#34 refactor: update plotting imports and utilize PlottingUtils for consistent style setup

Author: selimfirat

docs/examples/constraints/index.rst

@@ -27,7 +27,7 @@ Constraint handling and optimization techniques for communication systems design
 
 .. raw:: html
 
-    <div class="sphx-glr-thumbcontainer" tooltip="This example demonstrates how to combine multiple constraints in Kaira to satisfy complex signal requirements. We'll explore the composition utilities and see how constraints can be sequentially applied to meet practical transmission specifications.">
+    <div class="sphx-glr-thumbcontainer" tooltip="ax.text(0.5, 0.5, 'Spectral Mask Constraint Effects (Visualization placeholder)', ha='center', va='center', transform=ax.transAxes, fontsize=14) ax.set_title('Spectral Mask Constraint Effects', fontsize=16, fontweight='bold') plt.show()=============================================================================================================================================================== This example demonstrates how to combine multiple constraints in Kaira to satisfy complex signal requirements. We'll explore the composition utilities and see how constraints can be sequentially applied to meet practical transmission specifications.">
 
 .. only:: html
 

examples/channels/plot_awgn_channel.py

@@ -15,26 +15,22 @@
 # -------------------------------
 # We start by importing the necessary modules and setting up the environment.
 
+import matplotlib.pyplot as plt
 import numpy as np
 import torch
 
-from examples.example_utils.plotting import (
-    plot_signal_noise_comparison,
-    plot_snr_psnr_comparison,
-    plot_snr_vs_mse,
-    setup_plotting_style,
-)
 from kaira.channels import AWGNChannel
 from kaira.metrics.image import PSNR
 from kaira.metrics.signal import SNR
 from kaira.utils import snr_to_noise_power
+from kaira.utils.plotting import PlottingUtils
 
 # Set random seed for reproducibility
 torch.manual_seed(42)
 np.random.seed(42)
 
 # Configure plotting style
-setup_plotting_style()
+PlottingUtils.setup_plotting_style()
 
 # %%
 # Create Sample Signal
@@ -115,7 +111,55 @@
 # Compare the original clean signal with signals processed through
 # AWGN channels at different SNR levels to observe noise effects.
 
-fig = plot_signal_noise_comparison(t, signal, outputs, measured_metrics, "AWGN Channel Effects on Signal Transmission")
+fig, axes = plt.subplots(2, 2, figsize=(15, 10), constrained_layout=True)
+fig.suptitle("AWGN Channel Effects on Signal Transmission", fontsize=16, fontweight="bold")
+
+# Plot original and noisy signals
+ax1 = axes[0, 0]
+ax1.plot(t, signal, "b-", linewidth=2, label="Original Signal", alpha=0.8)
+for i, (snr_db, output) in enumerate(outputs[:3]):  # Show first 3 for clarity
+    color = PlottingUtils.MODERN_PALETTE[i % len(PlottingUtils.MODERN_PALETTE)]
+    ax1.plot(t, output, "--", color=color, linewidth=1.5, alpha=0.7, label=f"SNR: {snr_db} dB")
+ax1.set_xlabel("Time")
+ax1.set_ylabel("Amplitude")
+ax1.set_title("Signal Comparison")
+ax1.legend()
+ax1.grid(True, alpha=0.3)
+
+# Plot SNR comparison
+ax2 = axes[0, 1]
+target_snrs = [metric["target_snr_db"] for metric in measured_metrics]
+measured_snrs = [metric["measured_snr_db"] for metric in measured_metrics]
+ax2.plot(target_snrs, measured_snrs, "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=8)
+ax2.plot(target_snrs, target_snrs, "--", color="gray", alpha=0.7, label="Ideal (Target = Measured)")
+ax2.set_xlabel("Target SNR (dB)")
+ax2.set_ylabel("Measured SNR (dB)")
+ax2.set_title("SNR Validation")
+ax2.legend()
+ax2.grid(True, alpha=0.3)
+
+# Plot PSNR values
+ax3 = axes[1, 0]
+psnr_values = [metric["measured_psnr_db"] for metric in measured_metrics]
+ax3.plot(target_snrs, psnr_values, "s-", color=PlottingUtils.MODERN_PALETTE[1], linewidth=2, markersize=8)
+ax3.set_xlabel("Target SNR (dB)")
+ax3.set_ylabel("PSNR (dB)")
+ax3.set_title("PSNR vs Target SNR")
+ax3.grid(True, alpha=0.3)
+
+# Plot noise effects on signal
+ax4 = axes[1, 1]
+for i, (snr_db, output) in enumerate(outputs):
+    noise = output - signal
+    noise_power = np.mean(noise**2)
+    ax4.bar(i, noise_power, color=PlottingUtils.MODERN_PALETTE[i % len(PlottingUtils.MODERN_PALETTE)], alpha=0.7)
+ax4.set_xlabel("Channel Index")
+ax4.set_ylabel("Noise Power")
+ax4.set_title("Noise Power by Channel")
+ax4.set_xticks(range(len(outputs)))
+ax4.set_xticklabels([f"{snr}dB" for snr, _ in outputs], rotation=45)
+ax4.grid(True, alpha=0.3)
+
 fig.show()
 
