Update app.py

app.py

@@ -53,9 +53,7 @@ CUSTOM_CSS = """
   --bg1: #18004f;
   --bg2: #2a0a89;
   --panel: rgba(255,255,255,0.88);
-  --panel-2: rgba(255,255,255,0.75);
   --text: #20114f;
-  --muted: #6958a6;
   --gold: #f3c544;
   --orange: #ff8b1a;
   --line: rgba(255,255,255,0.20);
@@ -78,6 +76,7 @@ html, body, .gradio-container {
   padding-bottom: 32px !important;
 }
 
+/* tabs */
 .gr-tab-nav {
   background: rgba(34, 9, 110, 0.70) !important;
   border: 1px solid rgba(255,255,255,0.14) !important;
@@ -89,7 +88,9 @@ html, body, .gradio-container {
 .gr-tab-nav button span,
 .gr-tab-nav button p,
 .gr-tab-nav button div,
-.gr-tab-nav button label {
+.gr-tab-nav button label,
+button[aria-selected="false"],
+button[aria-selected="false"] * {
   color: #ffffff !important;
   opacity: 1 !important;
   font-weight: 800 !important;
@@ -98,46 +99,34 @@ html, body, .gradio-container {
 .gr-tab-nav button {
   border-radius: 14px !important;
   transition: background 0.2s ease, color 0.2s ease !important;
+  box-shadow: none !important;
 }
 
-.gr-tab-nav button:hover {
+.gr-tab-nav button:hover,
+button[aria-selected="false"]:hover {
   background: #000000 !important;
+  background-color: #000000 !important;
+  box-shadow: none !important;
 }
 
-.gr-tab-nav button:hover,
-.gr-tab-nav button:hover span,
-.gr-tab-nav button:hover p,
-.gr-tab-nav button:hover div,
-.gr-tab-nav button:hover label {
+.gr-tab-nav button:hover *,
+button[aria-selected="false"]:hover * {
   color: #ffffff !important;
 }
 
-.gr-tab-nav button.selected {
+.gr-tab-nav button.selected,
+button[aria-selected="true"] {
   background: transparent !important;
   border-bottom: 3px solid var(--orange) !important;
+  box-shadow: none !important;
 }
 
-.gr-tab-nav button.selected,
-.gr-tab-nav button.selected span,
-.gr-tab-nav button.selected p,
-.gr-tab-nav button.selected div,
-.gr-tab-nav button.selected label {
-  color: var(--orange) !important;
-}
-
-button[aria-selected="false"],
-button[aria-selected="false"] * {
-  color: #ffffff !important;
-}
-button[aria-selected="false"]:hover,
-button[aria-selected="false"]:hover * {
-  color: #ffffff !important;
-}
-button[aria-selected="true"],
+.gr-tab-nav button.selected *,
 button[aria-selected="true"] * {
   color: var(--orange) !important;
 }
 
+/* layout */
 .app-shell {
   background: rgba(28, 8, 94, 0.58);
   border: 1px solid var(--line);
@@ -270,7 +259,6 @@ button[aria-selected="true"] * {
 .section-title {
   color: var(--text) !important;
   font-weight: 900 !important;
-  letter-spacing: 0.2px;
 }
 
 .section-title-white,
@@ -288,12 +276,6 @@ label, .gr-form > div > label, .gr-box, .gr-panel {
   border: none !important;
 }
 
-.gr-button-secondary {
-  background: rgba(43, 29, 125, 0.14) !important;
-  color: var(--text) !important;
-  border: 1px solid rgba(43, 29, 125, 0.15) !important;
-}
-
 .gradio-container .block {
   border-radius: 16px !important;
 }
@@ -312,16 +294,6 @@ label, .gr-form > div > label, .gr-box, .gr-panel {
   color: white !important;
 }
 
-.dashboard-white-text h1,
-.dashboard-white-text h2,
-.dashboard-white-text h3,
-.dashboard-white-text h4,
-.dashboard-white-text strong,
-.dashboard-white-text li,
-.dashboard-white-text p {
-  color: white !important;
-}
-
 .dashboard-white-text {
   padding-bottom: 22px !important;
 }
@@ -442,29 +414,13 @@ def safe_df_from_records(records):
     return pd.DataFrame(records)
 
 
-def rename_for_coworking(df: pd.DataFrame) -> pd.DataFrame:
-    if df is None or df.empty:
-        return df
-
-    rename_map = {
-        "hotel_name": "location_name",
-        "room_type": "space_type",
-        "avg_price": "avg_space_price",
-    }
-    out = df.copy()
-    for old, new in rename_map.items():
-        if old in out.columns:
-            out = out.rename(columns={old: new})
-    return out
-
-
 def build_kpi_cards(pricing_df: pd.DataFrame, risk_alerts_df: pd.DataFrame) -> str:
     pricing_df = pricing_df.copy() if pricing_df is not None else pd.DataFrame()
     risk_alerts_df = risk_alerts_df.copy() if risk_alerts_df is not None else pd.DataFrame()
 
