wolfhece 2.2.15__py3-none-any.whl → 2.2.17__py3-none-any.whl

wolfhece/assets/mesh.py

@@ -0,0 +1,807 @@
+from matplotlib import pyplot as plt
+from matplotlib.axes import Axes
+from matplotlib.figure import Figure
+import numpy as np
+
+from ..wolf_array import header_wolf
+from ..wolfresults_2D import q_splitting, u_splitting, splitting_rule
+
+class Mesh2D(header_wolf):
+
+    def __init__(self, src_header:header_wolf):
+
+        self.set_origin(src_header.origx, src_header.origy)
+        self.set_resolution(src_header.dx, src_header.dy)
+        self.set_translation(src_header.translx, src_header.transly)
+        self.shape = src_header.shape
+        self._factor = None
+
+    def plot_cells(self, ax:Axes=None, transpose:bool= False, color='black', **kwargs):
+        """ Plot the grid of the mesh.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if transpose:
+
+            # plot the grid of the mesh in a transposed way
+            for x in np.linspace(xmin, xmax, endpoint=True, num=self.nbx + 1):
+                ax.plot([ymin, ymax], [x, x], color=color, **kwargs)
+
+            for y in np.linspace(ymin, ymax, endpoint=True, num=self.nby + 1):
+                ax.plot([y, y], [xmin, xmax], color=color, **kwargs)
+
+            self.set_aspect_labels_matrice(ax=ax, **kwargs)
+
+        else:
+
+            # plot the grid of the mesh
+            for y in np.linspace(ymin, ymax, endpoint=True, num=self.nby + 1):
+                ax.plot([xmin, xmax], [y, y], color=color, **kwargs)
+
+            for x in np.linspace(xmin, xmax, endpoint=True, num=self.nbx + 1):
+                ax.plot([x, x], [ymin, ymax], color=color, **kwargs)
+
+            self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax
+
+    def plot_center_cells(self, ax:Axes=None, color='black', linestyle='--', **kwargs):
+        """ Plot lines centered to the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            ax.plot([xmin, xmax], [y, y], color=color, linestyle=linestyle, **kwargs)
+
+        for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+            ax.plot([x, x], [ymin, ymax], color=color, linestyle=linestyle, **kwargs)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+        return fig, ax
+
+    def set_ticks_as_dxdy(self, ax:Axes=None, **kwargs):
+        """ Set the ticks of the axis as the dx and dy of the mesh.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        x_ticks = ['ox + tx', r'+$\Delta x$'] + [f'+{i}$\Delta x$' for i in range(2,self.nbx+1)]
+        y_ticks = ['oy + ty', r'+$\Delta y$'] + [f'{+i}$\Delta y$' for i in range(2,self.nby+1)]
+
+        ax.set_xticks(np.linspace(xmin, xmax, endpoint=True, num=self.nbx + 1))
+        ax.set_yticks(np.linspace(ymin, ymax, endpoint=True, num=self.nby + 1))
+        ax.set_xticklabels(x_ticks)
+        ax.set_yticklabels(y_ticks)
+
+        return fig, ax
+
+    def set_ticks_as_matrice(self, ax:Axes=None, Fortran_type:bool = True, **kwargs):
+        """ Set the ticks of the axis as the row and column of a matrice """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if Fortran_type:
+            x_ticks = [f'{i}' for i in range(1,self.nbx+1)]
+            y_ticks = [f'{i}' for i in range(1,self.nby+1)]
+        else:
+            x_ticks = [f'{i}' for i in range(self.nbx)]
+            y_ticks = [f'{i}' for i in range(self.nby)]
+
+        ax.set_yticks(np.linspace(xmin+self.dx/2., xmax-self.dx/2., endpoint=True, num=self.nbx))
+        ax.set_xticks(np.linspace(ymin+self.dy/2., ymax-self.dy/2., endpoint=True, num=self.nby))
+        
+        x_ticks.reverse()
+        ax.set_yticklabels(x_ticks)
+        ax.set_xticklabels(y_ticks)
+
+        self.set_aspect_labels_matrice(ax=ax, **kwargs)
+
+        return fig, ax
+
+    def plot_circle_at_centers(self, ax:Axes=None, color='black', radius:float=None, **kwargs):
+        """ Plot circles at the center of the cells.
