Redesign the Copulogram class, using matplotlib subplots instead of s…

…eaborn PaiGrid

copulogram/__init__.py

@@ -8,4 +8,4 @@
     "Copulogram",
     "NonParametricModel"
 ]
-__version__ = "0.0.3"
+__version__ = "0.0.4"

copulogram/deprecated_Copulogram.py

@@ -0,0 +1,288 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Copyright (C) EDF 2023
+
+@author: Elias Fekhari
+"""
+
+#%%
+import numpy as np
+import pandas as pd
+import seaborn as sns
+from matplotlib import rc
+import matplotlib.pyplot as plt
+
+# Interactive imports
+from itertools import product
+from bokeh.layouts import gridplot
+from bokeh.models import (BasicTicker, Circle, ColumnDataSource,
+                          DataRange1d, Grid, LassoSelectTool, LinearAxis,
+                          Plot, ResetTool)
+from bokeh.transform import factor_cmap, linear_cmap
+
+
+class Copulogram:
+    """
+    Draws a plot for multivariate distributions.
+    The lower triangle is a matrixplot of the data (without transformation), 
+    while the upper triangle is a matrixplot of the ranked data.
+
+    Parameters
+    ----------
+    data : pd.Dataframe()
+        Input dataset to be plotted. Must be a pandas DataFrame object. 
+        A preprocessing removing every missing data is applied.
+
+    Example
+    --------
+    >>> TBD
+    """
+    def __init__(
+        self, 
+        data, 
+        latex=False
+        ):
+        self.data = data.dropna()
+        self.N = data.shape[0]
+        if latex:
+            rc('font', **{'family': 'Times'})
+            rc('text', usetex=True)
+            rc('font', size=18)# Set the default text font size
+            rc('axes', titlesize=20)# Set the axes title font size
+            rc('axes', labelsize=18)# Set the axes labels font size
+            rc('xtick', labelsize=18)# Set the font size for x tick labels
+            rc('ytick', labelsize=18)# Set the font size for y tick labels
+            rc('legend', fontsize=18)# Set the legend font size
+
+    def draw(self, 
+            color='C0',
+            alpha=1.,
+            hue=None,
+            hue_palette=None,
+            kde_on_marginals=True,
+            quantile_contour_levels=None,
+            save_file=None,
+            marker='o',
+            subplot_size=2.5
+            ):
+        """
+        Draws the copulogram plot with a static or interactive option. 
+
+        Parameters
+        ----------
+        color : string
+            The matplotlib color on every element of the graph as long as "hue" is None.
+        alpha : float
+            The alpha blending value, between 0 (transparent) and 1 (opaque).
+        hue : string
+            Grouping variable that will produce points with different colors. 
+            Can be either categorical or numeric, although color mapping will behave differently in latter case.
+        hue_palette : string
+            Method for choosing the colors to use when mapping the hue semantic. 
+            By default "tab10" for categorical mapping, and "viridis" for continuous mapping.
+        kde_on_marginals : boolean
+            Defines the type of plot on the diagonal. Histogram when 
+            the variable is set to False, kernel density estimation otherwise.
+        quantile_contour_levels : 1-d list of floats
+            When the variable takes a value, the contours of the quantiles 
+            defined by the variable are plotted.
+        save_file : string 
+            When this variable is not None, it saves the plot in the current repertory.
+        marker : string
+            Defines the scatterplots markers according to Matplotlib formalism.
+        subplot_size : float
+            Defines the size of each subplot in inches.
+        
+        Returns
+        -------
+        copulogram : matplotlib.axes.Axes
+            The matplotlib axes containing the plot.
+        """
+        df = self.data.copy(deep=True)
+        df_numeric = df._get_numeric_data()
+        plotted_cols = np.array(df_numeric.columns)
+        plotted_cols = np.delete(plotted_cols, np.where(plotted_cols == hue)).tolist()
+        if hue is None:
+            copulogram = sns.PairGrid(df_numeric, height=subplot_size)
+            if kde_on_marginals:
+                copulogram.map_diag(sns.kdeplot, color=color)
+            else:
+                copulogram.map_diag(sns.histplot, color=color, bins=20)
+
+            if quantile_contour_levels is None:
+                copulogram.map_lower(plt.scatter, color=color, alpha=alpha, marker=marker)
+                temp = (df_numeric.rank() / self.N)
+                xmaxs = df_numeric.max().values
+                xmins = df_numeric.min().values
+                copulogram.data = temp * (xmaxs - xmins) + xmins
+                copulogram = copulogram.map_upper(plt.scatter, color=color, alpha=alpha, marker=marker)
+            else:
+                copulogram.map_lower(sns.kdeplot, levels=quantile_contour_levels, color=color)
+                temp = df_numeric[plotted_cols].rank() / self.N
+                xmaxs = df_numeric[plotted_cols].max().values
+                xmins = df_numeric[plotted_cols].min().values
+                copulogram.data = temp * (xmaxs - xmins) + xmins
+                copulogram.map_upper(sns.kdeplot, levels=quantile_contour_levels, color=color)
+        else : 
+            if hue_palette is None:
+                if df[hue].dtype =='O':
+                    hue_palette = 'tab10'
+                else:
+                    hue_palette='viridis'
+            copulogram = sns.PairGrid(df[plotted_cols + [hue]], hue=hue, palette=hue_palette, height=subplot_size)
+            if kde_on_marginals:
+                if df[hue].dtype =='O':
+                    copulogram.map_diag(sns.kdeplot, color=".3")
+                else:
+                    copulogram.map_diag(sns.kdeplot, hue=None, color=".3")
+            else:
+                if df[hue].dtype =='O':
+                    copulogram.map_diag(sns.histplot, color=".3", bins=20)
+                else:
