Release v0.1.0 (initial)

Remove obsolete GitHub workflows and update documentation for Python version; Enhance examples and performance settings in utilities.

Author: selimfirat

.github/workflows/publish-pypi.yml

@@ -8,7 +8,7 @@ version: 2
 build:
   os: ubuntu-22.04
   tools:
-    python: "3.11" # Updated to a more recent Python version
+    python: "3.11"
     # You can also specify other tool versions:
     # nodejs: "20"
     # rust: "1.70"
@@ -17,7 +17,6 @@ build:
     pre_build:
       # Run any pre-build steps like checking links or validating structure
       - pip install -e .
-      - python -m scripts.validate_docs
 
 # Build documentation in the "docs/" directory with Sphinx
 sphinx:

.github/workflows/release-drafter.yml

@@ -11,11 +11,11 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 - Future features and improvements will be listed here
 
-## [0.1.0] - 2025-03-01
+## [0.1.0] - 2025-05-09
 
 ### Added
 
-- Initial release of Kaira framework
+- Initial release of Kaira framework v0.1.0
 - Core modules for wireless communication simulation
 - DeepJSCC implementation
 - Channel models and modulation schemes

.github/workflows/update-changelog.yml

@@ -251,8 +251,16 @@ def calculate_ser(received, original_indices):
         detected_indices[(received_real < 0) & (received_imag > 0)] = 2  # -1+1j
         detected_indices[(received_real < 0) & (received_imag < 0)] = 3  # -1-1j
 
+        # Convert to numpy and ensure shapes match
+        original_np = original_indices.numpy()
+
+        # Make sure both arrays have the same length
+        min_length = min(len(detected_indices), len(original_np))
+        detected_indices = detected_indices[:min_length]
+        original_np = original_np[:min_length]
+
         # Calculate error rate
-        errors = detected_indices != original_indices.numpy()
+        errors = detected_indices != original_np
         ser = np.mean(errors)
 
         return ser

.readthedocs.yaml

@@ -357,20 +357,41 @@ def demonstrate_communication_pipeline(model, image, snr_db=10):
 
 # Modulated signal (take first few dimensions to visualize)
 modulated_data = results["modulated"][0].cpu().numpy()
-signal_length = min(100, modulated_data.shape[0])
+
+# For stem plot, use only the first dimension if it's multi-dimensional
+if len(modulated_data.shape) > 1:
+    # Extract just the first 16 values from the first dimension
+    plot_data = modulated_data[0, :16]
+    signal_length = len(plot_data)
+else:
+    # If it's already 1D, take first 16 values
+    plot_data = modulated_data[:16]
+    signal_length = len(plot_data)
 
 plt.subplot(2, 2, 2)
-plt.stem(range(signal_length), modulated_data[:signal_length])
-plt.title("Modulated Signal (First 100 symbols)")
+# Use the prepared data directly
+plt.stem(range(signal_length), plot_data)
+plt.title("Modulated Signal (First 16 symbols)")
 plt.xlabel("Symbol Index")
 plt.ylabel("Amplitude")
 plt.grid(True, alpha=0.3)
 
 # Show received signal (with channel effects)
 received_data = results["received"][0].cpu().numpy()
 
+# For stem plot, use only the first dimension if it's multi-dimensional
+if len(received_data.shape) > 1:
+    # Extract just the first 16 values from the first dimension
+    received_plot_data = received_data[0, :16]
+    received_length = len(received_plot_data)
+else:
+    # If it's already 1D, take first 16 values
+    received_plot_data = received_data[:16]
+    received_length = len(received_plot_data)
+
 plt.subplot(2, 2, 3)
-plt.stem(range(signal_length), received_data[:signal_length], linefmt="r-")
+# Use the prepared data directly
+plt.stem(range(received_length), received_plot_data, linefmt="r-")
 plt.title(f"Received Signal (After {model.channel.__class__.__name__})")
 plt.xlabel("Symbol Index")
 plt.ylabel("Amplitude")
@@ -417,11 +438,20 @@ def demonstrate_communication_pipeline(model, image, snr_db=10):
     # Get actual data size for modulated and received signals
     mod_data = results["modulated"][0].cpu().numpy()
     rec_data = results["received"][0].cpu().numpy()
-    signal_length = min(50, len(mod_data))  # Limit to at most 50 points
+
+    # For stem plot, use only the first dimension if it's multi-dimensional
+    if len(rec_data.shape) > 1:
+        # Extract values from the first dimension
+        rec_plot_data = rec_data[0, :16]
+        signal_length = len(rec_plot_data)
+    else:
+        # If it's already 1D, take first 16 values
+        rec_plot_data = rec_data[:16]
+        signal_length = len(rec_plot_data)
 
     # Show received signal (varies by SNR)
     plt.subplot(1, 4, i + 2)
-    plt.stem(range(signal_length), rec_data[:signal_length], linefmt="r-")
+    plt.stem(range(signal_length), rec_plot_data, linefmt="r-")
     plt.title(f"Received Signal (SNR={snr} dB)")
     plt.xlabel("Symbol Index")
     plt.ylabel("Amplitude")

CHANGELOG.md

@@ -37,7 +37,7 @@
 # -----------------------------------------
 
 # NEW: More granular performance control
-PERFORMANCE_MODE = "balanced"  # Options: "ultra_fast", "fast", "balanced", "detailed"
+PERFORMANCE_MODE = "ultra_fast"  # Options: "ultra_fast", "fast", "balanced", "detailed"
 
 # Configure visualization and computation based on performance mode
 if PERFORMANCE_MODE == "ultra_fast":

examples/channels/plot_fading_channels.py

@@ -17,3 +17,6 @@ isort>=5.12.0
 mypy>=1.0.0
 pre-commit
 coverage>=7.0.0
+seaborn
+ipython
+scikit-learn