Update README.md

README.md

@@ -12,7 +12,7 @@ pipeline_tag: image-classification
 This model predicts on 15 minutes spectrograms if they contain a burst or not, see paper:
 
 
-## Usage
+# Usage
 
 ```bash
 pip install torch torchvision huggingface_hub ecallisto_ng
@@ -39,11 +39,11 @@ from huggingface_hub import hf_hub_download
 from ecallisto_ng.data_download.downloader import get_ecallisto_data
 from ecallisto_ng.data_processing.utils import subtract_constant_background
 from ecallisto_ng.plotting.plotting import plot_spectrogram
+from plotly.io import show
 import torch.nn as nn
 from torchvision import models
 import os
 
-
 # ============================================================================
 # Model Definition
 # ============================================================================
@@ -312,8 +312,6 @@ def main():
     instrument = "Australia-ASSA_01"
     
     print(f"Example prediction on instrument: {instrument}")
-    duration_minutes = (end_time - start_time).total_seconds() / 60
-    print(f"Time window: {start_time} to {end_time} ({duration_minutes:.0f} minutes)\n")
     
     # Load model (downloaded and cached automatically)
     model = load_flaresense_model(device=device)
@@ -322,11 +320,9 @@ def main():
     print(f"Fetching data from e-Callisto...")
     df_dict = get_ecallisto_data(start_time, end_time, instrument)
     
-    if instrument not in df_dict:
-        print(f"Error: No data found for instrument {instrument}")
-        return
     
     df_spectrogram = df_dict[instrument]
+
     print(f"Data shape: {df_spectrogram.shape} (time x frequency)")
     print(f"Time range: {df_spectrogram.index[0]} to {df_spectrogram.index[-1]}")
     print(f"Frequency range: {df_spectrogram.columns[0]:.2f} - {df_spectrogram.columns[-1]:.2f} MHz\n")
@@ -347,21 +343,12 @@ def main():
     
     # Plot and save the spectrogram
     print("\nGenerating spectrogram plot...")
-    df_processed = subtract_constant_background(df_spectrogram)
-    fig = plot_spectrogram(df_processed)
-    
-    # Create output filename
-    burst_label = "burst" if burst_detected else "no_burst"
-    date_str = start_time.strftime("%Y-%m-%d_%H-%M-%S")
-    output_filename = f"{instrument}_{date_str}_{burst_label}.png"
-    output_path = os.path.join("ecallisto_ng", output_filename)
-    
-    # Create directory if it doesn't exist
-    os.makedirs("ecallisto_ng", exist_ok=True)
+    df_processed = subtract_constant_background(df_dict[instrument])
+
     
-    # Save the plot
-    fig.write_image(output_path)
-    print(f"Spectrogram saved to: {output_path}")
+    # Show the plot
+    fig = plot_spectrogram(df_processed)
+    show(fig)
 
 
 if __name__ == "__main__":