+        """
+
+        if radius is None:
+            radius = 0.1 * self.dx
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+                circle = plt.Circle((x, y), radius=radius, color=color, **kwargs)
+                ax.add_artist(circle)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax
+
+    def plot_indices_at_centers(self, ax:Axes=None, Fortran_type:bool = True, **kwargs):
+        """ Plot the indices of the cells at the center of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+                i,j = self.xy2ij(x, y)
+
+                if Fortran_type:
+                    i+=1
+                    j+=1
+
+                ax.text(x, y, f'({i},{j})', horizontalalignment='center', verticalalignment='center', **kwargs)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax
+    
+    def plot_memoryposition_at_centers(self, ax:Axes=None, 
+                                       transpose=False, 
+                                       Fortran_type:bool = True, 
+                                       f_contiguous:bool = True,
+                                       **kwargs):
+        """ Plot the position of the cells at the center of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if transpose:
+            k = 0
+            if Fortran_type:
+                k+=1
+
+            all_y = list(np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx))
+            all_x = list(np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby))
+
+            all_y.reverse()
+
+            if f_contiguous:
+
+                for x in all_x:
+                    for y in all_y:
+
+                        ax.text(x, y, f'{k}', horizontalalignment='center', verticalalignment='center', **kwargs)
+                        k+=1
+            else:
+
+                for y in all_y:
+                    for x in all_x:
+
+                        ax.text(x, y, f'{k}', horizontalalignment='center', verticalalignment='center', **kwargs)
+                        k+=1
+                        
+            self.set_aspect_labels_matrice(ax=ax, **kwargs)
+
+        else:
+            
+            k = 0
+            if Fortran_type:
+                k+=1
+
+            for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+                for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+
+                    ax.text(x, y, f'{k}', horizontalalignment='center', verticalalignment='center', **kwargs)
+                    k+=1
+
+            self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax
+
+    def plot_indices_at_bordersX(self, ax:Axes=None, Fortran_type:bool = True, **kwargs):
+        """ Plot the indices of the cells at the borders of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin, xmax, endpoint=True, num=self.nbx + 1):
+                i,j = self.xy2ij(x, y)
+
+                if Fortran_type:
+                    i+=1
+                    j+=1
+
+                ax.text(x, y, f'({i},{j})', horizontalalignment='center', verticalalignment='center', **kwargs)
+
+    def plot_indices_at_bordersY(self, ax:Axes=None, Fortran_type:bool = True, **kwargs):
+        """ Plot the indices of the cells at the borders of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        for y in np.linspace(ymin, ymax, endpoint=True, num=self.nby + 1):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+                i,j = self.xy2ij(x, y)
+
+                if Fortran_type:
+                    i+=1
+                    j+=1
+
+                ax.text(x, y, f'({i},{j})', horizontalalignment='center', verticalalignment='center', **kwargs)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax
+
+    def plot_Xarrows_at_center(self, ax:Axes=None,
+                               randomize:bool=False,
+                               amplitude:np.ndarray=None,
+                               color='black', **kwargs):
+        """ Plot arrows at the center of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if amplitude is None:
+            amplitude = np.ones((self.nbx, self.nby)) * 0.2 * self.dx
+        else:
+            amplitude = self.scale_amplitude(amplitude)
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+
+                i,j = self.xy2ij(x, y)
+                if randomize:
+                    dx = np.random.uniform(-0.5, 0.5) * self.dx
+                    amplitude[i,j] = dx
+                else:
+                    dx = amplitude[i,j]
+
+                ax.arrow(x - dx/2., y, dx, 0,
+                         head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                         fc=color, ec=color)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax, amplitude
+
+    def plot_Yarrows_at_center(self, ax:Axes=None,
+                               randomize:bool=False,
+                               amplitude:np.ndarray=None,
+                               color='black', **kwargs):
