Update app.py

app.py

@@ -3,6 +3,7 @@ import pandas as pd
 import matplotlib.pyplot as plt
 from fpdf import FPDF
 import io
+from datetime import datetime, timedelta
 
 # Global DataFrame to store the CSV data
 df = pd.DataFrame()
@@ -10,106 +11,128 @@ df = pd.DataFrame()
 def upload_csv(file):
     global df
     try:
-        # Check if a file was actually uploaded
         if file is None:
-            return ["All"], ["All"], "No file uploaded. Please upload a CSV file.", "All", "All"
+            return ["All"], ["All"], ["All"], "No file uploaded. Please upload a CSV file.", "All", "All", "All"
         
-        # Read the CSV file directly from the file object
+        # Read the CSV file
         df = pd.read_csv(file)
         
-        # Check if the DataFrame is empty
         if df.empty:
-            return ["All"], ["All"], "The uploaded CSV file is empty.", "All", "All"
+            return ["All"], ["All"], ["All"], "The uploaded CSV file is empty.", "All", "All", "All"
         
         # Define required columns
-        required_columns = {'DeviceID', 'Lab', 'Type', 'Timestamp', 'Status', 'UsageCount'}
-        
-        # Check if all required columns are present
+        required_columns = {'DeviceID', 'Lab', 'Type', 'Timestamp', 'Status', 'UsageCount', 'Health'}
         if not required_columns.issubset(df.columns):
             missing_cols = required_columns - set(df.columns)
-            return ["All"], ["All"], f"CSV is missing required columns: {', '.join(missing_cols)}", "All", "All"
+            return ["All"], ["All"], ["All"], f"CSV is missing required columns: {', '.join(missing_cols)}", "All", "All", "All"
+        
+        # Convert Timestamp to datetime
+        df['Timestamp'] = pd.to_datetime(df['Timestamp'])
         
-        # Extract unique values for dropdowns, ensuring no empty or invalid entries
+        # Extract unique values for dropdowns
         labs = ['All'] + sorted([str(lab) for lab in df['Lab'].dropna().unique() if str(lab).strip()])
         types = ['All'] + sorted([str(type_) for type_ in df['Type'].dropna().unique() if str(type_).strip()])
         
-        # If no valid labs or types are found, return default values
-        if len(labs) == 1:  # Only "All" is present
-            labs = ["All"]
-        if len(types) == 1:  # Only "All" is present
-            types = ["All"]
+        # Extract date range for filter
+        min_date = df['Timestamp'].min().strftime('%Y-%m-%d')
+        max_date = df['Timestamp'].max().strftime('%Y-%m-%d')
+        date_ranges = ['All', f"{min_date} to {max_date}"]
         
-        return labs, types, "", "All", "All"
+        return labs, types, date_ranges, "", "All", "All", "All"
     except Exception as e:
-        # Handle any other errors (e.g., file not readable, invalid CSV format)
-        return ["All"], ["All"], f"Failed to load CSV: {str(e)}", "All", "All"
+        return ["All"], ["All"], ["All"], f"Failed to load CSV: {str(e)}", "All", "All", "All"
 
-def filter_and_plot(selected_lab, selected_type):
+def filter_and_visualize(selected_lab, selected_type, selected_date_range):
     global df
-    # If no data is loaded, return None (no plot)
     if df.empty:
-        return None
+        return None, None, None, None, "No data available."
     
-    # Create a copy of the DataFrame for filtering
+    # Filter the DataFrame
     filtered_df = df.copy()
-    
-    # Apply filters based on dropdown selections
     if selected_lab != "All":
         filtered_df = filtered_df[filtered_df["Lab"] == selected_lab]
     if selected_type != "All":
         filtered_df = filtered_df[filtered_df["Type"] == selected_type]
+    if selected_date_range != "All" and selected_date_range != "No data available.":
+        start_date, end_date = selected_date_range.split(" to ")
+        start_date = pd.to_datetime(start_date)
+        end_date = pd.to_datetime(end_date) + timedelta(days=1)  # Include end date
+        filtered_df = filtered_df[(filtered_df["Timestamp"] >= start_date) & (filtered_df["Timestamp"] < end_date)]
     