 # %%
@@ -128,7 +172,29 @@
 # Compare theoretical vs measured SNR values and examine PSNR behavior
 # to validate the AWGN channel model performance.
 
-fig = plot_snr_psnr_comparison(measured_metrics, "SNR and PSNR Validation")
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 5), constrained_layout=True)
+fig.suptitle("SNR and PSNR Validation", fontsize=16, fontweight="bold")
+
+# SNR comparison
+target_snrs = [metric["target_snr_db"] for metric in measured_metrics]
+measured_snrs = [metric["measured_snr_db"] for metric in measured_metrics]
+psnr_values = [metric["measured_psnr_db"] for metric in measured_metrics]
+
+ax1.scatter(target_snrs, measured_snrs, color=PlottingUtils.MODERN_PALETTE[0], s=100, alpha=0.7, label="Measured SNR")
+ax1.plot(target_snrs, target_snrs, "--", color="gray", alpha=0.7, label="Ideal (Target = Measured)")
+ax1.set_xlabel("Target SNR (dB)")
+ax1.set_ylabel("Measured SNR (dB)")
+ax1.set_title("SNR Validation")
+ax1.legend()
+ax1.grid(True, alpha=0.3)
+
+# PSNR vs Target SNR
+ax2.plot(target_snrs, psnr_values, "o-", color=PlottingUtils.MODERN_PALETTE[1], linewidth=2, markersize=8)
+ax2.set_xlabel("Target SNR (dB)")
+ax2.set_ylabel("PSNR (dB)")
+ax2.set_title("PSNR vs Target SNR")
+ax2.grid(True, alpha=0.3)
+
 fig.show()
 
 # %%
@@ -158,7 +224,31 @@
 mse_vals = [mse for _, mse in mse_values]
 signal_power = amplitude**2 / 2
 
-fig = plot_snr_vs_mse(snr_levels, mse_vals, signal_power, "SNR vs Mean Squared Error Analysis")
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 5), constrained_layout=True)
+fig.suptitle("SNR vs Mean Squared Error Analysis", fontsize=16, fontweight="bold")
+
+# Plot MSE vs SNR
+ax1.semilogy(snr_levels, mse_vals, "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=8)
+ax1.set_xlabel("SNR (dB)")
+ax1.set_ylabel("MSE")
+ax1.set_title("Measured MSE vs SNR")
+ax1.grid(True, alpha=0.3)
+
+# Plot theoretical MSE (noise power)
+theoretical_mse = []
+for snr_db in snr_levels:
+    snr_linear = 10 ** (snr_db / 10)
+    theoretical_noise_power = signal_power / snr_linear
+    theoretical_mse.append(theoretical_noise_power)
+
+ax2.semilogy(snr_levels, mse_vals, "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=8, label="Measured MSE")
+ax2.semilogy(snr_levels, theoretical_mse, "--", color=PlottingUtils.MODERN_PALETTE[1], linewidth=2, label="Theoretical MSE")
+ax2.set_xlabel("SNR (dB)")
+ax2.set_ylabel("MSE")
+ax2.set_title("Measured vs Theoretical MSE")
+ax2.legend()
+ax2.grid(True, alpha=0.3)
+
 fig.show()
 
 # %%

examples/channels/plot_binary_channels.py

@@ -18,24 +18,19 @@
 # -------------------------------
 # We start by importing the necessary modules and setting up the environment.
 