+        """ Plot arrows at the center of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if amplitude is None:
+            amplitude = np.ones((self.nbx, self.nby)) * 0.2 * self.dy
+        else:
+            amplitude = self.scale_amplitude(amplitude)
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+
+                i,j = self.xy2ij(x, y)
+                if randomize:
+                    dy = np.random.uniform(-0.5, 0.5) * self.dy
+                    amplitude[i,j] = dy
+                else:
+                    dy = amplitude[i,j]
+
+                ax.arrow(x, y - dy/2., 0, dy,
+                         head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                         fc=color, ec=color)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax, amplitude
+
+    def plot_Xarrows_at_borders(self, ax:Axes=None,
+                                randomize:bool=False,
+                                amplitudeX:np.ndarray=None,
+                                amplitudeY:np.ndarray=None,
+                                color='black', **kwargs):
+        """ Plot arrows at the borders of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if amplitudeX is None:
+            amplitudeX = np.ones((self.nbx+1, self.nby)) * 0.2 * self.dx
+        else:
+            if self._factor is not None:
+                amplitudeX *= self._factor
+
+        if amplitudeY is None:
+            amplitudeY = np.ones((self.nbx, self.nby+1)) * 0.2 * self.dy
+        else:
+            if self._factor is not None:
+                amplitudeY *= self._factor
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin, xmax, endpoint=True, num=self.nbx + 1):
+
+                i,j = self.xy2ij(x, y)
+                if randomize:
+                    dx = np.random.uniform(-0.5, 0.5) * self.dx
+                    amplitudeX[i,j] = dx
+                else:
+                    dx = amplitudeX[i,j]
+
+                if dx != 0.:
+                    ax.arrow(x - dx/2., y, dx, 0,
+                            head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                            fc=color, ec=color)
+
+        for y in np.linspace(ymin, ymax, endpoint=True, num=self.nby+1):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+
+                i,j = self.xy2ij(x, y)
+                if randomize:
+                    dx = np.random.uniform(-0.5, 0.5) * self.dy
+                    amplitudeY[i,j] = dx
+                else:
+                    dx = amplitudeY[i,j]
+
+                if dx != 0.:
+                    ax.arrow(x - dx, y, dx, 0,
+                            head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                            fc=color, ec=color)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax, amplitudeX, amplitudeY
+
+    def plot_Yarrows_at_borders(self, ax:Axes=None,
+                                randomize:bool=False,
+                                amplitudeX:np.ndarray=None,
+                                amplitudeY:np.ndarray=None,
+                                color='black', **kwargs):
+        """ Plot arrows at the borders of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        if amplitudeX is None:
+            amplitudeX = np.ones((self.nbx+1, self.nby)) * 0.2 * self.dx
+        else:
+            if self._factor is not None:
+                amplitudeX *= self._factor
+
+        if amplitudeY is None:
+            amplitudeY = np.ones((self.nbx, self.nby+1)) * 0.2 * self.dy
+        else:
+            if self._factor is not None:
+                amplitudeY *= self._factor
+
+        for y in np.linspace(ymin, ymax, endpoint=True, num=self.nby + 1):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+
+                i,j = self.xy2ij(x, y)
+                if randomize:
+                    dy = np.random.uniform(-0.5, 0.5) * self.dy
+                    amplitudeY[i,j] = dy
+                else:
+                    dy = amplitudeY[i,j]
+
+                if dy != 0.:
+                    ax.arrow(x, y - dy/2., 0, dy,
+                            head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                            fc=color, ec=color)
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin, xmax, endpoint=True, num=self.nbx + 1):
+
+                i,j = self.xy2ij(x, y)
+                if randomize:
+                    dy = np.random.uniform(-0.5, 0.5) * self.dx
+                    amplitudeX[i,j] = dy
+                else:
+                    dy = amplitudeX[i,j]
+
+                if dy != 0.:
+                    ax.arrow(x, y - dy/2., 0., dy,
+                            head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                            fc=color, ec=color)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax, amplitudeX, amplitudeY
+
+    def plot_normal_arrows_at_borders(self, ax:Axes=None, color='black', **kwargs):
+        """ Plot arrows at the borders of the cells.
