0.31.8

bartzbeielstein · bartzbeielstein · commit b3ab51b6562c · 2025-08-02T20:43:02.000+02:00
new Spot plot contour
diff --git a/notebooks/00_spotPython_tests.ipynb b/notebooks/00_spotPython_tests.ipynb
diff --git a/pyproject.toml b/pyproject.toml
@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "spotpython"
-version = "0.31.7"
+version = "0.31.8"
 authors = [
   { name="T. Bartz-Beielstein", email="tbb@bartzundbartz.de" }
 ]
diff --git a/src/spotpython/spot/spot.py b/src/spotpython/spot/spot.py
@@ -51,7 +51,7 @@
 from typing import Callable
 from spotpython.utils.numpy2json import NumpyEncoder
 from spotpython.utils.file import load_result
-from spotpython.surrogate.plot import plot_3d_contour
+from spotpython.surrogate.plot import plot_3d_contour, plotkd
 
 # Setting up the backend to use QtAgg
 # matplotlib.use("TkAgg")
@@ -2432,77 +2432,6 @@ def process_z00(self, z00, use_min=True) -> list:
                     result.append(max_value)
         return result
 
-    def plot_contour(
-        self,
-        i=0,
-        j=1,
-        min_z=None,
-        max_z=None,
-        show=True,
-        title=None,
-        filename=None,
-        n_grid=50,
-        contour_levels=10,
-        dpi=200,
-        figsize=(12, 5),
-        use_min=False,
-        use_max=True,
-        tkagg=False,
-        cmap="jet",
-    ) -> None:
-        """
-        Plot the contour and 3D surface for any pair of dimensions of the surrogate model.
-        This method visualizes the surrogate model's predictions over a grid for two selected dimensions.
-        It creates both a filled contour plot and a 3D surface plot, allowing users to inspect the surrogate's
-        response surface. The remaining dimensions are fixed to either their minimum or maximum values, depending
-        on the `use_min` and `use_max` flags.
-
-        Args:
-            i (int, optional): Index of the first dimension to plot. Default is 0.
-            j (int, optional): Index of the second dimension to plot. Default is 1.
-            min_z (float, optional): Minimum value for the color scale (z-axis). If None, determined automatically.
-            max_z (float, optional): Maximum value for the color scale (z-axis). If None, determined automatically.
-            show (bool, optional): Whether to display the plot interactively. Default is True.
-            filename (str, optional): If provided, saves the plot to this file. Default is None.
-            n_grid (int, optional): Number of grid points per dimension. Default is 50.
-            contour_levels (int, optional): Number of contour levels. Default is 10.
-            dpi (int, optional): Dots per inch for saved figure. Default is 200.
-            title (str, optional): Title for the plot. Default is None.
-            figsize (tuple, optional): Figure size in inches (width, height). Default is (12, 6).
-            use_min (bool, optional): If True, fix hidden dimensions to their minimum values. Default is False.
-            use_max (bool, optional): If True, fix hidden dimensions to their maximum values. Default is True.
-            tkagg (bool, optional): If True, use TkAgg backend for matplotlib. Default is False.
-            cmap (str, optional): Colormap to use for the contour plot. Default is "jet".
-
-        Returns:
-            None
-        """
-        X, Y, Z = self.prepare_plot(
-            i=i,
-            j=j,
-            n_grid=n_grid,
-            use_min=use_min,
-            use_max=use_max,
-        )
-        plot_3d_contour(
-            X=X,
-            Y=Y,
-            Z=Z,
-            vmin=min_z if min_z is not None else np.min(Z),
-            vmax=max_z if max_z is not None else np.max(Z),
-            var_name=self.var_name,
-            i=i,
-            j=j,
-            show=show,
-            filename=filename,
-            contour_levels=contour_levels,
-            dpi=dpi,
-            title=title,
-            figsize=figsize,
-            tkagg=tkagg,
-            cmap=cmap,
-        )
-
     def prepare_plot(
         self,
         i=0,
@@ -2522,7 +2451,10 @@ def prepare_plot(
             use_max (bool, optional): If True, fix hidden dimensions to their maximum values. Default is True.
 
         Returns:
-            dict: Dictionary containing X_combined, Y_combined, Z_combined, min_z, max_z.
+            (tuple): A tuple containing:
+                - X (numpy.ndarray): Mesh grid for the first dimension.
+                - Y (numpy.ndarray): Mesh grid for the second dimension.
+                - Z (numpy.ndarray): Surrogate predictions over the grid.
 