-    avg_price = pricing_df["avg_price"].mean() if "avg_price" in pricing_df.columns and not pricing_df.empty else None
-    avg_occ = pricing_df["avg_occupancy"].mean() if "avg_occupancy" in pricing_df.columns and not pricing_df.empty else None
-    avg_cancel = pricing_df["avg_cancellation"].mean() if "avg_cancellation" in pricing_df.columns and not pricing_df.empty else None
+    avg_price = pricing_df["avg_space_price"].mean() if "avg_space_price" in pricing_df.columns and not pricing_df.empty else None
+    avg_util = pricing_df["avg_utilization"].mean() if "avg_utilization" in pricing_df.columns and not pricing_df.empty else None
+    avg_cancel = pricing_df["avg_member_cancellation"].mean() if "avg_member_cancellation" in pricing_df.columns and not pricing_df.empty else None
     total_revenue = pricing_df["total_revenue"].sum() if "total_revenue" in pricing_df.columns and not pricing_df.empty else None
     raise_count = int((pricing_df["pricing_action"] == "Raise price").sum()) if "pricing_action" in pricing_df.columns and not pricing_df.empty else 0
     alert_count = len(risk_alerts_df)
@@ -472,8 +428,8 @@ def build_kpi_cards(pricing_df: pd.DataFrame, risk_alerts_df: pd.DataFrame) -> s
     cards = [
         ("Locations/Segments", len(pricing_df), "#5f44cc"),
         ("Avg Space Price", fmt_num(avg_price), "#2fbf9f"),
-        ("Avg Occupancy", fmt_pct(avg_occ), "#f3c544"),
-        ("Avg Cancellation", fmt_pct(avg_cancel), "#e05b77"),
+        ("Avg Utilization", fmt_pct(avg_util), "#f3c544"),
+        ("Avg Member Cancellation", fmt_pct(avg_cancel), "#e05b77"),
         ("Total Revenue", fmt_num(total_revenue), "#3ba0ff"),
         ("Raise Opportunities", raise_count, "#8a5cff"),
         ("Risk Alerts", alert_count, "#ff7a5c"),
@@ -541,12 +497,7 @@ def chart_action_distribution(pricing_df: pd.DataFrame) -> go.Figure:
         title="Pricing Action Distribution",
         color_discrete_sequence=["#5f44cc", "#2fbf9f", "#f3c544", "#e05b77", "#3ba0ff"],
     )
-    fig.update_layout(
-        template="plotly_white",
-        paper_bgcolor="rgba(255,255,255,0.95)",
-        height=420,
-        showlegend=False,
-    )
+    fig.update_layout(template="plotly_white", paper_bgcolor="rgba(255,255,255,0.95)", height=420, showlegend=False)
     return fig
 
 
@@ -554,10 +505,8 @@ def chart_theme_counts(theme_counts: dict) -> go.Figure:
     if not theme_counts:
         return empty_figure("Top Complaint / Satisfaction Themes")
 
-    df = pd.DataFrame({
-        "theme": list(theme_counts.keys()),
-        "count": list(theme_counts.values())
-    }).sort_values("count", ascending=True).tail(10)
+    df = pd.DataFrame({"theme": list(theme_counts.keys()), "count": list(theme_counts.values())})
+    df = df.sort_values("count", ascending=True).tail(10)
 
     fig = px.bar(
         df,
@@ -568,53 +517,41 @@ def chart_theme_counts(theme_counts: dict) -> go.Figure:
         color="count",
         color_continuous_scale=["#d8cdfa", "#5f44cc"],
     )
-    fig.update_layout(
-        template="plotly_white",
-        paper_bgcolor="rgba(255,255,255,0.95)",
-        height=420,
-    )
+    fig.update_layout(template="plotly_white", paper_bgcolor="rgba(255,255,255,0.95)", height=420)
     return fig
 
 
 def chart_avg_price_by_city(pricing_df: pd.DataFrame) -> go.Figure:
-    if pricing_df is None or pricing_df.empty or "city" not in pricing_df.columns or "avg_price" not in pricing_df.columns:
+    if pricing_df is None or pricing_df.empty or "city" not in pricing_df.columns or "avg_space_price" not in pricing_df.columns:
         return empty_figure("Average Space Price by City")
 