+import matplotlib.pyplot as plt
 import numpy as np
 import torch
 
-from examples.example_utils.plotting import (
-    plot_binary_channel_comparison,
-    plot_channel_capacity_analysis,
-    plot_channel_error_rates,
-    plot_transition_matrices,
-    setup_plotting_style,
-)
 from kaira.channels import BinaryErasureChannel, BinarySymmetricChannel, BinaryZChannel
+from kaira.utils.plotting import PlottingUtils
 
 # Set random seed for reproducibility
 torch.manual_seed(42)
 np.random.seed(42)
 
 # Configure plotting style
-setup_plotting_style()
+PlottingUtils.setup_plotting_style()
 
 # %%
 # Generate Binary Data
@@ -167,7 +162,32 @@
 
 # Visualize channel effects
 original_data = binary_data[0].numpy()
-fig = plot_binary_channel_comparison(original_data, channel_outputs, segment_start, segment_length, "Binary Channel Effects Comparison")
+
+# Create binary channel comparison plot
+fig, axes = plt.subplots(2, 2, figsize=(15, 10), constrained_layout=True)
+fig.suptitle("Binary Channel Effects Comparison", fontsize=16, fontweight="bold")
+
+# Plot original data segment
+segment_end = segment_start + segment_length
+ax1 = axes[0, 0]
+ax1.plot(range(segment_length), original_data[segment_start:segment_end], "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=6, label="Original")
+ax1.set_title("Original Binary Data")
+ax1.set_xlabel("Bit Index")
+ax1.set_ylabel("Bit Value")
+ax1.set_ylim(-0.1, 1.1)
+ax1.grid(True, alpha=0.3)
+
+# Plot each channel output
+for i, (channel_name, output, error_prob) in enumerate(channel_outputs[:3]):
+    ax = axes.flat[i + 1]
+    ax.plot(range(segment_length), output[segment_start:segment_end], "o-", color=PlottingUtils.MODERN_PALETTE[(i + 1) % len(PlottingUtils.MODERN_PALETTE)], linewidth=2, markersize=6, label=f"{channel_name} (p={error_prob:.2f})")
+    ax.set_title(f"{channel_name} Output")
+    ax.set_xlabel("Bit Index")
+    ax.set_ylabel("Bit Value")
+    ax.set_ylim(-0.1, 1.1)
+    ax.grid(True, alpha=0.3)
+    ax.legend()
+
 fig.show()
 
 # %%
@@ -185,15 +205,29 @@
 observed_bsc = [err_rate for _, _, err_rate in bsc_outputs]
 
 # Plot BSC error rates
-fig1 = plot_channel_error_rates(error_probs, theoretical_bsc, observed_bsc, ["BSC"], "Binary Symmetric Channel Error Rates")
+fig1, ax1 = plt.subplots(figsize=(10, 6), constrained_layout=True)
+ax1.plot(error_probs, theoretical_bsc, "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=8, label="Theoretical BSC")
+ax1.plot(error_probs, observed_bsc, "s-", color=PlottingUtils.MODERN_PALETTE[1], linewidth=2, markersize=8, label="Observed BSC")
+ax1.set_xlabel("Error Probability")
+ax1.set_ylabel("Error Rate")
+ax1.set_title("Binary Symmetric Channel Error Rates", fontsize=14, fontweight="bold")
+ax1.legend()
+ax1.grid(True, alpha=0.3)
 fig1.show()
 
 # Prepare BEC erasure rate data
 theoretical_bec = erasure_probs  # Theoretical erasure rate equals p
 observed_bec = [erasure_rate for _, _, erasure_rate in bec_outputs]
 
