Base app.py interface v1.1

app.py

@@ -1,78 +1,120 @@
-# app.py — Financial RAG Dashboard
+# app.py — Financial RAG → Power BI
 import gradio as gr
 import asyncio
+from itertools import cycle
+from datetime import datetime
 import requests
 import pandas as pd
-from itertools import cycle
 from transformers import pipeline
-from datetime import datetime
 
-# 🔧 Настройки
-FRED_KEY = "YOUR_FRED_API_KEY"
+# ============================================================
+# ⚙️ Конфигурация
+# ============================================================
+FRED_API_KEY = "YOUR_FRED_API_KEY"  # зарегистрируй на https://fred.stlouisfed.org
 FRED_URL = "https://api.stlouisfed.org/fred/series/observations"
+
 INDICATORS = {
     "GDP": "GDP",
     "Inflation (CPI)": "CPIAUCSL",
     "Unemployment Rate": "UNRATE",
-    "Interest Rate (Fed Funds)": "FEDFUNDS"
+    "Interest Rate (Fed Funds)": "FEDFUNDS",
 }
 
-# 🧠 Модель генерации аналитики
 generator = pipeline("text2text-generation", model="google/flan-t5-base")
 
-# 🔁 Анимация “обработка...”
-async def async_loader(update_fn, delay=0.15):
+# ============================================================
+# 🌀 Асинхронная анимация "Loading..."
+# ============================================================
+async def async_loader(update_fn, delay=0.12):
     frames = cycle(["⠋","⠙","⠹","⠸","⠼","⠴","⠦","⠧","⠇","⠏"])
     for frame in frames:
-        update_fn(f"💭 Fetching FRED data {frame}")
+        update_fn(f"💭 Fetching financial data... {frame}")
         await asyncio.sleep(delay)
 
-# 📈 Получение данных FRED
-def fetch_data(indicator):
-    series = INDICATORS[indicator]
+# ============================================================
+# 📈 Основная логика: загрузка, анализ, экспорт
+# ============================================================
+def get_fred_data(series_id, start="2024-01-01"):
     params = {
-        "series_id": series,
-        "api_key": FRED_KEY,
+        "series_id": series_id,
+        "api_key": FRED_API_KEY,
         "file_type": "json",
-        "observation_start": "2024-01-01",
+        "observation_start": start,
     }
-    data = requests.get(FRED_URL, params=params).json().get("observations", [])
+    r = requests.get(FRED_URL, params=params)
+    data = r.json().get("observations", [])
     df = pd.DataFrame(data)
+    if df.empty:
+        return pd.DataFrame()
     df["value"] = pd.to_numeric(df["value"], errors="coerce")
-    return df.tail(8)
+    return df.tail(10)
 
-# 🧩 Аналитика и экспорт
-def generate_and_export(indicator):
-    df = fetch_data(indicator)
+def analyze_and_export(indicator_name, progress_text):
+    """Потоковая функция, полностью в стиле MBTI-интерфейса."""
+    if not indicator_name:
+        yield "⚠️ Select indicator first.", "", progress_text
+        return
+
+    # шаг 1 — загрузка данных
+    yield "📡 Requesting latest data from FRED...", "⏳ Fetching...", "1/3"
+    series_id = INDICATORS[indicator_name]
+    df = get_fred_data(series_id)
     if df.empty:
-        return "⚠️ No data found.", None
+        yield "⚠️ No data received.", "", "1/3"
+        return
+
+    # шаг 2 — подготовка контекста
+    recent = df.tail(5)
+    trend = recent["value"].pct_change().mean() * 100
+    context = f"Recent {indicator_name} values:\n{recent[['date','value']].to_string(index=False)}"
+    yield "📊 Data fetched successfully.", "💭 Generating analytical summary...", "2/3"
 
-    trend = df["value"].pct_change().mean() * 100
-    context = f"Recent {indicator} data:\n{df[['date','value']].to_string(index=False)}"
-    prompt = f"Analyze this economic indicator and summarize its recent trend:\n{context}\nAverage change: {trend:.2f}%"
-    summary = generator(prompt, max_new_tokens=120)[0]["generated_text"]
+    # шаг 3 — генерация аналитики
+    prompt = (
+        f"Analyze the following economic indicator data and describe the recent trend:\n"
+        f"{context}\nAverage change: {trend:.2f}%"
+    )
+    try:
+        response = generator(prompt, max_new_tokens=150)[0]["generated_text"]
+    except Exception as e:
+        response = f"⚠️ LLM generation error: {e}"
 
-    fname = f"powerbi_{indicator.lower().replace(' ','_')}.csv"
-    df.to_csv(fname, index=False)
-    return summary, fname
+    # шаг 4 — экспорт в CSV
+    filename = f"powerbi_{indicator_name.lower().replace(' ','_')}.csv"
+    df.to_csv(filename, index=False)
+    yield response, f"✅ Exported to {filename}", "3/3"
 
-# 🧱 Gradio Interface
+# ============================================================
+# 🧱 Gradio UI (в том же формате, что MBTI)
+# ============================================================
 with gr.Blocks(theme=gr.themes.Soft(), title="🏦 Financial RAG → Power BI") as demo:
     gr.Markdown(
-        "## 🏦 Financial RAG Dashboard\n"
-        "Автоматическая аналитика по банковским данным (FRED API) с экспортом в Power BI.\n\n"
-        "_Интерфейс с потоковой анимацией и аккуратной вёрсткой._"
+        "## 🏦 Financial RAG → Power BI\n"
+        "Получай аналитические сводки по банковским показателям (FRED API) "
+        "и экспортируй их прямо в Power BI CSV.\n\n"
+        "_Интерфейс в том же формате, что MBTI Interviewer._"
     )
 
     with gr.Row():
         with gr.Column(scale=1):
-            indicator = gr.Dropdown(list(INDICATORS.keys()), label="Выберите показатель", value="Inflation (CPI)")
-            btn = gr.Button("📊 Получить аналитику", variant="primary")
-        with gr.Column(scale=2):
-            summary = gr.Textbox(label="📈 Сводка аналитики", lines=8)
-            export_file = gr.File(label="📂 Файл для Power BI")
+            indicator_in = gr.Dropdown(
+                choices=list(INDICATORS.keys()),
+                label="Выберите показатель",
+                value="Inflation (CPI)"
+            )
+            run_btn = gr.Button("📈 Сформировать отчёт", variant="primary")
 
-    # Обработчик
-    btn.click(generate_and_export, inputs=indicator, outputs=[summary, export_file])
+        with gr.Column(scale=1):
+            summary_out = gr.Textbox(label="📊 Аналитическая сводка", lines=7)
+            status_out = gr.Textbox(label="💬 Статус", lines=2)
+            progress_out = gr.Textbox(label="📊 Прогресс", value="0/3")
+
+    # Кнопка → потоковая функция
+    run_btn.click(
+        analyze_and_export,
+        inputs=[indicator_in, progress_out],
+        outputs=[summary_out, status_out, progress_out],
+        show_progress=True,
+    )
 
 demo.queue(max_size=32).launch(server_name="0.0.0.0", server_port=7860)