adding motherduch

app3.py

@@ -0,0 +1,166 @@
+import streamlit as st
+import pandas as pd
+import plotly.express as px
+import duckdb
+import os
+
+# --- Set page layout ---
+st.set_page_config(layout="wide", page_title="Sentiment Analysis")
+
+# --- CSS for Flag Toggle ---
+st.markdown("""
+    <style>
+    .flag-container { display: flex; justify-content: center; gap: 10px; }
+    .flag { cursor: pointer; width: 40px; height: 30px; }
+    </style>
+""", unsafe_allow_html=True)
+
+# --- State & Language ---
+flag_selection = st.radio("", ["🇬🇧", "🇧🇬"], horizontal=True, label_visibility="collapsed")
+LANG = "English" if flag_selection == "🇬🇧" else "Български"
+
+T = {
+    "English": {
+        "title": "📊 Sentiment Analysis Dashboard",
+        "select_entities": "Select Entities",
+        "select_domains": "Select Domains",
+        "score_type": "Select Score Type",
+        "group_by_domain": "Group by Domain",
+        "all_scores": "Show All Score Types",
+        "no_data": "No data matches the selected filters.",
+        "avg_over_time": "Average {} Over Time",
+        "scores_over_time": "Sentiment Scores Over Time",
+        "select_entities_prompt": "Please select at least one entity."
+    },
+    "Български": {
+        "title": "📊 Табло за анализ на настроенията",
+        "select_entities": "Изберете обекти",
+        "select_domains": "Изберете източници",
+        "score_type": "Изберете тип оценка",
+        "group_by_domain": "Групирай по сайт",
+        "all_scores": "Покажи всички типове оценки",
+        "no_data": "Няма данни за избраните филтри.",
+        "avg_over_time": "Средна стойност на {} във времето",
+        "scores_over_time": "Оценки на настроенията във времето",
+        "select_entities_prompt": "Моля, изберете поне един обект."
+    }
+}
+
+# --- MotherDuck Connection ---
+# Set MOTHERDUCK_TOKEN in Hugging Face Secrets
+@st.cache_resource
+def get_connection():
+    token = st.secrets["MOTHERDUCK_TOKEN"]
+    return duckdb.connect(f"md:?motherduck_token={token}")
+
+con = get_connection()
+
+# --- Cached Metadata for Filters ---
+@st.cache_data(ttl=600)
+def get_filter_options():
+    # Use DuckDB to get counts extremely fast
+    entities = con.sql("SELECT entity, COUNT(*) as c FROM sentiment_analysis GROUP BY 1 ORDER BY c DESC").df()
+    domains = con.sql("SELECT domain, COUNT(*) as c FROM sentiment_analysis GROUP BY 1 ORDER BY c DESC").df()
+    return entities, domains
+
+df_entities, df_domains = get_filter_options()
+
+# --- Sidebar ---
+st.title(T[LANG]["title"])
+
+with st.sidebar:
+    selected_entities = st.multiselect(
+        T[LANG]["select_entities"], 
+        options=df_entities['entity'].tolist(),
+        format_func=lambda x: f"{x} ({df_entities[df_entities['entity']==x]['c'].values[0]})"
+    )
+    
+    selected_domains = st.multiselect(
+        T[LANG]["select_domains"], 
+        options=df_domains['domain'].tolist(),
+        format_func=lambda x: f"{x} ({df_domains[df_domains['domain']==x]['c'].values[0]})"
+    )
+
+    score_type = st.selectbox(T[LANG]["score_type"], ["entity_score", "title_score", "overall_score"])
+    group_by_domain = st.checkbox(T[LANG]["group_by_domain"])
+    all_scores = st.checkbox(T[LANG]["all_scores"])
+    
+    timeframes = {"All": 9999, "Last 7 Days": 7, "Last 30 Days": 30, "Last Year": 365}
+    time_choice = st.selectbox("Timeframe", list(timeframes.keys()))
+
+# --- Query Building ---
+if not selected_entities:
+    st.info(T[LANG]["select_entities_prompt"])
+    st.stop()
+
+# Build dynamic SQL
+where_clause = f"WHERE entity IN ({str(selected_entities)[1:-1]})"
+if selected_domains:
+    where_clause += f" AND domain IN ({str(selected_domains)[1:-1]})"
+if time_choice != "All":
+    where_clause += f" AND created_at >= (epoch(now()) - {timeframes[time_choice] * 86400})"
+
+# Column selection logic
+if all_scores:
+    # Use DuckDB UNPIVOT (equivalent to pandas melt but faster)
+    sql_query = f"""
+    SELECT 
+        time_bucket(interval '1 day', to_timestamp(created_at)) as date,
+        entity,
+        domain,
+        score_name as score_type,
+        AVG(score_value) as score
+    FROM (
+        UNPIVOT sentiment_analysis 
+        ON entity_score, title_score, overall_score
+        INTO NAME score_name VALUE score_value
+    )
+    {where_clause}
+    GROUP BY ALL
+    """
+else:
+    sql_query = f"""
+    SELECT 
+        time_bucket(interval '1 day', to_timestamp(created_at)) as date,
+        entity,
+        domain,
+        AVG({score_type}) as score
+    FROM sentiment_analysis
+    {where_clause}
+    GROUP BY ALL
+    """
+
+# --- Execution & Plotting ---
+filtered_df = con.sql(sql_query).df()
+
+if filtered_df.empty:
+    st.warning(T[LANG]["no_data"])
+else:
+    # Create the Legend Label
+    color_col = "entity"
+    if group_by_domain:
+        filtered_df["label"] = filtered_df["entity"] + " | " + filtered_df["domain"]
+        color_col = "label"
+    if all_scores:
+        filtered_df["label"] = filtered_df.get("label", filtered_df["entity"]) + " | " + filtered_df["score_type"]
+        color_col = "label"
+
+    fig = px.line(
+        filtered_df.sort_values("date"),
+        x="date", y="score", color=color_col,
+        title=T[LANG]["scores_over_time"] if all_scores else T[LANG]["avg_over_time"].format(score_type),
+        labels={"score": "Sentiment Score", "date": "Date"}
+    )
+
+    # Grid line styling
+    fig.update_layout(
+        yaxis=dict(range=[-10, 10], gridcolor="lightgrey"),
+        plot_bgcolor="white",
+        legend=dict(orientation="h", y=-0.2)
+    )
+    
+    # Add emphasis lines at -5, 0, 5
+    for val in [-5, 0, 5]:
+        fig.add_hline(y=val, line_width=2 if val==0 else 1, line_dash="dash", line_color="black")
+
+    st.plotly_chart(fig, use_container_width=True)