 # Plot BEC erasure rates
-fig2 = plot_channel_error_rates(erasure_probs, theoretical_bec, observed_bec, ["BEC"], "Binary Erasure Channel Erasure Rates")
+fig2, ax2 = plt.subplots(figsize=(10, 6), constrained_layout=True)
+ax2.plot(erasure_probs, theoretical_bec, "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=8, label="Theoretical BEC")
+ax2.plot(erasure_probs, observed_bec, "s-", color=PlottingUtils.MODERN_PALETTE[1], linewidth=2, markersize=8, label="Observed BEC")
+ax2.set_xlabel("Erasure Probability")
+ax2.set_ylabel("Erasure Rate")
+ax2.set_title("Binary Erasure Channel Erasure Rates", fontsize=14, fontweight="bold")
+ax2.legend()
+ax2.grid(True, alpha=0.3)
 fig2.show()
 
 # Prepare Z-Channel error rate data
@@ -203,7 +237,14 @@
 observed_z = [err_rate * p_one for _, _, err_rate in z_outputs]
 
 # Plot Z-Channel error rates
-fig3 = plot_channel_error_rates(z_error_probs, theoretical_z, observed_z, ["Z-Channel"], "Z-Channel Error Rates")
+fig3, ax3 = plt.subplots(figsize=(10, 6), constrained_layout=True)
+ax3.plot(z_error_probs, theoretical_z, "o-", color=PlottingUtils.MODERN_PALETTE[0], linewidth=2, markersize=8, label="Theoretical Z-Channel")
+ax3.plot(z_error_probs, observed_z, "s-", color=PlottingUtils.MODERN_PALETTE[1], linewidth=2, markersize=8, label="Observed Z-Channel")
+ax3.set_xlabel("Error Probability")
+ax3.set_ylabel("Error Rate")
+ax3.set_title("Z-Channel Error Rates", fontsize=14, fontweight="bold")
+ax3.legend()
+ax3.grid(True, alpha=0.3)
 fig3.show()
 
 # %%
@@ -230,7 +271,24 @@
 # Plot transition matrices
 matrices = [("Binary Symmetric Channel", bsc_matrix, p_bsc), ("Binary Erasure Channel", bec_matrix, p_bec), ("Z-Channel", z_matrix, p_z)]
 
-fig4 = plot_transition_matrices(matrices, "Binary Channel Transition Matrices")
+fig4, axes = plt.subplots(1, 3, figsize=(15, 5), constrained_layout=True)
+fig4.suptitle("Binary Channel Transition Matrices", fontsize=16, fontweight="bold")
+
+for i, (name, matrix, p) in enumerate(matrices):
+    ax = axes[i]
+    im = ax.imshow(matrix, cmap=PlottingUtils.MATRIX_CMAP, interpolation="nearest", aspect="auto")
+    ax.set_title(f"{name}\n(p={p:.2f})")
+    ax.set_xlabel("Output")
+    ax.set_ylabel("Input")
+
+    # Add text annotations
+    for row in range(matrix.shape[0]):
+        for col in range(matrix.shape[1]):
+            color = "white" if matrix[row, col] > 0.5 else "black"
+            ax.text(col, row, f"{matrix[row, col]:.2f}", ha="center", va="center", color=color, fontsize=12, fontweight="bold")
+
+    plt.colorbar(im, ax=ax, shrink=0.8)
+
 fig4.show()
 
 # %%
@@ -278,7 +336,7 @@ def calculate_z_capacity(p):
 # Plot capacity analysis
 capacities = {"BSC": bsc_capacities, "BEC": bec_capacities, "Z-Channel": z_capacities}
 
-fig5 = plot_channel_capacity_analysis(p_range, capacities, "Binary Channel Capacity Analysis")
+fig5 = PlottingUtils.plot_capacity_analysis(p_range, capacities, "Binary Channel Capacity Analysis")
 fig5.show()
 
 # %%

examples/channels/plot_fading_channels.py

@@ -17,15 +17,16 @@
 import torch
 from scipy import signal
 
-# Plotting imports
-from examples.example_utils.plotting import setup_plotting_style
 from kaira.channels import AWGNChannel, FlatFadingChannel, PerfectChannel
 from kaira.metrics.signal import BitErrorRate
 from kaira.modulations import QPSKModulator
 from kaira.modulations.utils import calculate_theoretical_ber
 from kaira.utils import snr_to_noise_power
 
-setup_plotting_style()
+# Plotting imports
+from kaira.utils.plotting import PlottingUtils
+
+PlottingUtils.setup_plotting_style()
 
 # %%
 # Imports and Setup