-    df = pricing_df.groupby("city", dropna=False)["avg_price"].mean().reset_index()
+    df = pricing_df.groupby("city", dropna=False)["avg_space_price"].mean().reset_index()
     fig = px.bar(
-        df.sort_values("avg_price", ascending=False),
+        df.sort_values("avg_space_price", ascending=False),
         x="city",
-        y="avg_price",
+        y="avg_space_price",
         title="Average Space Price by City",
-        color="avg_price",
+        color="avg_space_price",
         color_continuous_scale=["#d4d0ff", "#4320b5"],
     )
-    fig.update_layout(
-        template="plotly_white",
-        paper_bgcolor="rgba(255,255,255,0.95)",
-        height=420,
-    )
+    fig.update_layout(template="plotly_white", paper_bgcolor="rgba(255,255,255,0.95)", height=420)
     return fig
 
 
-def chart_avg_occupancy_by_space(pricing_df: pd.DataFrame) -> go.Figure:
-    if pricing_df is None or pricing_df.empty or "room_type" not in pricing_df.columns or "avg_occupancy" not in pricing_df.columns:
+def chart_avg_utilization_by_space(pricing_df: pd.DataFrame) -> go.Figure:
+    if pricing_df is None or pricing_df.empty or "space_type" not in pricing_df.columns or "avg_utilization" not in pricing_df.columns:
         return empty_figure("Average Utilization by Space Type")
 
-    df = pricing_df.groupby("room_type", dropna=False)["avg_occupancy"].mean().reset_index()
+    df = pricing_df.groupby("space_type", dropna=False)["avg_utilization"].mean().reset_index()
     fig = px.bar(
-        df.sort_values("avg_occupancy", ascending=False),
-        x="room_type",
-        y="avg_occupancy",
+        df.sort_values("avg_utilization", ascending=False),
+        x="space_type",
+        y="avg_utilization",
         title="Average Utilization by Space Type",
-        color="avg_occupancy",
+        color="avg_utilization",
         color_continuous_scale=["#d4fff2", "#2fbf9f"],
     )
-    fig.update_layout(
-        template="plotly_white",
-        paper_bgcolor="rgba(255,255,255,0.95)",
-        height=420,
-    )
+    fig.update_layout(template="plotly_white", paper_bgcolor="rgba(255,255,255,0.95)", height=420)
     fig.update_yaxes(tickformat=".0%")
     return fig
 
@@ -632,11 +569,7 @@ def chart_revenue_by_city(pricing_df: pd.DataFrame) -> go.Figure:
         color="total_revenue",
         color_continuous_scale=["#f9ddb0", "#f3c544"],
     )
-    fig.update_layout(
-        template="plotly_white",
-        paper_bgcolor="rgba(255,255,255,0.95)",
-        height=420,
-    )
+    fig.update_layout(template="plotly_white", paper_bgcolor="rgba(255,255,255,0.95)", height=420)
     return fig
 
 
@@ -655,12 +588,7 @@ def chart_alert_levels(risk_alerts_df: pd.DataFrame) -> go.Figure:
         title="Risk Alert Levels",
         color_discrete_map={"High": "#e05b77", "Medium": "#f3c544", "Low": "#2fbf9f"},
     )
-    fig.update_layout(
-        template="plotly_white",
-        paper_bgcolor="rgba(255,255,255,0.95)",
-        height=420,
-        showlegend=False,
-    )
+    fig.update_layout(template="plotly_white", paper_bgcolor="rgba(255,255,255,0.95)", height=420, showlegend=False)
     return fig
 
 
@@ -737,9 +665,7 @@ def run_pipeline(merged_file):
     risk_alerts_df = safe_df_from_records(risk_alerts)
 
     dashboard_kpis = build_kpi_cards(pricing_df, risk_alerts_df)
-    preview_df = merged_df.head(MAX_PREVIEW_ROWS)
-    display_pricing_df = rename_for_coworking(pricing_df.head(20))
-    display_alerts_df = rename_for_coworking(risk_alerts_df.head(20))
+    preview_df = merged_df.head(MAX_PREVIEW_ROWS).copy()
 
     coworking_summary_md = f"""
 ### Coworking Pricing Summary
@@ -749,7 +675,7 @@ def run_pipeline(merged_file):
 ### Automation Notes
 - Workflow 1 cleaned and standardized the uploaded merged dataset.
 - Workflow 2 generated pricing actions and chart-ready outputs.
-- Workflow 3 flagged risky locations and space segments for management review.
+- Workflow 3 flagged risky coworking segments for management review.
 """
 
     risk_summary_md = f"""
@@ -783,11 +709,11 @@ def run_pipeline(merged_file):
         chart_action_distribution(pricing_df),
         chart_theme_counts(theme_counts),
         chart_avg_price_by_city(pricing_df),
-        chart_avg_occupancy_by_space(pricing_df),
+        chart_avg_utilization_by_space(pricing_df),
         chart_revenue_by_city(pricing_df),
         chart_alert_levels(risk_alerts_df),
-        display_pricing_df,
-        display_alerts_df,
+        pricing_df.head(20),
+        risk_alerts_df.head(20),
         analysis_state,
     )
 
