mktgtech commited on
Commit
9b204cc
·
verified ·
1 Parent(s): 5fcd213

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +46 -20
app.py CHANGED
@@ -156,7 +156,6 @@ def enrich_leads_with_visits(df, start_date, end_date, max_rows=None, progress=g
156
  print(f"🕵️ Deep enriching {total} rows ({start_date} to {end_date})...")
157
 
158
  for index, row in target_df.iterrows():
159
- # Only enrich if data is missing
160
  if row.get("landing_page_path") and row.get("exit_page_path"):
161
  continue
162
 
@@ -319,11 +318,11 @@ def build_trend(df, preset, grain, filter_values):
319
  return px.line(ts, x="period", y=val, markers=True, template="plotly_white", title=preset)
320
 
321
  # =====================================================
322
- # ADVANCED ANALYTICS (WITH AUTO-ENRICHMENT)
323
  # =====================================================
324
  def build_advanced_plots(df, search_lead_id, search_company, start_date, end_date):
325
  if df is None or df.empty:
326
- return None, None
327
 
328
  # --- 1. FILTER ---
329
  filtered_df = df.copy()
@@ -339,31 +338,23 @@ def build_advanced_plots(df, search_lead_id, search_company, start_date, end_dat
339
  title_suffix = f" (Company: {search_company})"
340
 
341
  if filtered_df.empty:
342
- return go.Figure(layout=dict(title="No matching data found")), go.Figure(layout=dict(title="No matching data found"))
 
 
343
 
344
  # --- 2. AUTO-ENRICHMENT ---
345
- # Check if the filtered rows are missing landing/exit data.
346
- # If so, fetch it immediately!
347
  missing_data = filtered_df[filtered_df["landing_page_path"].isnull()]
348
-
349
  if not missing_data.empty:
350
- # We pass the subset to the enrich function
351
  print(f"🔄 Auto-enriching {len(missing_data)} rows for visualization...")
352
-
353
- # We need to update the original filtered_df with the new data
354
  enriched_subset = enrich_leads_with_visits(missing_data, start_date, end_date)
355
-
356
- # Update the rows in filtered_df using the enriched data
357
  filtered_df.update(enriched_subset)
358
 
359
  # --- 3. SANKEY DIAGRAM ---
360
  flow_df = filtered_df.dropna(subset=["landing_page_path", "exit_page_path"])
361
-
362
  if flow_df.empty:
363
- sankey_fig = go.Figure(layout=dict(title="No visit data available for selected lead(s)"))
364
  else:
365
  flow_counts = flow_df.groupby(["landing_page_path", "exit_page_path"]).size().reset_index(name="count")
366
-
367
  if not search_lead_id and not search_company:
368
  flow_counts = flow_counts.sort_values("count", ascending=False).head(20)
369
 
@@ -402,7 +393,33 @@ def build_advanced_plots(df, search_lead_id, search_company, start_date, end_dat
402
  title=f"Activity Heatmap{title_suffix}"
403
  )
404
 
405
- return sankey_fig, heatmap_fig
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
406
 
407
  # =====================================================
408
  # UI LAYOUT
@@ -462,7 +479,7 @@ with gr.Blocks(title="Leadfeeder Pro Analytics") as demo:
462
  plot
463
  )
464
 
465
- # --- TAB 4: ADVANCED ANALYTICS ---
466
  with gr.Tab("🔬 Advanced Analytics"):
467
  gr.Markdown("Visualize user behavior flows. Leave search blank to see aggregate Top 20 flows.")
468
 
@@ -470,14 +487,23 @@ with gr.Blocks(title="Leadfeeder Pro Analytics") as demo:
470
  search_lead = gr.Textbox(label="Search by Lead ID (Exact)")
471
  search_comp = gr.Textbox(label="Search by Company Name (Partial)")
472
 
473
- btn_adv = gr.Button("Generate User Flow", variant="primary")
474
 
 
475
  with gr.Row():
476
  sankey_plot = gr.Plot(label="User Flow (Sankey)")
477
  heatmap_plot = gr.Plot(label="Activity Heatmap")
478
 
479
- # NOTE: Passed start/end dates so we can auto-enrich on the fly!
480
- btn_adv.click(build_advanced_plots, [df_state, search_lead, search_comp, start, end], [sankey_plot, heatmap_plot])
 
 
 
 
 
 
 
 
481
 
482
  # --- TAB 5: DEBUGGER ---
483
  with gr.Tab("🛠️ Debugger"):
 
156
  print(f"🕵️ Deep enriching {total} rows ({start_date} to {end_date})...")
157
 
158
  for index, row in target_df.iterrows():
 
159
  if row.get("landing_page_path") and row.get("exit_page_path"):
160
  continue
161
 