         Examples:
             >>> plot_data = S.prepare_plot(i=0, j=1)
@@ -2577,6 +2509,104 @@ def predict_contour_values(X, Y, z0):
 
         return X_combined, Y_combined, Z_combined
 
+    def plot_contour(
+        self,
+        i=0,
+        j=1,
+        min_z=None,
+        max_z=None,
+        show=True,
+        title=None,
+        filename=None,
+        n_grid=50,
+        contour_levels=10,
+        dpi=200,
+        figsize=(12, 5),
+        use_min=False,
+        use_max=True,
+        tkagg=False,
+        cmap="jet",
+    ) -> None:
+        """
+        Plot the contour and 3D surface for any pair of dimensions of the surrogate model.
+        This method visualizes the surrogate model's predictions over a grid for two selected dimensions.
+        It creates both a filled contour plot and a 3D surface plot, allowing users to inspect the surrogate's
+        response surface. The remaining dimensions are fixed to either their minimum or maximum values, depending
+        on the `use_min` and `use_max` flags.
+        """
+        plotkd(model=self.surrogate, X=self.X, y=self.y, i=i, j=j, num=n_grid)
+
+    def plot_contour_old(
+        self,
+        i=0,
+        j=1,
+        min_z=None,
+        max_z=None,
+        show=True,
+        title=None,
+        filename=None,
+        n_grid=50,
+        contour_levels=10,
+        dpi=200,
+        figsize=(12, 5),
+        use_min=False,
+        use_max=True,
+        tkagg=False,
+        cmap="jet",
+    ) -> None:
+        """
+        Plot the contour and 3D surface for any pair of dimensions of the surrogate model.
+        This method visualizes the surrogate model's predictions over a grid for two selected dimensions.
+        It creates both a filled contour plot and a 3D surface plot, allowing users to inspect the surrogate's
+        response surface. The remaining dimensions are fixed to either their minimum or maximum values, depending
+        on the `use_min` and `use_max` flags.
+
+        Args:
+            i (int, optional): Index of the first dimension to plot. Default is 0.
+            j (int, optional): Index of the second dimension to plot. Default is 1.
+            min_z (float, optional): Minimum value for the color scale (z-axis). If None, determined automatically.
+            max_z (float, optional): Maximum value for the color scale (z-axis). If None, determined automatically.
+            show (bool, optional): Whether to display the plot interactively. Default is True.
+            filename (str, optional): If provided, saves the plot to this file. Default is None.
+            n_grid (int, optional): Number of grid points per dimension. Default is 50.
+            contour_levels (int, optional): Number of contour levels. Default is 10.
+            dpi (int, optional): Dots per inch for saved figure. Default is 200.
+            title (str, optional): Title for the plot. Default is None.
+            figsize (tuple, optional): Figure size in inches (width, height). Default is (12, 6).
+            use_min (bool, optional): If True, fix hidden dimensions to their minimum values. Default is False.
+            use_max (bool, optional): If True, fix hidden dimensions to their maximum values. Default is True.
+            tkagg (bool, optional): If True, use TkAgg backend for matplotlib. Default is False.
+            cmap (str, optional): Colormap to use for the contour plot. Default is "jet".
+
+        Returns:
+            None
+        """
+        X, Y, Z = self.prepare_plot(
+            i=i,
+            j=j,
+            n_grid=n_grid,
+            use_min=use_min,
+            use_max=use_max,
+        )
+        plot_3d_contour(
+            X=X,
+            Y=Y,
+            Z=Z,
+            vmin=min_z if min_z is not None else np.min(Z),
+            vmax=max_z if max_z is not None else np.max(Z),
+            var_name=self.var_name,
+            i=i,
+            j=j,
+            show=show,
+            filename=filename,
+            contour_levels=contour_levels,
+            dpi=dpi,
+            title=title,
+            figsize=figsize,
+            tkagg=tkagg,
+            cmap=cmap,
+        )
+
     def plot_important_hyperparameter_contour(
         self,
         threshold=0.0,
diff --git a/src/spotpython/surrogate/plot.py b/src/spotpython/surrogate/plot.py
@@ -34,8 +34,8 @@ def plot1d(model, X: np.ndarray, y: np.ndarray, show: Optional[bool] = True) ->
         raise ValueError("plot1d is only supported for 1D input data.")
 
     _ = plt.figure(figsize=(9, 6))
-    n_grid = 100
-    x = linspace(X[:, 0].min(), X[:, 0].max(), num=n_grid).reshape(-1, 1)
+    num = 100
+    x = linspace(X[:, 0].min(), X[:, 0].max(), num=num).reshape(-1, 1)
     y_pred, y_std = model.predict(x, return_std=True)
 
     plt.plot(x, y_pred, "k", label="Prediction")
@@ -55,30 +55,47 @@ def plot1d(model, X: np.ndarray, y: np.ndarray, show: Optional[bool] = True) ->
         plt.show()
 
 
-def generate_mesh_grid(X: np.ndarray, i: int, j: int, n_grid: int = 100):
+def generate_mesh_grid(
+    X: Optional[np.ndarray] = None,
+    i: int = 0,
+    j: int = 1,
+    num: int = 100,
+    lower: Optional[np.ndarray] = None,
+    upper: Optional[np.ndarray] = None,
+):
     """
-    Generate a mesh grid for two selected dimensions of X, and fill the remaining dimensions with their mean values.
+    Generate a mesh grid for two selected dimensions, filling remaining dimensions with their mean values
+    (if X is given) or the mean of the lower and upper bound (if lower and upper are given).
 