-    # If no data remains after filtering, return None
     if filtered_df.empty:
-        return None
+        return None, None, None, None, "No data matches the selected filters."
+    
+    # Device Cards (as a table)
+    device_cards = filtered_df[['DeviceID', 'Lab', 'Type', 'Health', 'UsageCount', 'Timestamp']].sort_values(by='Timestamp', ascending=False)
+    
+    # Daily Log Trends (Line Chart)
+    daily_logs = filtered_df.groupby(filtered_df['Timestamp'].dt.date).size()
+    plt.figure(figsize=(8, 4))
+    daily_logs.plot(kind='line', marker='o', color='blue')
+    plt.title("Daily Log Trends")
+    plt.xlabel("Date")
+    plt.ylabel("Number of Logs")
+    plt.xticks(rotation=45)
+    buf1 = io.BytesIO()
+    plt.savefig(buf1, format="png", bbox_inches="tight")
+    plt.close()
+    buf1.seek(0)
     
-    # Prepare the plot
+    # Weekly Uptime % (Bar Chart)
+    end_date = filtered_df['Timestamp'].max()
+    start_date = end_date - timedelta(days=7)
+    weekly_df = filtered_df[(filtered_df['Timestamp'] >= start_date) & (filtered_df['Timestamp'] <= end_date)]
+    uptime = weekly_df.groupby(weekly_df['Timestamp'].dt.date)['Status'].apply(lambda x: (x == 'Up').mean() * 100)
     plt.figure(figsize=(8, 4))
-    status_counts = filtered_df["Status"].value_counts()
-    status_counts.plot(kind="bar", color=["green", "red"])
-    plt.title("Status Counts")
-    plt.xlabel("Status")
-    plt.ylabel("Count")
+    uptime.plot(kind='bar', color='green')
+    plt.title("Weekly Uptime %")
+    plt.xlabel("Date")
+    plt.ylabel("Uptime %")
+    plt.xticks(rotation=45)
+    buf2 = io.BytesIO()
+    plt.savefig(buf2, format="png", bbox_inches="tight")
+    plt.close()
+    buf2.seek(0)
     
-    # Save the plot to a buffer
-    buf = io.BytesIO()
-    plt.savefig(buf, format="png")
-    plt.close()  # Close the plot to free memory
-    buf.seek(0)
-    return buf
+    # Anomaly Alerts (Text)
+    anomalies = filtered_df[(filtered_df['UsageCount'] > 80) | (filtered_df['Status'] == 'Down')]
+    if anomalies.empty:
+        anomaly_text = "No anomalies detected."
+    else:
+        anomaly_text = "Anomalies Detected:\n" + anomalies[['DeviceID', 'Lab', 'Type', 'Status', 'UsageCount']].to_string(index=False)
+    
+    return device_cards, buf1, buf2, anomaly_text, ""
 
-def download_pdf(selected_lab, selected_type):
+def download_pdf(selected_lab, selected_type, selected_date_range):
     global df
-    # If no data is loaded, return None (no PDF)
     if df.empty:
         return None
     
-    # Create a copy of the DataFrame for filtering
     filtered_df = df.copy()
-    
-    # Apply filters based on dropdown selections
     if selected_lab != "All":
         filtered_df = filtered_df[filtered_df["Lab"] == selected_lab]
     if selected_type != "All":
         filtered_df = filtered_df[filtered_df["Type"] == selected_type]
+    if selected_date_range != "All" and selected_date_range != "No data available.":
+        start_date, end_date = selected_date_range.split(" to ")
+        start_date = pd.to_datetime(start_date)
+        end_date = pd.to_datetime(end_date) + timedelta(days=1)
+        filtered_df = filtered_df[(filtered_df["Timestamp"] >= start_date) & (filtered_df["Timestamp"] < end_date)]
     
-    # If no data remains after filtering, return None
     if filtered_df.empty:
         return None
     
-    # Generate the PDF
     pdf = FPDF()
     pdf.add_page()
     pdf.set_font("Arial", size=12)
-    pdf.cell(200, 10, txt="LabOps Summary Report", ln=True, align='C')
+    pdf.cell(200, 10, txt="LabOps Dashboard Report", ln=True, align='C')
     pdf.ln(10)
     
