update app.py

app.py

@@ -342,7 +342,7 @@ def ecg_analysis():
 
     # Customize the plot layout
     fig.update_layout(
-        title='Latent space 2D (PCA reduction)',
+        #title='Latent space 2D (PCA reduction)',
         xaxis_title='component 1',
         yaxis_title='component 2'
     )
@@ -372,13 +372,38 @@ def ecg_analysis():
     
     # DECODER RECONSTRUCTION PLOT
     
-    fig_reconstruction = plt.figure(figsize=(10,8))
-    sns.set(font_scale = 2)
+    # fig_reconstruction = plt.figure(figsize=(10,8))
+    # sns.set(font_scale = 2)
+    # sns.set_style("white")
+    # plt.plot(df_data[0], 'black',linewidth=2)
+    # plt.plot(heartbeat_decoder_preds[0], 'red',linewidth=2)
+    # plt.fill_between(np.arange(140), heartbeat_decoder_preds[0], df_data[0], color='lightcoral')
+    # plt.legend(labels=["Input", "Reconstruction", "Error"])
+    
+    fig_reconstruction = go.Figure()
+
+    sns.set(font_scale=2)
     sns.set_style("white")
-    plt.plot(df_data[0], 'black',linewidth=2)
-    plt.plot(heartbeat_decoder_preds[0], 'red',linewidth=2)
-    plt.fill_between(np.arange(140), heartbeat_decoder_preds[0], df_data[0], color='lightcoral')
-    plt.legend(labels=["Input", "Reconstruction", "Error"])
+
+    # Plot 'Input' and 'Reconstruction' lines
+    fig_reconstruction.add_trace(
+        go.Scatter(x=np.arange(140), y=df_data[0], fill=None, mode='lines', name='Input', line=dict(color='black', width=3)))
+    fig_reconstruction.add_trace(
+        go.Scatter(x=np.arange(140), y=heartbeat_decoder_preds[0], fill=None, mode='lines', name='Reconstruction',
+                   line=dict(color='red', width=3)))
+
+    # Create a custom fill area
+    fill_x = list(np.arange(140)) + list(reversed(np.arange(140)))
+    fill_y = list(heartbeat_decoder_preds[0]) + list(reversed(df_data[0]))
+    fig_reconstruction.add_trace(go.Scatter(x=fill_x, y=fill_y, fill='tozeroy', fillcolor='rgba(255, 182, 193, 10.0)', mode='lines', line=dict(color='rgba(255, 182, 193, 0.5)', width=0), name='Error'))
+
+    # Customize the legend's position (outside the graph)
+    fig_reconstruction.update_layout(
+        legend=dict(
+            x=1.1,  # Adjust the x-coordinate to position the legend outside
+            y=1.05,  # Adjust the y-coordinate to position the legend
+        )
+    )
     
     #classification probability
 
@@ -482,29 +507,61 @@ def ecg_analysis():
 
     #clusters
 
-    fig_llm_cluster = plt.figure()
-    x = df_pca_llm['comp1']
-    y = df_pca_llm['comp2']
+    # fig_llm_cluster = plt.figure()
+    # x = df_pca_llm['comp1']
+    # y = df_pca_llm['comp2']
 
-    labels = ['Cluster 0', 'Cluster 1', 'Cluster 2', 'Cluster 3']
+    # labels = ['Cluster 0', 'Cluster 1', 'Cluster 2', 'Cluster 3']
 
-    # Create a dictionary to map 'RestingECG' values to colors
-    color_mapping = {0: 'r', 1: 'b', 2: 'g', 3: 'y'}
+    # # Create a dictionary to map 'RestingECG' values to colors
+    # color_mapping = {0: 'r', 1: 'b', 2: 'g', 3: 'y'}
 
-    for i in df_pca_llm['cluster'].unique():
-        color = color_mapping.get(i, 'k')  # Use 'k' (black) for undefined values
-        plt.scatter(x[df_pca_llm['cluster'] == i], y[df_pca_llm['cluster'] == i], c=color, label=labels[i])
+    # for i in df_pca_llm['cluster'].unique():
+    #     color = color_mapping.get(i, 'k')  # Use 'k' (black) for undefined values
+    #     plt.scatter(x[df_pca_llm['cluster'] == i], y[df_pca_llm['cluster'] == i], c=color, label=labels[i])
 