+                    copulogram.map_diag(sns.histplot, hue=None, color=".3", bins=20)
+
+            if quantile_contour_levels is None:
+                copulogram.map_lower(sns.scatterplot, alpha=alpha, marker=marker)
+                temp = df_numeric[plotted_cols].rank() / self.N
+                xmaxs = df_numeric[plotted_cols].max().values
+                xmins = df_numeric[plotted_cols].min().values
+                temp = temp * (xmaxs - xmins) + xmins
+                temp[hue] = df[hue]
+                copulogram.data = temp
+                copulogram = copulogram.map_upper(sns.scatterplot, alpha=alpha, marker=marker)
+            else:
+                copulogram.map_lower(sns.kdeplot, levels=quantile_contour_levels)
+                temp = df_numeric[plotted_cols].rank() / self.N
+                xmaxs = df_numeric[plotted_cols].max().values
+                xmins = df_numeric[plotted_cols].min().values
+                temp = temp * (xmaxs - xmins) + xmins
+                temp[hue] = df[hue]
+                copulogram.data = temp
+                copulogram = copulogram.map_upper(sns.kdeplot, levels=quantile_contour_levels)
+
+            copulogram.add_legend(title=hue)
+
+        if save_file is not None:
+            plt.savefig(save_file, dpi=300, bbox_inches='tight')
+        return copulogram
+
+    def draw_interactive(self, 
+            color='navy',
+            alpha=1.,
+            hue=None,
+            hue_palette=None,
+            marker='o',
+            subplot_size=5
+            ):
+        """
+        Draws the copulogram plot with a static or interactive option. 
+
+        Parameters
+        ----------
+        TBD
+        
+        Returns
+        -------
+        copulogram : TBD
+        """
+
+        df = self.data.copy(deep=True)
+        df_numeric = df._get_numeric_data()
+        rdf = df_numeric.rank()
+
+        plotted_cols = np.array(df_numeric.columns)
+        plotted_cols = np.delete(plotted_cols, np.where(plotted_cols == hue)).tolist()
+        dim = len(plotted_cols)
+
+        if hue is not None: 
+            df_numeric[hue] = df[hue]
+            rdf[hue] = df[hue]
+        source = ColumnDataSource(data=df_numeric)
+        rsource = ColumnDataSource(data=rdf)
+
+        plot_list = []
+        for i, (y, x) in enumerate(product(plotted_cols, plotted_cols)):
+            # Scatter plot
+            if hue is None: 
+                scatter_color = color
+            else: 
+                if df[hue].dtype =='O':
+                    if hue_palette is None:
+                        hue_palette = "Category10_3"
+                    scatter_color = factor_cmap(hue, hue_palette, sorted(df[hue].unique()))
+                else:
+                    if hue_palette is None:
+                        hue_palette="Spectral6"
+                    scatter_color = linear_cmap(hue, hue_palette, low=df[hue].min(), high=df[hue].max())
+
+
+            circle = Circle(x=x, y=y, fill_alpha=alpha, size=5, line_color=None,
+                            fill_color=scatter_color)
+            # Lower triangle
+            if (i%dim) <= (i//dim): # Column index smaller than row index (i.e., lower triangle)
+                # Define one empty plot
+                p = Plot(x_range=DataRange1d(bounds=(df[x].min(), df[x].max())), y_range=DataRange1d(bounds=(df[y].min(), df[y].max())),
+                        background_fill_color="#fafafa",
+                        border_fill_color="white", width=200, height=200, min_border=subplot_size)
+                r = p.add_glyph(source, circle)
+                # Delete diagonal plot
+                if (i%dim) == (i//dim):
+                    r.visible = False
+                    p.grid.grid_line_color = None
+            # Upper triangle
+            elif (i%dim) > (i//dim): 
+                # Define one empty plot
+                p = Plot(x_range=DataRange1d(bounds=(rdf[x].min(), rdf[x].max())), y_range=DataRange1d(bounds=(rdf[y].min(), rdf[y].max())),
+                        background_fill_color="#fafafa",
+                        border_fill_color="white", width=200, height=200, min_border=5)
+                r = p.add_glyph(rsource, circle)
+            p.x_range.renderers.append(r)
+            p.y_range.renderers.append(r)
+            # First column ticks
+            if i % dim == 0:
+                p.min_border_left = p.min_border + 4
+                p.width += 40
+                yaxis = LinearAxis(axis_label=y)
+                yaxis.major_label_orientation = "vertical"
+                p.add_layout(yaxis, "left")
+                yticker = yaxis.ticker
+            else:
+                yticker = BasicTicker()
+            p.add_layout(Grid(dimension=1, ticker=yticker))
+
+            # Last row ticks
+            if i >= dim * (dim-1):
+                p.min_border_bottom = p.min_border + 40
+                p.height += 40
+                xaxis = LinearAxis(axis_label=x)
+                p.add_layout(xaxis, "below")
+                xticker = xaxis.ticker
+            else:
+                xticker = BasicTicker()
+            p.add_layout(Grid(dimension=0, ticker=xticker))
+            p.add_tools(LassoSelectTool(), ResetTool())
+            plot_list.append(p)
+
+        grid_plot = gridplot(plot_list, ncols=dim)
+        return grid_plot
+
+##TODO:
+# Add docstrings
+# Remove the misleading yticks from the top left plot? Ideally we should add the index ticks on the top left of the plot
+# Add color bar on the interactive method using : https://docs.bokeh.org/en/latest/docs/examples/basic/data/color_mappers.html
+
+
+#%%
+if __name__ == "__main__":
+    #data = sns.load_dataset('iris')
+    import pandas as pd
+    data = pd.read_csv("../examples/data/wind_waves_ANEMOC_1H.csv", index_col=0)
+    data = data.iloc[:1000]
+
+    output = data["U_hub (m/s)"] ** 3 * ((np.pi / 180) * data["θ_wind (deg)"]) + (data["Hs (m)"] ** 2 * data["Tp (s)"]) / ((np.pi / 180) * data["θ_wave_new (deg)"])  
+    data['output'] = np.log10(output)
+
+    copulogram = Copulogram(data)
+    copulogram.draw_interactive(hue="output")
+# %%

