Add MCDA module for spatial multi-criteria decision analysis (#1030)

New xrspatial.mcda subpackage covering the four stages of raster MCDA:

Standardize (6 value functions):
  linear, sigmoidal, gaussian, triangular, piecewise, categorical

Weight derivation:
  AHP eigenvector method with consistency ratio, rank-order weighting
  (ROC, rank sum, reciprocal of ranks)

Combination methods:
  WLC, WPM, OWA, fuzzy overlay (AND/OR/sum/product/gamma), boolean overlay

Validation:
  Constraint masking, one-at-a-time and Monte Carlo sensitivity analysis

All operations are element-wise or reductions across layers, so they work
on numpy, cupy, dask+numpy, and dask+cupy without ArrayTypeFunctionMapping
dispatch. No new dependencies.

148 tests covering correctness, edge cases (degenerate triangles, inf/NaN
propagation, single-criterion datasets, AHP validation, dask graph safety,
weight boundary clamping, MC reproducibility), and a full end-to-end
suitability workflow.

Includes docs reference page, README feature matrix section, and user
guide notebook (35_MCDA.ipynb).

Author: brendancol

README.md

@@ -441,6 +441,23 @@ write_vrt('mosaic.vrt', ['tile1.tif', 'tile2.tif'])  # generate VRT
 
 -------
 
+### **Multi-Criteria Decision Analysis (MCDA)**
+
+| Name | Description | Source | NumPy xr.DataArray | Dask xr.DataArray | CuPy GPU xr.DataArray | Dask GPU xr.DataArray |
+|:----------:|:------------|:------:|:----------------------:|:--------------------:|:-------------------:|:------:|
+| [Standardize](xrspatial/mcda/standardize.py) | Converts criterion rasters to 0-1 suitability scale (linear, sigmoidal, gaussian, triangular, piecewise, categorical) | Standard | ✅️ | ✅️ | ✅️ | ✅️ |
+| [AHP Weights](xrspatial/mcda/weights.py) | Derives criterion weights from pairwise comparisons using the Saaty eigenvector method with consistency ratio | Saaty 1980 | ✅️ | ✅️ | ✅️ | ✅️ |
+| [Rank Weights](xrspatial/mcda/weights.py) | Derives weights from a rank ordering (ROC, rank sum, reciprocal) | Standard | ✅️ | ✅️ | ✅️ | ✅️ |
+| [WLC](xrspatial/mcda/combine.py) | Weighted Linear Combination (fully compensatory weighted sum) | Malczewski 2006 | ✅️ | ✅️ | ✅️ | ✅️ |
+| [WPM](xrspatial/mcda/combine.py) | Weighted Product Model (multiplicative, penalizes low scores) | Standard | ✅️ | ✅️ | ✅️ | ✅️ |
+| [OWA](xrspatial/mcda/combine.py) | Ordered Weighted Averaging with tunable risk attitude | Yager 1988 | ✅️ | ✅️ | ✅️ | ✅️ |
+| [Fuzzy Overlay](xrspatial/mcda/combine.py) | Combines criteria using fuzzy set operators (AND, OR, sum, product, gamma) | Eastman 1999 | ✅️ | ✅️ | ✅️ | ✅️ |
+| [Boolean Overlay](xrspatial/mcda/combine.py) | Combines binary criterion masks using AND/OR logic | Standard | ✅️ | ✅️ | ✅️ | ✅️ |
+| [Constrain](xrspatial/mcda/constrain.py) | Masks exclusion zones from a suitability surface | Standard | ✅️ | ✅️ | ✅️ | ✅️ |
+| [Sensitivity](xrspatial/mcda/sensitivity.py) | Assesses weight stability via one-at-a-time or Monte Carlo perturbation | Standard | ✅️ | ✅️ | ✅️ | ✅️ |
+
+-------
+
 ### **Pathfinding**
 
 | Name | Description | Source | NumPy xr.DataArray | Dask xr.DataArray | CuPy GPU xr.DataArray | Dask GPU xr.DataArray |

docs/source/reference/index.rst

@@ -15,6 +15,7 @@ Reference
    focal
    hydrology
    interpolation
+   mcda
    morphology
    multispectral
    pathfinding

docs/source/reference/mcda.rst

@@ -0,0 +1,45 @@
+.. _reference.mcda:
+
+*********************************************
+Multi-Criteria Decision Analysis (MCDA)
+*********************************************
+
+Standardize
+===========
+.. autosummary::
+    :toctree: _autosummary
+
+    xrspatial.mcda.standardize.standardize
+
+Weights
+=======
+.. autosummary::
+    :toctree: _autosummary
+
+    xrspatial.mcda.weights.ahp_weights
+    xrspatial.mcda.weights.rank_weights
+
+Combination
+===========
+.. autosummary::
+    :toctree: _autosummary
+
+    xrspatial.mcda.combine.wlc
+    xrspatial.mcda.combine.wpm
+    xrspatial.mcda.combine.owa
+    xrspatial.mcda.combine.fuzzy_overlay
+    xrspatial.mcda.combine.boolean_overlay
+
+Constraints
+===========
+.. autosummary::
+    :toctree: _autosummary
+
+    xrspatial.mcda.constrain.constrain
+
+Sensitivity
+===========
+.. autosummary::
+    :toctree: _autosummary
+
+    xrspatial.mcda.sensitivity.sensitivity