+        """
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            for x in np.linspace(xmin, xmax, endpoint=True, num=self.nbx+1):
+                ax.arrow(x - 0.1 * self.dx, y, 0.2 * self.dx, 0,
+                         head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                         fc=color, ec=color)
+
+        for y in np.linspace(ymin, ymax, endpoint=True, num=self.nby+1):
+            for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+                ax.arrow(x, y - 0.1 * self.dy, 0., 0.2 * self.dy,
+                         head_width=0.1 * self.dx, head_length=0.1 * self.dy,
+                         fc=color, ec=color)
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+        return fig, ax
+
+    def scale_axes(self, ax:Axes=None, factor:float = 0.1, **kwargs):
+        """ Scale the axes of the plot to fit the data.
+        """
+        # augmenter légèrement la taille visible sans
+        # ajouter de ticks
+        ticksX = ax.get_xticks()
+        ticksY = ax.get_yticks()
+        ticklabelsX = ax.get_xticklabels()
+        ticklabelsY = ax.get_yticklabels()
+
+        deltaX = ticksX[-1] - ticksX[0]
+        deltaY = ticksY[-1] - ticksY[0]
+        ax.set_xlim(ticksX[0] - factor * deltaX, ticksX[-1] + factor * deltaX)
+        ax.set_ylim(ticksY[0] - factor * deltaY, ticksY[-1] + factor * deltaY)
+
+        ax.set_xticks(ticksX)
+        ax.set_yticks(ticksY)
+        ax.set_xticklabels(ticklabelsX) #, rotation=45)
+        ax.set_yticklabels(ticklabelsY) #, rotation=45)
+
+        # remove up border and right border
+        ax.spines['top'].set_visible(False)
+        ax.spines['right'].set_visible(False)
+
+        return ax.figure, ax
+
+    def plot_reconstructed_values_at_borders(self, ax:Axes=None,
+                                             colors=['green', 'blue', 'red', 'brown'],
+                                             radius:float = None, **kwargs):
+        """ Plot 4 small circles on each side of the border.
+        """
+
+        if radius is None:
+            radius = 0.02 * self.dx
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        deltay = self.dy / 6.
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            x = xmin
+            for i in range(4):
+                ax.add_artist(plt.Circle((x+.1*self.dx, y - 1.5*deltay + i*deltay),
+                                         radius=radius, color=colors[-i-1], **kwargs))
+            x = xmax
+            for i in range(4):
+                ax.add_artist(plt.Circle((x-.1*self.dx, y - 1.5*deltay + i*deltay),
+                                         radius=radius, color=colors[-i-1], **kwargs))
+
+            for x in np.linspace(xmin + self.dx, xmax - self.dx, endpoint=True, num=self.nbx-1):
+                for i in range(4):
+                    ax.add_artist(plt.Circle((x+.1*self.dx, y - 1.5*deltay + i*deltay),
+                                             radius=radius, color=colors[-i-1], **kwargs))
+                    ax.add_artist(plt.Circle((x-.1*self.dx, y - 1.5*deltay + i*deltay),
+                                             radius=radius, color=colors[-i-1], **kwargs))
+
+        for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+            y = ymin
+            for i in range(4):
+                ax.add_artist(plt.Circle((x - 1.5*deltay + i*deltay, y+.1*self.dy),
+                                         radius=radius, color=colors[i], **kwargs))
+            y = ymax
+            for i in range(4):
+                ax.add_artist(plt.Circle((x - 1.5*deltay + i*deltay, y-.1*self.dy),
+                                         radius=radius, color=colors[i], **kwargs))
+
+            for y in np.linspace(ymin + self.dy, ymax - self.dy, endpoint=True, num=self.nby-1):
+                for i in range(4):
+                    ax.add_artist(plt.Circle((x - 1.5*deltay + i*deltay, y+.1*self.dy),
+                                             radius=radius, color=colors[i], **kwargs))
+                    ax.add_artist(plt.Circle((x - 1.5*deltay + i*deltay, y-.1*self.dy),
+                                             radius=radius, color=colors[i], **kwargs))
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+    def plot_splitted_values_at_borders(self, ax:Axes=None,
+                                        qx:np.ndarray=None,
+                                        qy:np.ndarray=None,
+                                        colors=['green', 'blue', 'red', 'brown'],
+                                        radius:float = None, **kwargs):
+        """ Plot 4 small circles on each side of the border.