examples/iris_dataset.py

@@ -8,12 +8,9 @@
 
 import seaborn as sns
 import copulogram as cp
-from bokeh.io import show
 
 data = sns.load_dataset('iris')
 copulogram = cp.Copulogram(data)
 copulogram.draw(save_file="figures/iris1.jpg")
 copulogram.draw(alpha=0.8, hue='species', kde_on_marginals=False, save_file="figures/iris2.jpg")
-copulogram.draw(hue='species', quantile_contour_levels=[0.2, 0.4, 0.6, 0.8], save_file="figures/iris_contours.jpg")
-interactive_plot =  copulogram.draw_interactive(hue="species")
-show(interactive_plot)
+copulogram.draw(hue='species', quantile_contour_levels=[0.2, 0.4, 0.6, 0.8], save_file="figures/iris_contours.jpg")

examples/ishigami.py

@@ -20,4 +20,4 @@
 data = pd.DataFrame(np.array(X[:plotting_size]), columns=list(X.getDescription()))
 data['Y'] = np.array(Y[:plotting_size])
 copulogram = cp.Copulogram(data)
-copulogram.draw(color='C7', marker='.', alpha=0.5, save_file="figures/ishigami.jpg")
+copulogram.draw(color='C7', marker='.', alpha=0.5, save_file="figures/ishigami.jpg")

examples/wind_waves_dataset.py

@@ -31,9 +31,18 @@
 output = data["$U$"] ** 3 * ((np.pi / 180) * data["$\\theta_{wind}$"]) + (data["$H_s$"] ** 2 * data["$T_p$"]) / ((np.pi / 180) * data["$\\theta_{wave}$"])  
 data['output'] = np.log10(output)
 data.columns = ["$\\theta_{wind}$", "$U$", "$\\theta_{wave}$", "$H_s$", "$T_p$", "output"]
-# Draw static copulogram on data
+
+# Draw copulogram on data
+copulogram = cp.Copulogram(data, latex=True)
+copulogram.draw(hue="output", hue_colorbar="plasma", alpha=alpha, marker="o", kde_on_marginals=False, save_file="figures/wind_waves_woutput.jpg")
+
+# Draw threshold event 
+import seaborn as sns
+from matplotlib.colors import to_rgba
+
+threshold = 4.
+data['is_failed'] = "False"
+data.loc[data[data['output'] > threshold].index, 'is_failed'] = "True"
+color_palette = sns.color_palette([to_rgba('C0', 0.2), to_rgba('C1', 0.9)], as_cmap=True)
 copulogram = cp.Copulogram(data, latex=True)
-copulogram.draw(color="C7", hue="output", alpha=alpha, kde_on_marginals=False, save_file="figures/wind_waves_woutput.jpg", marker="o")
-# Draw interactive copulogram on data
-#interactive_plot = copulogram.draw_interactive(color="C7", hue="output", alpha=0.8)
-#show(interactive_plot)
+copulogram.draw(hue='is_failed', hue_colorbar=color_palette, save_file="figures/wind_waves_threshold.jpg")

setup.py

@@ -46,7 +46,7 @@
           "seaborn",
           "matplotlib",
           "openturns>=1.20", 
-          "bokeh>=3.2",
+          #"bokeh>=3.2",
       ],
     include_package_data=True,
     classifiers=[