New tabs and functions v2.1

app.py

@@ -97,28 +97,29 @@ with gr.Blocks(theme=dark_theme, css="""
             chart_out = gr.Plot(label="Alpha vs BTC")
             chart_btn.click(fn=build_alpha_chart, inputs=chart_in, outputs=chart_out)
 
-        # --- Crypto Intelligence Dashboard (final Power BI style) ---
+        # --- Crypto Intelligence Dashboard (Coinlore, Altair) ---
         with gr.TabItem("Crypto Intelligence Dashboard"):
-            gr.Markdown("### 💹 Multi-Source Market Dashboard")
+            gr.Markdown("### 💹 Coinlore Market Dashboard (Altair Edition)")
 
             with gr.Row():
-                data_source = gr.Radio(
-                    ["Binance", "Coinlore", "Yahoo"], label="Select Data Source", value="Binance")
-                asset_select = gr.CheckboxGroup(
-                    label="Assets", choices=["BTCUSDT", "ETHUSDT", "BNBUSDT", "SOLUSDT", "DOGEUSDT"],
-                    value=["BTCUSDT", "ETHUSDT"])
-                load_btn = gr.Button("Build Dashboard", variant="primary")
+                top_slider = gr.Slider(label="Top N coins", minimum=20, maximum=100, step=10, value=50)
+                load_btn = gr.Button("Generate Dashboard", variant="primary")
 
-            chart = gr.Plot(label="Market Visualization")
-            ai_box = gr.Textbox(label="AI Summary", lines=8, elem_id="llm_comment_box")
+            treemap_plot = gr.Plot(label="Market Composition")
+            movers_plot = gr.Plot(label="Top Movers")
+            scatter_plot = gr.Plot(label="Market Cap vs Volume")
+            ai_box = gr.Textbox(label="AI Market Summary", lines=8, elem_id="llm_comment_box")
 
             from core.crypto_dashboard import build_crypto_dashboard
 
-            load_btn.click(fn=build_crypto_dashboard,
-                           inputs=[data_source, asset_select],
-                           outputs=[chart, ai_box],
-                           show_progress="minimal")
 
+            def run_dash(n): return build_crypto_dashboard(n)
+
+
+            load_btn.click(fn=run_dash,
+                           inputs=top_slider,
+                           outputs=[treemap_plot, movers_plot, scatter_plot, ai_box],
+                           show_progress="minimal")
     gr.Markdown("---")
     gr.Markdown(
         "<center><small style='color:#6e7681;'>Developed with Featherless.ai • Powered by OpenAI-compatible API</small></center>",