 
318
  return px.line(ts, x="period", y=val, markers=True, template="plotly_white", title=preset)
319
 
320
  # =====================================================
321
+ # ADVANCED ANALYTICS (WITH MATRIX, SANKEY, HEATMAP)
322
  # =====================================================
323
  def build_advanced_plots(df, search_lead_id, search_company, start_date, end_date):
324
  if df is None or df.empty:
325
+ return None, None, None
326
 
327
  # --- 1. FILTER ---
328
  filtered_df = df.copy()
 
338
  title_suffix = f" (Company: {search_company})"
339
 
340
  if filtered_df.empty:
341
+ return (go.Figure(layout=dict(title="No data")),
342
+ go.Figure(layout=dict(title="No data")),
343
+ go.Figure(layout=dict(title="No data")))
344
 
345
  # --- 2. AUTO-ENRICHMENT ---
 
 
346
  missing_data = filtered_df[filtered_df["landing_page_path"].isnull()]
 
347
  if not missing_data.empty:
 
348
  print(f"🔄 Auto-enriching {len(missing_data)} rows for visualization...")
 
 
349
  enriched_subset = enrich_leads_with_visits(missing_data, start_date, end_date)
 
 
350
  filtered_df.update(enriched_subset)
351
 
352
  # --- 3. SANKEY DIAGRAM ---
353
  flow_df = filtered_df.dropna(subset=["landing_page_path", "exit_page_path"])
 
354
  if flow_df.empty:
355
+ sankey_fig = go.Figure(layout=dict(title="No visit data available"))
356
  else:
357
  flow_counts = flow_df.groupby(["landing_page_path", "exit_page_path"]).size().reset_index(name="count")
 
358
  if not search_lead_id and not search_company:
359
  flow_counts = flow_counts.sort_values("count", ascending=False).head(20)
360
 
 
393
  title=f"Activity Heatmap{title_suffix}"
394
  )
395
 
396
+ # --- 5. INDUSTRY VALUE MATRIX (NEW) ---
397
+ # Only meaningful on aggregate data, so we use the MAIN df, not filtered_df
398
+ # unless filtered by date.
399
+ roi_df = df.groupby("primary_industry").agg({
400
+ "total_visits": "sum",
401
+ "lead_quality_score": "mean",
402
+ "company_name": "count" # Bubble size
403
+ }).reset_index()
404
+
405
+ # Filter small industries to reduce noise
406
+ roi_df = roi_df[roi_df["company_name"] > 2]
407
+
408
+ matrix_fig = px.scatter(
409
+ roi_df,
410
+ x="total_visits",
411
+ y="lead_quality_score",
412
+ size="company_name",
413
+ color="primary_industry",
414
+ hover_name="primary_industry",
415
+ title="Industry Value Matrix (Volume vs Quality)",
416
+ labels={"total_visits": "Volume (Total Visits)", "lead_quality_score": "Avg Quality Score"}
417
+ )
418
+ if not roi_df.empty:
419
+ matrix_fig.add_hline(y=roi_df["lead_quality_score"].mean(), line_dash="dot", annotation_text="Avg Quality")
420
+ matrix_fig.add_vline(x=roi_df["total_visits"].mean(), line_dash="dot", annotation_text="Avg Volume")
421
+
422
+ return sankey_fig, heatmap_fig, matrix_fig
423
 
424
  # =====================================================
425
  # UI LAYOUT
 
479
  plot
480
  )
481
 
482
+ # --- TAB 4: ADVANCED ANALYTICS (NEW MATRIX ADDED) ---
483
  with gr.Tab("🔬 Advanced Analytics"):
484
  gr.Markdown("Visualize user behavior flows. Leave search blank to see aggregate Top 20 flows.")
485
 
 
487
  search_lead = gr.Textbox(label="Search by Lead ID (Exact)")
488
  search_comp = gr.Textbox(label="Search by Company Name (Partial)")
489
 
490
+ btn_adv = gr.Button("Generate Advanced Charts", variant="primary")
491
 
492
+ # Row 1: Flow and Heatmap
493
  with gr.Row():
494
  sankey_plot = gr.Plot(label="User Flow (Sankey)")
495
  heatmap_plot = gr.Plot(label="Activity Heatmap")
496
 
497
+ # Row 2: Value Matrix
498
+ with gr.Row():
499
+ matrix_plot = gr.Plot(label="Industry Value Matrix")
500
+
501
+ # Pass all inputs and outputs
502
+ btn_adv.click(
503
+ build_advanced_plots,
504
+ [df_state, search_lead, search_comp, start, end],
505
+ [sankey_plot, heatmap_plot, matrix_plot]
506
+ )
507
 
508
  # --- TAB 5: DEBUGGER ---
509
  with gr.Tab("🛠️ Debugger"):