-    # Add data to the PDF
+    # Add filtered data
     for index, row in filtered_df.iterrows():
-        line = f"{row['Timestamp']} | {row['DeviceID']} | {row['Lab']} | {row['Type']} | {row['Status']} | {row['UsageCount']}"
+        line = f"{row['Timestamp']} | {row['DeviceID']} | {row['Lab']} | {row['Type']} | {row['Status']} | {row['UsageCount']} | {row['Health']}"
         pdf.multi_cell(0, 10, txt=line)
     
-    # Save the PDF to a buffer
     output = io.BytesIO()
     pdf.output(output)
     output.seek(0)
@@ -117,27 +140,50 @@ def download_pdf(selected_lab, selected_type):
 
 # Build the Gradio interface
 with gr.Blocks() as demo:
-    gr.Markdown("🧪 **LabOps Dashboard with Filters**\nUpload lab device logs, filter by Lab and Equipment Type, visualize uptime/downtime & generate PDF reports.")
+    gr.Markdown("🧪 **Multi-Device LabOps Dashboard**\nMonitor smart lab devices, visualize logs, and generate PDF reports.")
 
     with gr.Row():
-        csv_input = gr.File(label="Upload Log CSV", file_types=[".csv"])
+        csv_input = gr.File(label="Upload Device Logs CSV", file_types=[".csv"])
     
     with gr.Row():
-        lab_dropdown = gr.Dropdown(label="Select Lab", choices=["All"], value="All")
-        type_dropdown = gr.Dropdown(label="Select Equipment Type", choices=["All"], value="All")
+        lab_dropdown = gr.Dropdown(label="Filter by Lab", choices=["All"], value="All")
+        type_dropdown = gr.Dropdown(label="Filter by Equipment Type", choices=["All"], value="All")
+        date_dropdown = gr.Dropdown(label="Filter by Date Range", choices=["All"], value="All")
+
+    with gr.Row():
+        device_table = gr.DataFrame(label="Device Cards (Health, Usage, Last Log)")
+        plot_daily = gr.Image(label="Daily Log Trends")
+        plot_uptime = gr.Image(label="Weekly Uptime %")
 
-    plot_output = gr.Image(label="Plot")
-    download_btn = gr.Button("Download PDF Summary")
+    anomaly_output = gr.Textbox(label="Anomaly Alerts")
+    download_btn = gr.Button("Download PDF Report")
     error_box = gr.Textbox(label="Status/Error Message", visible=True, interactive=False)
 
     # Connect the components
     csv_input.change(
         fn=upload_csv,
         inputs=csv_input,
-        outputs=[lab_dropdown, type_dropdown, error_box, lab_dropdown, type_dropdown]
+        outputs=[lab_dropdown, type_dropdown, date_dropdown, error_box, lab_dropdown, type_dropdown, date_dropdown]
+    )
+    lab_dropdown.change(
+        fn=filter_and_visualize,
+        inputs=[lab_dropdown, type_dropdown, date_dropdown],
+        outputs=[device_table, plot_daily, plot_uptime, anomaly_output, error_box]
+    )
+    type_dropdown.change(
+        fn=filter_and_visualize,
+        inputs=[lab_dropdown, type_dropdown, date_dropdown],
+        outputs=[device_table, plot_daily, plot_uptime, anomaly_output, error_box]
+    )
+    date_dropdown.change(
+        fn=filter_and_visualize,
+        inputs=[lab_dropdown, type_dropdown, date_dropdown],
+        outputs=[device_table, plot_daily, plot_uptime, anomaly_output, error_box]
+    )
+    download_btn.click(
+        fn=download_pdf,
+        inputs=[lab_dropdown, type_dropdown, date_dropdown],
+        outputs=gr.File(label="Download PDF")
     )
-    lab_dropdown.change(fn=filter_and_plot, inputs=[lab_dropdown, type_dropdown], outputs=plot_output)
-    type_dropdown.change(fn=filter_and_plot, inputs=[lab_dropdown, type_dropdown], outputs=plot_output)
-    download_btn.click(fn=download_pdf, inputs=[lab_dropdown, type_dropdown], outputs=gr.File(label="Download PDF"))
 
 demo.launch()