core/crypto_dashboard.py

@@ -1,47 +1,74 @@
 """
-Power-BI-style dashboard combining Binance, Coinlore, and Yahoo Finance
+Coinlore Dashboard — Power BI style (Altair)
 """
-import plotly.express as px, plotly.graph_objects as go
+import requests, pandas as pd, altair as alt
 from services.llm_client import llm_service
-from core.data_binance import get_binance_data
-from core.data_coinlore import get_coinlore_data
-from core.data_yfinance import get_yf_history
-
-def build_crypto_dashboard(source: str, symbols: list[str]):
-    if source == "Binance":
-        df = get_binance_data(symbols)
-        fig = px.line(df, x="timestamp", y="close", color="symbol",
-                      title="Binance Price History (1y)", template="plotly_dark")
-    elif source == "Yahoo":
-        df = get_yf_history([s.replace("USDT","-USD") for s in symbols])
-        df["returns"] = df.groupby("symbol")["close"].apply(lambda x: x/x.iloc[0]-1)
-        fig = px.line(df, x="timestamp", y="returns", color="symbol",
-                      title="Yahoo Finance – Normalized Returns (2y)", template="plotly_dark")
-    elif source == "Coinlore":
-        df = get_coinlore_data().head(25)
-        fig = px.treemap(df,
-                         path=["symbol"], values="market_cap_usd",
-                         color="percent_change_24h",
-                         color_continuous_scale="RdYlGn",
-                         title="Coinlore Market Treemap (Top 25)",
-                         template="plotly_dark")
-    else:
-        fig = go.Figure()
-        fig.add_annotation(text="Unknown data source", x=0.5, y=0.5, showarrow=False)
-
-    insight = _generate_ai_summary(source, symbols)
-    return fig, insight
-
-
-def _generate_ai_summary(source, symbols):
+
+
+def fetch_coinlore_data(limit=100):
+    url = "https://api.coinlore.net/api/tickers/"
+    data = requests.get(url).json()["data"]
+    df = pd.DataFrame(data)
+    for col in ["price_usd", "market_cap_usd", "volume24",
+                "percent_change_1h", "percent_change_24h", "percent_change_7d"]:
+        df[col] = pd.to_numeric(df[col], errors="coerce")
+    return df.head(limit)
+
+
+def build_crypto_dashboard(top_n=50):
+    df = fetch_coinlore_data(top_n)
+
+    # === 1️⃣ Treemap imitation ===
+    base = alt.Chart(df).mark_rect().encode(
+        x=alt.X('symbol:N', title=None, axis=None),
+        y=alt.Y('market_cap_usd:Q', title=None, axis=None),
+        color=alt.Color('percent_change_24h:Q',
+                        scale=alt.Scale(scheme='redyellowgreen'),
+                        title='24h Change %'),
+        tooltip=['symbol', 'price_usd', 'market_cap_usd', 'percent_change_24h']
+    ).properties(title="Market Composition (Top Coins)", height=300)
+
+    # === 2️⃣ Bar chart (Top Movers 24h) ===
+    movers = df.sort_values("percent_change_24h", ascending=False).head(15)
+    bar = alt.Chart(movers).mark_bar(cornerRadiusEnd=4).encode(
+        x=alt.X('percent_change_24h:Q', title='Change %'),
+        y=alt.Y('symbol:N', sort='-x', title=None),
+        color=alt.Color('percent_change_24h:Q', scale=alt.Scale(scheme='greenblue')),
+        tooltip=['symbol', 'price_usd', 'percent_change_24h']
+    ).properties(title="Top 15 Gainers (24h)", height=300)
+
+    # === 3️⃣ Scatter (Market Cap vs Volume) ===
+    scatter = alt.Chart(df).mark_circle(size=70).encode(
+        x=alt.X('market_cap_usd:Q', scale=alt.Scale(type='log')),
+        y=alt.Y('volume24:Q', scale=alt.Scale(type='log')),
+        color=alt.Color('percent_change_7d:Q',
+                        scale=alt.Scale(scheme='redyellowgreen'),
+                        title='7d %'),
+        tooltip=['symbol', 'price_usd', 'market_cap_usd', 'volume24', 'percent_change_7d']
+    ).properties(title="Market Cap vs 24h Volume", height=350)
+
+    # === LLM summary ===
+    ai_comment = _generate_ai_summary(df)
+
+    return base, bar, scatter, ai_comment
+
+
+def _generate_ai_summary(df):
+    leaders = df.sort_values("percent_change_24h", ascending=False).head(3)
+    laggards = df.sort_values("percent_change_24h").head(3)
     prompt = f"""
-Summarize crypto market from {source} for assets: {', '.join(symbols)}.
-Describe trend, volatility, and investor sentiment.
+Provide a brief professional summary of today's crypto market based on Coinlore data.
+Top gainers: {', '.join(leaders['symbol'].tolist())}.
+Top losers: {', '.join(laggards['symbol'].tolist())}.
+Include:
+- general sentiment
+- volatility trend
+- short-term outlook (bullish/bearish)
 """
-    summary = ""
+    text = ""
     for delta in llm_service.stream_chat(
         messages=[{"role": "user", "content": prompt}],
         model="meta-llama/Meta-Llama-3.1-8B-Instruct"
     ):
-        summary += delta
-    return summary
+        text += delta
+    return text

requirements.txt

@@ -7,4 +7,4 @@ matplotlib
 plotly
 yfinance>=0.2.43
 plotly>=6.3.1
-
+altair