+        """
+
+        split_x = self.zeros_bordersX()
+        split_y = self.zeros_bordersY()
+
+        for i in range(1, self.nbx):
+            for j in range(self.nby):
+                split_x[i,j] = splitting_rule(qx[i-1,j], qx[i,j])
+        for i in range(self.nbx):
+            for j in range(1, self.nby):
+                split_y[i,j] = splitting_rule(qy[i,j-1], qy[i,j])
+
+        if radius is None:
+            radius = 0.02 * self.dx
+
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        deltay = self.dy / 6.
+
+        # X borders
+        for y in np.linspace(ymin + self.dy/2., ymax - self.dy/2., endpoint=True, num=self.nby):
+            x = xmin
+            i,j = self.xy2ij(x, y)
+            for i_unk in range(4):
+                ax.add_artist(plt.Circle((x+.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                         radius=radius, color=colors[-i_unk-1], **kwargs))
+            x = xmax
+            i,j = self.xy2ij(x, y)
+            for i_unk in range(4):
+                ax.add_artist(plt.Circle((x-.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                         radius=radius, color=colors[-i_unk-1], **kwargs))
+
+            for x in np.linspace(xmin + self.dx, xmax - self.dx, endpoint=True, num=self.nbx-1):
+                i,j = self.xy2ij(x, y)
+
+                pond = split_x[i,j]
+                if pond == 0.5:
+                    for i_unk in range(4):
+                        ax.add_artist(plt.Circle((x+.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                                radius=radius, color=colors[-i_unk-1], **kwargs))
+                        ax.add_artist(plt.Circle((x-.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                                radius=radius, color=colors[-i_unk-1], **kwargs))
+                elif pond == 1.0:
+                    # left is upstream
+                    for i_unk in [1,2]:
+                        ax.add_artist(plt.Circle((x-.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                                radius=radius, color=colors[-i_unk-1], **kwargs))
+                    for i_unk in [0,3]:
+                        ax.add_artist(plt.Circle((x+.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                                radius=radius, color=colors[-i_unk-1], **kwargs))
+                elif pond == 0.0:
+                    # right is upstream
+                    for i_unk in [0,3]:
+                        ax.add_artist(plt.Circle((x-.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                                radius=radius, color=colors[-i_unk-1], **kwargs))
+                    for i_unk in [1,2]:
+                        ax.add_artist(plt.Circle((x+.1*self.dx, y - 1.5*deltay + i_unk*deltay),
+                                                radius=radius, color=colors[-i_unk-1], **kwargs))
+
+        # Y borders
+        for x in np.linspace(xmin + self.dx/2., xmax - self.dx/2., endpoint=True, num=self.nbx):
+            y = ymin
+            i,j = self.xy2ij(x, y)
+            for i_unk in range(4):
+                ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y+.1*self.dy),
+                                         radius=radius, color=colors[i_unk], **kwargs))
+            y = ymax
+            i,j = self.xy2ij(x, y)
+            for i_unk in range(4):
+                ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y-.1*self.dy),
+                                         radius=radius, color=colors[i_unk], **kwargs))
+
+            for y in np.linspace(ymin + self.dy, ymax - self.dy, endpoint=True, num=self.nby-1):
+                i,j = self.xy2ij(x, y)
+                pond = split_y[i,j]
+                if pond == 0.5:
+                    for i_unk in range(4):
+                        ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y+.1*self.dy),
+                                                radius=radius, color=colors[i_unk], **kwargs))
+                        ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y-.1*self.dy),
+                                                radius=radius, color=colors[i_unk], **kwargs))
+                elif pond == 1.0:
+                    for i_unk in [1,2]:
+                        ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y-.1*self.dy),
+                                                radius=radius, color=colors[i_unk], **kwargs))
+                    for i_unk in [0,3]:
+                        ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y+.1*self.dy),
+                                                radius=radius, color=colors[i_unk], **kwargs))
+                elif pond == 0.0:
+                    for i_unk in [0,3]:
+                        ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y-.1*self.dy),
+                                                radius=radius, color=colors[i_unk], **kwargs))
+                    for i_unk in [1,2]:
+                        ax.add_artist(plt.Circle((x - 1.5*deltay + i_unk*deltay, y+.1*self.dy),
+                                                radius=radius, color=colors[i_unk], **kwargs))
+
+
+        self.set_aspect_labels(ax=ax, **kwargs)
+
+    def set_aspect_labels(self, ax:Axes=None, **kwargs):
+        """ Set the aspect of the plot to be equal.
