geostat_engine.py¶

Overview¶

This file contains 34 documented elements.

Classes¶

GeostatEngine¶

Motore geostatistico avanzato con ML e analisi composizionale

Methods¶

init(self)¶

No description available. Initializes a new GeostatEngine instance by setting up its core attributes. An empty dictionary models and an empty dictionary cache are created to store models and cached data respectively. The ml_available and plotly_available flags are set from the module-level constants ML_AVAILABLE and PLOTLY_AVAILABLE, indicating the availability of optional ML and Plotly dependencies.

calculate_variogram(self, points, values, max_distance, model_type, check_anisotropy)¶

Calcola variogramma con supporto anisotropia

ordinary_kriging(self, points, values, extent, pixel_size, variogram_params)¶

Ordinary Kriging implementation - optimized and robust version

co_kriging(self, primary_points, primary_values, secondary_data, extent, pixel_size)¶

Co-Kriging multivariabile

spatiotemporal_kriging(self, space_time_data, target_time, extent)¶

Kriging spazio-temporale

compositional_analysis(self, points, compositions, transform_type)¶

Analisi composizionale con trasformazioni log-ratio

prepare_ml_features(self, layers)¶

Prepara features per Machine Learning da layer QGIS

extract_labels(self, training_layer)¶

Estrai etichette da layer di training

ml_pattern_recognition(self, layers_data, method)¶

Pattern recognition con Machine Learning

train_ml_model(self, features, labels, method)¶

Training modello ML supervisionato

unsupervised_ml(self, features, method)¶

Clustering o anomaly detection non supervisionato

create_tiles(self, extent, tile_size, overlap_percent)¶

Crea tiles con overlap per batch processing

extract_points_in_tile(self, layer, tile, max_points)¶

Estrae punti in un tile specifico

extract_points_in_tile_with_field(self, layer, tile, field_name, max_points)¶

Estrae punti in un tile con campo specifico

kriging_tile(self, tile_data, tile)¶

Esegue kriging su un singolo tile

merge_tiles(self, tile_results, extent)¶

Unisce tiles processati in raster finale

optimal_sampling_design(self, existing_points, boundary_geom, n_new, method)¶

Calcola design campionamento ottimale

create_interactive_variogram(self, variogram_data)¶

Crea grafico variogramma interattivo con Plotly

create_interactive_kriging_map(self, x, y, predictions, variances, data_points)¶

Crea mappa kriging interattiva con Plotly

create_ml_visualization(self, features, labels, method_name)¶

Visualizzazione risultati Machine Learning

compositional_clr_transform(self, comp_data)¶

Centered Log-Ratio transformation per dati composizionali

compositional_ilr_transform(self, comp_data)¶

Isometric Log-Ratio transformation per dati composizionali

DummyBBox¶

No description available. A locally-defined surrogate bounding box class used as a fallback when no boundary geometry is available. It stores minimum and maximum coordinate values along both axes and exposes them through xMinimum(), xMaximum(), yMinimum(), yMaximum(), width(), and height() methods, mirroring the interface of a standard geometry bounding box object. Instances are initialised with explicit xmin, xmax, ymin, and ymax values; width() and height() are computed as the difference between the respective maximum and minimum values.

Methods¶

init(self, xmin, xmax, ymin, ymax)¶

Initializes a stratified sampling instance on a regular grid. If boundary_geom is not provided, a DummyBBox object is constructed internally using the minimum and maximum coordinate values derived from the data, exposing xMinimum, xMaximum, yMinimum, yMaximum, and width accessors. The DummyBBox serves as a fallback bounding box when no explicit boundary geometry is supplied.

xMinimum(self)¶

No description available. Returns the minimum x-coordinate of the bounding extent. This value corresponds to the _xmin field set during initialization. No parameters are accepted and the stored numeric value is returned directly.

xMaximum(self)¶

No description available. Returns the maximum x-coordinate of the bounding rectangle. This value corresponds to the _xmax field set during object initialisation. Use this method to retrieve the right-hand boundary of the extent.

yMinimum(self)¶

No description available. Returns the minimum Y-coordinate value of the bounding box. This value corresponds to the _ymin instance attribute set during initialization.

yMaximum(self)¶

No description available. Returns the maximum Y-coordinate value of the bounding box. This value is stored in the internal _ymax attribute and represents the upper boundary of the bounding box along the Y axis.

width(self)¶

No description available. Returns the width of the bounding box as a numeric value. It is calculated by subtracting _xmin from _xmax, representing the horizontal extent of the bounding box. No parameters are accepted and no exceptions are documented in the source.

height(self)¶

No description available. Returns the height of the bounding box as a numeric value. It is computed by subtracting the minimum y-coordinate (_ymin) from the maximum y-coordinate (_ymax). This method provides the vertical extent of the bounding box along the y-axis.

Functions¶

objective(params)¶

No description available. An inner objective function used for variogram model fitting that computes the sum of squared residuals between observed semivariances and model-predicted values. It accepts a parameter vector params containing nugget, sill, and range_val, and returns a penalty value of 1e10 if any parameter violates physical constraints (negative nugget, sill less than nugget, or non-positive range). Otherwise, it evaluates the specified variogram model via self._evaluate_model and returns the total squared error, which is minimized by the enclosing minimize call using the L-BFGS-B method.

Parameters: - params

log(msg)¶

No description available. Appends a timestamped message to a log file located at ~/Desktop/geoarchaeo_kriging_log.txt. Each entry is prefixed with the current time formatted as HH:MM:SS.ffffff and followed by a newline, with the file buffer flushed immediately after writing. All exceptions raised during the write operation are silently suppressed.

Parameters: - msg