     Args:
-        X (np.ndarray): Input data of shape (n_samples, k).
+        X (np.ndarray, optional): Input data of shape (n_samples, k). Required if lower/upper are not given.
         i (int): Index of the first dimension to vary.
         j (int): Index of the second dimension to vary.
-        n_grid (int): Number of grid points per dimension.
+        num (int): Number of grid points per dimension.
+        lower (np.ndarray, optional): Lower bounds for each dimension (shape (k,)).
+        upper (np.ndarray, optional): Upper bounds for each dimension (shape (k,)).
 
     Returns:
         X_i (np.ndarray): Meshgrid for the i-th dimension.
         X_j (np.ndarray): Meshgrid for the j-th dimension.
-        grid_points (np.ndarray): Grid points of shape (n_grid*n_grid, k) for prediction.
+        grid_points (np.ndarray): Grid points of shape (num*num, k) for prediction.
     """
-    k = X.shape[1]
-    mean_values = X.mean(axis=0)
+    # Check that exactly one of (X) or (lower and upper) is provided
+    if (X is not None and (lower is not None or upper is not None)) or (X is None and (lower is None or upper is None)):
+        raise ValueError("Provide either X or both lower and upper, but not both or neither.")
+
+    if X is not None:
+        k = X.shape[1]
+        mean_values = X.mean(axis=0)
+        x_i = linspace(X[:, i].min(), X[:, i].max(), num=num)
+        x_j = linspace(X[:, j].min(), X[:, j].max(), num=num)
+    else:
+        k = len(lower)
+        mean_values = (np.array(lower) + np.array(upper)) / 2.0
+        x_i = linspace(lower[i], upper[i], num=num)
+        x_j = linspace(lower[j], upper[j], num=num)
 
-    # Create a grid for the two varied dimensions
-    x_i = linspace(X[:, i].min(), X[:, i].max(), num=n_grid)
-    x_j = linspace(X[:, j].min(), X[:, j].max(), num=n_grid)
     X_i, X_j = meshgrid(x_i, x_j)
-
-    # Prepare the grid points for prediction
     grid_points = np.zeros((X_i.size, k))
     grid_points[:, i] = X_i.ravel()
     grid_points[:, j] = X_j.ravel()
@@ -318,7 +335,7 @@ def plotkd(
     max_error: float = 1e-3,
     var_names: Optional[List[str]] = None,
     cmap: str = "jet",
-    n_grid: int = 100,
+    num: int = 100,
     vmin: Optional[float] = None,
     vmax: Optional[float] = None,
     add_points: bool = False,
@@ -338,7 +355,7 @@ def plotkd(
         max_error (float): Maximum error for color scaling. Default is 1e-3.
         var_names (list of str, optional): List of variable names for axis labeling. If None, generic labels are used.
         cmap (str): Colormap for the surface and contour plots. Default is "jet".
-        n_grid (int): Number of grid points per dimension for the mesh grid. Default is 100.
+        num (int): Number of grid points per dimension for the mesh grid. Default is 100.
         vmin (float, optional): Minimum value for the color scale. If None, determined from predictions.
         vmax (float, optional): Maximum value for the color scale. If None, determined from predictions.
         add_points (bool): If True, adds scatter points to the surface and contour plots. Default is False.
@@ -358,7 +375,7 @@ def plotkd(
     """
     k = X.shape[1]
     check_ij(i, j, k)
-    X_i, X_j, grid_points = generate_mesh_grid(X, i, j, n_grid)
+    X_i, X_j, grid_points = generate_mesh_grid(X, i, j, num)
 
     # Predict the values and standard deviations
     y_pred, y_std = model.predict(grid_points, return_std=True)
@@ -491,7 +508,7 @@ def plot_3d_contour(X, Y, Z, vmin, vmax, var_name=None, i=0, j=1, show=True, fil
     Examples:
         >>> # Example 1: Using output from Spot
         >>> # Assume S is a Spot object with a fitted surrogate
-        >>> plot_data = S.prepare_plot(i=0, j=1, n_grid=100)
+        >>> plot_data = S.prepare_plot(i=0, j=1, num=100)
         >>> from spotpython.surrogate.plot import plot_3d_contour
         >>> plot_3d_contour(
         ...     plot_data,

Original file line number	Diff line number	Diff line change
`@@ -7,7 +7,7 @@ build-backend = "setuptools.build_meta"`
`7`	`7`
`8`	`8`	`[project]`
`9`	`9`	`name = "spotpython"`
`10`		`-version = "0.31.7"`
	`10`	`+version = "0.31.8"`
`11`	`11`	`authors = [`
`12`	`12`	`{ name="T. Bartz-Beielstein", email="tbb@bartzundbartz.de" }`
`13`	`13`	`]`