-    plt.scatter(pca_llm_point['comp1'], pca_llm_point['comp1'], c='k', marker='D')
+    # plt.scatter(pca_llm_point['comp1'], pca_llm_point['comp1'], c='k', marker='D')
 
-    # Remove the numerical labels from the x and y axes
-    plt.xticks([])
-    plt.yticks([])
+    # # Remove the numerical labels from the x and y axes
+    # plt.xticks([])
+    # plt.yticks([])
+
+    # plt.xlabel('Principal Component 1')
+    # plt.ylabel('Principal Component 2')
+    # plt.legend()
+    # plt.grid(False)
+
+    fig_llm_cluster = go.Figure()  #use plotly for this, otherwise with matplotlib the legend will be ouside of the image when we use gradio
+
+    for cluster in df_pca_llm['cluster'].unique():
+        cluster_data = df_pca_llm[df_pca_llm['cluster'] == cluster]
+        fig_llm_cluster.add_trace(
+            go.Scatter(x=cluster_data['comp1'], y=cluster_data['comp2'], mode='markers', name=f'Cluster {cluster}'))
+
+    # Customize the marker size
+    fig_llm_cluster.update_traces(marker=dict(size=12))
+
+    # Set axis labels
+    fig_llm_cluster.update_xaxes(title_text="Principal Component 1")
+    fig_llm_cluster.update_yaxes(title_text="Principal Component 2")
+
+    # Add the additional point
+    fig_llm_cluster.add_trace(
+        go.Scatter(x=pca_llm_point['comp1'], y=pca_llm_point['comp2'], mode='markers', name='Patient',
+                   marker=dict(size=12, symbol='diamond', line=dict(width=2, color='Black'))))
+
+    # Customize the legend's position (outside the graph)
+    fig_llm_cluster.update_layout(
+        legend=dict(
+            x=1.05,  # Adjust the x-coordinate to position the legend outside
+            y=1  # Adjust the y-coordinate to position the legend
+        )
+    )
+
+    # Deactivate the grid
+    fig_llm_cluster.update_xaxes(showgrid=False)
+    fig_llm_cluster.update_yaxes(showgrid=False)
 
-    plt.xlabel('Principal Component 1')
-    plt.ylabel('Principal Component 2')
-    plt.legend()
-    plt.grid(False)
 
 
 
@@ -524,9 +581,11 @@ def ecg_analysis():
     
 #demo app
 
-demo = gr.Blocks()
+with gr.Blocks(title="TIQUE - AI DEMO CAPABILITIES") as demo:
+
+# demo = gr.Blocks()
 
-with demo:
+# with demo:
     
     gr.Markdown("<h1><center>TIQUE: AI DEMO CAPABILITIES<center><h1>")
 
@@ -543,23 +602,23 @@ with demo:
     
     with gr.Row():
         
-        input_file = gr.UploadButton("Click to Upload an ECG 📁")
+        input_file = gr.UploadButton("Upload patient's data and latest ECG 📁")
         text_upload_results = gr.Textbox()
         
         input_file.upload(upload_ecg,inputs=[input_file],outputs=text_upload_results)
         
     with gr.Row():
-        ecg_start_analysis_button = gr.Button(value="Start ECG analysis",scale=1)
+        ecg_start_analysis_button = gr.Button(value="Start data analysis",scale=1)
 
 
-    gr.Markdown("## Large Language Model clustering")
+    gr.Markdown("## Patient positioning on clusters")
 
     with gr.Row():
 
         llm_cluster = gr.Plot()
         
         
-    gr.Markdown("## Autoencoder results:")
+    gr.Markdown("## ECG analysis:")
     
     with gr.Row():
         
@@ -573,7 +632,7 @@ with demo:
             
             classifier_nn_prediction = gr.Textbox()
 
-    gr.Markdown("## Decision Tree results:")
+    gr.Markdown("## Patient's classification based on Chest Pain Type and ST Slope:")
 
     with gr.Row():