@@ -823,8 +749,8 @@ def fallback_ai_answer(question: str, analysis_state: dict) -> str:
             if not candidates.empty:
                 top = candidates.iloc[0]
                 return (
-                    f"The strongest current raise-price opportunity is {top.get('hotel_name', 'Unknown location')} "
-                    f"in {top.get('city', 'Unknown city')} for {top.get('room_type', 'Unknown space type')}. "
+                    f"The strongest current raise-price opportunity is {top.get('coworking_space_name', 'Unknown location')} "
+                    f"in {top.get('city', 'Unknown city')} for {top.get('space_type', 'Unknown space type')}. "
                     f"Rationale: {top.get('rationale', 'No rationale returned.')}"
                 )
         return "No raise-price opportunity was returned by the automation."
@@ -834,8 +760,8 @@ def fallback_ai_answer(question: str, analysis_state: dict) -> str:
             first = risk_alerts[0]
             return (
                 f"{alerts_summary}\n\n"
-                f"One flagged segment is {first.get('hotel_name', 'Unknown location')} in "
-                f"{first.get('city', 'Unknown city')} for {first.get('room_type', 'Unknown space type')}. "
+                f"One flagged segment is {first.get('coworking_space_name', 'Unknown location')} in "
+                f"{first.get('city', 'Unknown city')} for {first.get('space_type', 'Unknown space type')}. "
                 f"Reason: {first.get('reasons', 'No reason returned.')}"
             )
         return "No active risk alerts were returned by the automation."
@@ -844,13 +770,13 @@ def fallback_ai_answer(question: str, analysis_state: dict) -> str:
         return management_summary or "No management summary is currently available."
 
     if "occupancy" in q or "utilization" in q:
-        if "avg_occupancy" in pricing_df.columns and not pricing_df.empty:
-            avg_occ = pricing_df["avg_occupancy"].mean()
-            return f"The average utilization across coworking segments is {fmt_pct(avg_occ)}."
+        if "avg_utilization" in pricing_df.columns and not pricing_df.empty:
+            avg_util = pricing_df["avg_utilization"].mean()
+            return f"The average utilization across coworking segments is {fmt_pct(avg_util)}."
         return "Utilization data is not currently available."
 
     return (
-        "I can answer questions about pricing actions, coworking themes, risk alerts, utilization, and overall management summary. "
+        "I can answer questions about pricing actions, coworking themes, risk alerts, utilization, and the overall management summary. "
         "Try asking: 'Where should prices be raised?' or 'What are the main complaint themes?'"
     )
 
@@ -860,7 +786,7 @@ def build_llm_prompt(question: str, analysis_state: dict) -> str:
 You are an AI assistant for a coworking space pricing and satisfaction dashboard.
 
 You must answer as a concise business analyst.
-Use the automation outputs below.
+Use coworking language only. Never refer to hotels, guests, or rooms.
 
 Management summary:
 {analysis_state.get("management_summary", "")}
@@ -882,7 +808,7 @@ User question:
 
 Instructions:
 - Answer directly.
-- Use coworking language, not hotel language.
+- Use coworking language only.
 - Mention pricing implications when relevant.
 - Keep it clear and business-focused.
 """
@@ -925,10 +851,10 @@ def ask_ai(question, history, analysis_state):
     else:
         answer = fallback_ai_answer(question, analysis_state)
 
-    history.append({"role": "user", "content": question})
-    history.append({"role": "assistant", "content": answer})
+    history.append((question, answer))
     return history, ""
 
+
 # =========================================================
 # UI
 # =========================================================
@@ -1012,7 +938,7 @@ with gr.Blocks(title="AI Coworking Space Pricing Optimizer") as demo:
 
         with gr.Row():
             price_city_chart = gr.Plot(label="Average Space Price by City")
-            occupancy_space_chart = gr.Plot(label="Average Utilization by Space Type")
+            utilization_space_chart = gr.Plot(label="Average Utilization by Space Type")
 
         with gr.Row():
             revenue_city_chart = gr.Plot(label="Revenue by City")
@@ -1026,9 +952,7 @@ with gr.Blocks(title="AI Coworking Space Pricing Optimizer") as demo:
 
     with gr.Tab('"AI" Dashboard'):
         with gr.Group(elem_classes=["panel-card", "ai-panel"]):
-            llm_status = gr.HTML(
-                f'<div class="status-card"><p>{get_llm_status_text()}</p></div>'
-            )
+            gr.HTML(f'<div class="status-card"><p>{get_llm_status_text()}</p></div>')
 
             gr.Markdown(
                 """
@@ -1068,7 +992,7 @@ Example questions:
             action_chart,
             theme_chart,
             price_city_chart,
-            occupancy_space_chart,
+            utilization_space_chart,
             revenue_city_chart,
             alerts_level_chart,
             pricing_table,