+        """
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        ax.set_aspect('equal')
+        ax.set_xlim(xmin, xmax)
+        ax.set_ylim(ymin, ymax)
+        ax.set_xlabel('X (m)')
+        ax.set_ylabel('Y (m)')
+        return fig, ax
+    
+    def set_aspect_labels_matrice(self, ax:Axes=None, **kwargs):
+        """ Set the aspect of the plot to be equal.
+        """
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        ax.set_aspect('equal')
+        ax.set_ylim(xmin, xmax)
+        ax.set_xlim(ymin, ymax)
+        ax.set_xlabel('columns')
+        ax.set_ylabel('rows')
+
+        #set x ais on the upper side
+        ax.xaxis.set_ticks_position('top')
+        ax.xaxis.set_label_position('top')
+        return fig, ax
+
+    def zeros(self):
+        """ Return a 2D array of zeros with the shape of the mesh.
+        """
+        return np.zeros((self.nbx, self.nby))
+
+    def ones(self):
+        """ Return a 2D array of ones with the shape of the mesh.
+        """
+        return np.ones((self.nbx, self.nby))
+
+    def zeros_bordersX(self):
+        """ Return a 2D array of zeros with the shape of the mesh + 1 in x direction.
+        """
+        return np.zeros((self.nbx+1, self.nby))
+
+    def zeros_bordersY(self):
+        """ Return a 2D array of zeros with the shape of the mesh + 1 in y direction.
+        """
+        return np.zeros((self.nbx, self.nby+1))
+
+    def ones_bordersX(self):
+        """ Return a 2D array of ones with the shape of the mesh + 1 in x direction.
+        """
+        return np.ones((self.nbx+1, self.nby))
+
+    def ones_bordersY(self):
+        """ Return a 2D array of ones with the shape of the mesh + 1 in y direction.
+        """
+        return np.ones((self.nbx, self.nby+1))
+
+    def apply_splitting_X(self, q:np.ndarray):
+        """ Apply the splitting rule to the X direction.
+        """
+        q_borderX = self.zeros_bordersX()
+        for i in range(1,self.nbx):
+            for j in range(self.nby):
+                q_borderX[i,j] = q_splitting(q[i-1,j], q[i,j])
+        return q_borderX
+
+    def apply_splitting_Y(self, q:np.ndarray):
+        """ Apply the splitting rule to the Y direction.
+        """
+        q_borderY = self.zeros_bordersY()
+        for i in range(self.nbx):
+            for j in range(1,self.nby):
+                q_borderY[i,j] = q_splitting(q[i,j-1], q[i,j])
+        return q_borderY
+
+    def scale_amplitude(self, amplitude:np.ndarray, factor:float = None):
+        """ Scale the amplitude of the arrows.
+        """
+
+        if self._factor is not None:
+            return amplitude * self._factor
+
+        if factor is not None:
+                return amplitude * factor
+
+        factor = min(self.dx, self.dy) * .5 / np.max(amplitude)
+        self._factor = factor
+
+        return amplitude * factor
+
+    def plot_outside_domain(self, ax:Axes=None, color='black', **kwargs):
+        """ Plot a hashed zone outside of the domain.
+        """
+        if ax is None:
+            fig, ax = plt.subplots()
+        else:
+            fig = ax.figure
+
+        [xmin, xmax], [ymin, ymax] = self.get_bounds()
+
+        # Create a rectangle outside the domain
+        rect = plt.Rectangle((xmin-self.dx, ymin-self.dy), xmax - xmin + 2.*self.dx, ymax - ymin + 2.*self.dy,
+                     color=color, alpha=0.2, hatch='//', **kwargs)
+        ax.add_patch(rect)
+
+        rect2 = plt.Rectangle((xmin, ymin), xmax - xmin, ymax - ymin,
+                     color='white', alpha=1., hatch='', **kwargs)
+        ax.add_patch(rect2)
+
+        return fig, ax