What Types of Stock Data Does the Daily‑Stock‑Analysis Repository Analyze?

The ZhuLinsen/daily_stock_analysis engine normalizes four distinct categories of equity information—historical daily OHLCV bars, real‑time price quotes, broad market indices, and fundamental financial metrics—behind a unified Python interface regardless of whether the underlying source is Yahoo Finance, Akshare, or Tushare.

The repository provides a multi‑source aggregation layer that standardizes disparate financial feeds into consistent pandas DataFrames and named‑tuples. By abstracting vendor‑specific quirks through the BaseFetcher class defined in data_provider/base.py, the system guarantees that downstream stock data workflows receive identical column schemas whether they are back‑testing on five years of history or querying live market snapshots.

Historical Daily K‑Line Data (OHLCV)

The foundation of the analysis pipeline is end‑of‑day price history adjusted for splits and dividends. All fetchers inherit from BaseFetcher and must normalize raw payloads into a DataFrame containing the columns ['date', 'open', 'high', 'low', 'close', 'volume', 'amount', 'pct_chg'].

Concrete implementations include:

• YfinanceFetcher – Global equities via Yahoo Finance (data_provider/yfinance_fetcher.py)
• AkshareFetcher – Chinese A‑share and Hong Kong markets (data_provider/akshare_fetcher.py)
• TushareFetcher – Pro‑API tier for institutional‑grade history
• PytdxFetcher and BaostockFetcher – Direct exchange protocols and free fundamental‑enhanced histories

To retrieve data, instantiate DataFetcherManager and call get_daily_data():

from data_provider.base import DataFetcherManager

manager = DataFetcherManager()
df, source = manager.get_daily_data('600519', days=60)
print(f"Fetched {len(df)} rows from {source}")
print(df.head())  # Columns: date, open, high, low, close, volume, amount, pct_chg

The manager implements a cascading priority chain—EfinanceFetcher → AkshareFetcher → PytdxFetcher → TushareFetcher → BaostockFetcher → YfinanceFetcher → LongbridgeFetcher—automatically failing over to the next provider if the primary source is rate‑limited or unavailable.

Real‑Time Price Quotes

Beyond historical bars, the system captures live market snapshots through DataFetcherManager.get_realtime_quote(). This method returns a UnifiedRealtimeQuote named‑tuple containing the latest price, change percentage, volume, turnover rate, PE/PB ratios, and market capitalization, normalized across US, A‑share, and Hong Kong exchanges.

Primary real‑time fetchers include:

• AkshareFetcher – Aggregates East Money, Sina, and Tencent channels, applying a 20‑minute TTL cache for bulk A‑share and ETF quotes (_realtime_cache and _etf_realtime_cache in data_provider/akshare_fetcher.py)
• YfinanceFetcher – US equities and major indices
• LongbridgeFetcher – Authenticated fallback for US/HK securities when free sources exceed limits

manager = DataFetcherManager()

# US equity

quote = manager.get_realtime_quote('AAPL')

# Chinese A‑share (prefix auto‑converted)

quote = manager.get_realtime_quote('sh600519')

# Hong Kong

quote = manager.get_realtime_quote('hk00700')

print(f"{quote.name}: {quote.price} ({quote.change_pct:.2f}%)")

Market‑Wide Statistics and Index Levels

The repository aggregates macro‑level indicators through dedicated methods in DataFetcherManager:

• get_main_indices(region) – Current price, change, and volume for benchmarks such as S&P 500, CSI 300, and Hang Seng
• get_market_stats() – Exchange‑wide up/down/flat security counts
• get_sector_rankings(n) – Top‑performing industry groups by momentum

These methods normalize disparate vendor formats into plain dictionaries with keys such as code, name, current, change_pct, and amplitude, enabling consistent dashboard rendering regardless of whether the underlying feed originates from yfinance_fetcher.py or akshare_fetcher.py.

Fundamental and Company‑Level Data

For valuation and earnings analysis, the AkshareFundamentalAdapter (data_provider/fundamental_adapter.py) extracts periodic report data including balance sheets, income statements, cash‑flow statements, and chip‑structure metrics. The adapter is exposed internally via DataFetcherManager._fundamental_adapter and returns dictionaries containing ratios such as pe_ratio and pb_ratio.

from data_provider.fundamental_adapter import AkshareFundamentalAdapter

adapter = AkshareFundamentalAdapter()
fundamentals = adapter.get_fundamental('600519')
print(f"PE: {fundamentals['pe_ratio']}, PB: {fundamentals['pb_ratio']}")

Summary

• Historical daily data: Standardized OHLCV DataFrames with split/dividend‑adjusted prices, sourced from six distinct fetchers and unified by the BaseFetcher interface in data_provider/base.py.
• Real‑time quotes: Snapshot prices and market metrics returned as UnifiedRealtimeQuote named‑tuples, with 20‑minute caching for A‑share bulk queries.
• Market statistics: Index levels, sector rankings, and broad market breadth indicators accessible through region‑specific DataFetcherManager methods.
• Fundamental data: Quarterly and annual financial statements plus valuation ratios via AkshareFundamentalAdapter.

The architecture guarantees that analysis scripts never handle vendor‑specific column naming or API quirks directly; the manager orchestrates failover, caching, and normalization transparently.

Frequently Asked Questions

Does the repository support real‑time intraday tick data?

No. The real‑time capabilities are limited to snapshot quotes (last price, volume, daily high/low) rather than full Level‑2 order books or tick‑by‑tick time‑and‑sales streams. For historical granularity, the fetchers provide daily‑bar resolution only; sub‑day data is not currently normalized in the BaseFetcher interface.

How does the system handle data provider outages?

DataFetcherManager implements a cascading priority list defined in _init_default_fetchers (data_provider/base.py). If EfinanceFetcher fails, it automatically retries with AkshareFetcher, then PytdxFetcher, continuing down the chain to LongbridgeFetcher. Each fetcher handles its own rate‑limiting and exponential back‑off, while the manager logs the successful source for debugging.

What markets and exchanges are covered?

The codebase explicitly supports US equities (NYSE/NASDAQ via Yahoo Finance and Longbridge), Chinese A‑shares (SSE/SZSE via Akshare, Tushare, Pytdx, and Efinance), and Hong Kong (HKEX via Akshare and Longbridge). The get_main_indices() method distinguishes regions using the region parameter ('us', 'cn', 'hk') to return localized benchmark indices.

Is the historical data adjusted for corporate actions?

Yes. The standard column pct_chg and the adjusted close prices returned by get_daily_data() reflect splits and dividends, not raw nominal prices. This is ensured by the underlying provider libraries (e.g., Akshare and Baostock) and verified through the test suite, which asserts that the standardized DataFrame schema remains consistent across all fetchers.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}