-OrdinaryKriging
+GeoStatsModels.OrdinaryKriging
 ```
 
 ## Universal Kriging
@@ -206,7 +175,7 @@ variance at location ``\x_0`` are given by:
 ```
 
 ```@docs
-UniversalKriging
+GeoStatsModels.UniversalKriging
 ```
 
 ## External Drift Kriging
@@ -224,5 +193,5 @@ estimated.
 External drifts are known to cause numerical instability. Give preference to other Kriging variants if possible.
 
 ```@docs
-ExternalDriftKriging
+GeoStatsModels.ExternalDriftKriging
 ```
diff --git a/docs/src/miscellaneous.md b/docs/src/miscellaneous.md
@@ -25,27 +25,6 @@ using GeoStatsImages
 Π = geosplit(𝒟, 0.5, (1.0, 1.0))
 ```
 
-## Trend
-
-```@docs
-trend
-```
-
-```@example misc
-# quadratic + noise
-r = range(-1,stop=1,length=100)
-μ = [x^2 + y^2 for x in r, y in r]
-ϵ = 0.1rand(100,100)
-𝒟 = georef((Z=μ+ϵ,))
-
-ℳ = trend(𝒟, :Z, degree=2)
-
-fig = Mke.Figure(resolution = (800, 400))
-viz(fig[1,1], 𝒟.geometry, color = 𝒟.Z)
-viz(fig[1,2], ℳ.geometry, color = ℳ.Z)
-fig
-```
-
 ## Integrate
 
 ```@docs

diff --git a/docs/src/problems.md b/docs/src/problems.md
@@ -33,11 +33,11 @@ coord = [(25.,25.), (50.,75.), (75.,50.)]
 𝒫 = EstimationProblem(𝒟, 𝒢, :Z)
 ```
 
-Solve the problem with [`Kriging`](@ref) solver:
+Solve the problem with the [`KrigingSolver`](@ref):
 
 ```@example problems
 # ordinary Kriging
-𝒮 = Kriging(:Z => (variogram=GaussianVariogram(range=35.),))
+𝒮 = KrigingSolver(:Z => (variogram=GaussianVariogram(range=35.),))
 
 # perform estimation
 Ω = solve(𝒫, 𝒮)

diff --git a/docs/src/solvers/builtin.md b/docs/src/solvers/builtin.md
@@ -5,9 +5,9 @@ Below is the list of solvers that are readily available after loading the projec
 ## Estimation
 
 ```@docs
-Kriging
-IDW
-LWR
+KrigingSolver
+IDWSolver
+LWRSolver
 ```
 
 ## Simulation

diff --git a/docs/src/solvers/overview.md b/docs/src/solvers/overview.md
@@ -56,9 +56,9 @@ Please consult
 
 | Solver | Description | References |
 |:------:|:------------|:-----------|
-| [`Kriging`](@ref) | Kriging (SK, OK, UK, EDK) | [Matheron 1971](https://books.google.com/books/about/The_Theory_of_Regionalized_Variables_and.html?id=TGhGAAAAYAAJ) |
-| [`IDW`](@ref) | Inverse distance weighting | [Shepard 1968](https://dl.acm.org/citation.cfm?id=810616) |
-| [`LWR`](@ref) | Locally weighted regression | [Cleveland 1979](https://www.jstor.org/stable/2286407) |
+| [`KrigingSolver`](@ref) | Kriging (SK, OK, UK, EDK) | [Matheron 1971](https://books.google.com/books/about/The_Theory_of_Regionalized_Variables_and.html?id=TGhGAAAAYAAJ) |
+| [`IDWSolver`](@ref) | Inverse distance weighting | [Shepard 1968](https://dl.acm.org/citation.cfm?id=810616) |
+| [`LWRSolver`](@ref) | Locally weighted regression | [Cleveland 1979](https://www.jstor.org/stable/2286407) |
 
 ## Simulation
 

diff --git a/docs/src/variography/theoretical.md b/docs/src/variography/theoretical.md
@@ -240,7 +240,7 @@ for (i, θ) in enumerate(θs)
   γ = GaussianVariogram(e)
 
   # solve the problem with Kriging
-  sol = solve(problem, Kriging(:Z => (variogram=γ,)))
+  sol = solve(problem, KrigingSolver(:Z => (variogram=γ,)))
 
   # visualize solution at subplot i
   viz(fig[pos(i)...], sol.geometry, color = sol.Z,

diff --git a/src/GeoStats.jl b/src/GeoStats.jl
@@ -29,9 +29,9 @@ using Reexport
 @reexport using TableTransforms
 @reexport using GeoStatsBase
 @reexport using Variography
-@reexport using KrigingEstimators
 @reexport using PointPatterns
-@reexport using GeoClustering
+@reexport using GeoStatsModels
+@reexport using GeoStatsTransforms
 @reexport using GeoStatsSolvers
 @reexport using Chain: @chain