#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ Created on Tue Jul 1 22:02:35 2025 @author: jameslofty """ import numpy as np import pandas as pd import matplotlib.pyplot as plt import seaborn as sns from scipy.stats import gaussian_kde import matplotlib.ticker as ticker from matplotlib.ticker import ScalarFormatter, LogLocator, FormatStrFormatter import glob data = pd.read_excel("synthetic_data/synthetic_data_top25.xlsx") data['CSF'] = data['L3 (cm)'] / np.sqrt(data['L1 (cm)'] * data['L2 (cm)']) # %% ####################################################################### unique_materials = np.array([ 'PO soft', 'PO hard', 'EPS', 'PS', 'Multilayer', 'Glass', 'PET', 'Metal', 'Textiles', 'Other plastics', 'Paper', 'Rubber', 'Wood' ], dtype=object) palette = sns.color_palette('Set2', len(unique_materials)) material_colors = {material: palette[i] for i, material in enumerate(unique_materials)} material_colors['Other plastics'] = 'tomato' material_colors['Rubber'] = '#66bad9' # %% # If percentages are in 0–100 range, convert them to 0–1 data['river_prop'] = data['Average_Percentage_Rivers'] / 100 data['bank_prop'] = data['Average_Percentage_Riverbanks'] / 100 weight_data = data[['Common name', 'river_prop', 'bank_prop']].drop_duplicates() mean_by_category = data.groupby('Common name')['L1 (cm)'].mean() merged = weight_data.merge(mean_by_category, on='Common name') merged.rename(columns={'L1 (cm)': 'mean_size'}, inplace=True) merged['river_prop'] /= merged['river_prop'].sum() merged['bank_prop'] /= merged['bank_prop'].sum() # Reuse the weighted mean sizes river_mean_size = np.dot(merged['river_prop'], merged['mean_size']) bank_mean_size = np.dot(merged['bank_prop'], merged['mean_size']) # Calculate weighted standard deviation river_std = np.sqrt(np.dot(merged['river_prop'], (merged['mean_size'] - river_mean_size) ** 2)) bank_std = np.sqrt(np.dot(merged['bank_prop'], (merged['mean_size'] - bank_mean_size) ** 2)) print(f"Estimated mean size in rivers: {river_mean_size:.2f} cm ± {river_std:.2f} cm") print(f"Estimated mean size on riverbanks: {bank_mean_size:.2f} cm ± {bank_std:.2f} cm") # %% mean_by_category = data.groupby('Common name')['L1 (cm)'].mean() sorted_means = mean_by_category.sort_values(ascending=False) print(sorted_means) # %% ordered_names = ( data.groupby('Common name') .apply(lambda x: x['Average_Percentage_Rivers'].max()) .sort_values(ascending=False) .index ) # Ensure 'Common name' follows this order data['Common name'] = pd.Categorical(data['Common name'], categories=ordered_names, ordered=True) ####################################################################### # --- Assign Colors to Common Names --- common_name_colors = {} for name in data['Common name'].unique(): material = data[data['Common name'] == name]['material'].iloc[0] common_name_colors[name] = material_colors.get(material, 'gray') # fallback to gray if material not found # %% def ridgeplot(data, column, xlabel, x_min, x_max, log, scale, scale_markers): figsize = (1.3 , 7) sns.set_theme(rc={"axes.facecolor": (0, 0, 0, 0), "font.family": "DejaVu Sans"}) g = sns.FacetGrid( data, row="Common name", hue="Common name", palette=common_name_colors, aspect=5, height=0.8, sharey=False, # <-- IMPORTANT ) g.fig.set_size_inches(figsize) # Overall figure size g.map_dataframe(sns.kdeplot, x=column, fill=True, alpha=0.8, log_scale=log, bw_adjust=2) g.map_dataframe(sns.kdeplot, x=column, fill=False, alpha=0.8, log_scale=log, lw = 1, bw_adjust=2, color = "k") for ax, label in zip(g.axes.flat, ordered_names): ax.text(-0.1, 0.1, label, color='black', ha="right", va="center", transform=ax.transAxes) g.fig.subplots_adjust(hspace=-0.6) g.set_titles("") g.set(ylabel='') # g.set(yscale='log') g.set(yticks=[]) g.set(xlabel=xlabel) g.set(xlim=(x_min, x_max)) # g.set(ylim=(0, 5)) for ax in g.axes.flat: ymax = 0.0 # Filled KDEs (PolyCollection) for coll in ax.collections: for path in coll.get_paths(): v = path.vertices if v is not None and len(v): ymax = max(ymax, np.max(v[:, 1])) # Outline KDEs (Line2D) for line in ax.lines: y = np.asarray(line.get_ydata()) if y.size: ymax = max(ymax, np.max(y)) ylim =3 if ymax > ylim: ax.set_ylim(0, ymax + 0.6) else: ymax = ylim ax.set_ylim(0, ymax) for ax in g.axes.flat: ax.grid(False) ax.plot([0, 1], [0.02, 0.02], transform=ax.transAxes, color='black', alpha=0.5, linewidth=0.7, clip_on=False) for ax in g.axes.flat: ax.set_xticks(scale) ax.set_xticklabels(scale_markers, rotation=45) # Rotated for readability ax.xaxis.set_visible(True) tick_height = 0.20 # 20% of the axis height for ax in g.axes.flat: trans = ax.get_xaxis_transform() for tick in scale: ax.plot([tick, tick], [0, tick_height], transform=trans, color='black', ls='--', alpha=0.3, linewidth=1, clip_on=False) # plt.savefig(f"python_figures/{column}.svg", format='svg') ridgeplot(data, 'L1 (cm)', '$L1$ (cm)', 0.08, 500, True, [ 0.1, 1, 10, 100], [ "0.1", "1", '10', '100']) ridgeplot(data, 'L2 (cm)', '$L2$ (cm)', 0.008, 105, True, [0.01, 0.1, 1, 10, 100], ["0.01", "0.1", "1", "10", "100"]) ridgeplot(data, 'L3 (cm)', '$L3$ (cm)', 0.0008, 20, True, [0.001, 0.01, 0.1, 1, 10], ["0.001", '0.01', "0.1", "1", "10" ]) # %% # ridgeplot(data, 'CSF', 'CSF (-)', 0, 1, False, [0, 0.5, 1], ['0', '0.5', '1']) ridgeplot(data, 'Elongation_E', '$EL$ (-)', 0, 1, False, [0, 0.5, 1], ['0', '0.5', '1']) ridgeplot(data, 'Flatness_F', '$FL$ (-)', 0, 1, False, [0, 0.5, 1], ['0', '0.5', '1']) ridgeplot(data, 'Mass (g)', 'Mass (g)', 0.00008, 500, True, [0.0001, 0.01, 1, 100], ['0.0001', '0.01', '1', '100']) ridgeplot(data, 'Volume (cm^3)', 'Volume (cm$^3$)', 0.00008, 10000, True, [0.0001, 0.01 , 1, 100, 10000], ['0.0001', '0.01', '1', '100', '10000']) ridgeplot(data, 'Density (g/cm^3)', 'Density (g/cm$^3$)', 0, 10, True, [0.01, 0.1, 1, 10], ['0.01', '0.1','1', '10']) plt.show()