Daily Segment Analyzer

Language: 中文 | English

This project is a Python tool for identifying daily stock price segments, confirmed tops/bottoms, pivots, turn candles, and unconfirmed candidates. It can run from the command line or through a Tkinter GUI.

For the full Chinese documentation and detailed rule definitions, see README.zh-CN.md.

Setup

pip install -r requirements.txt

Dependencies:

• pandas
• requests
• akshare

Daily market data is cached under data/cache by default. If a network endpoint is temporarily unavailable, the program can reuse existing cached data.

Usage

Open the GUI:

python3 main.py gui

The GUI lets users enter a stock symbol, analyze all detected segments, see whether each segment comes from pivot confirmation or an extreme anchor, and export either a TXT summary or detailed JSON.

Analyze one stock:

python3 main.py stock --symbol 001259 --start 19900101 --end 20260529 --output out/001259.json

Analyze all limit-up stocks on a given day:

python3 main.py limit-up --date 20260515 --lookback-days 520 --direction up --output out/zt_20260515.json

Analyze a local CSV file:

python3 main.py csv --file data/sample.csv --direction both --output out/sample.json

CSV files may use date/open/high/low/close, or common AkShare Chinese column names such as 日期/开盘/最高/最低/收盘.

Common options:

• --scan-mode lowest: start from the historical lowest low and complete segments forward/backward.
• --scan-mode history: scan from the left side of history in chronological order.
• --direction auto|up|down|both: initial direction for history mode.
• --adjust ""|qfq|hfq: adjustment mode; default is qfq.
• --max-segments: maximum number of consecutive segments.
• --cache-dir / --no-cache: control market data cache behavior.

Project Structure

• main.py: command-line entry point.
• segment_analyzer/data_fetcher.py: daily market data and limit-up pool fetching.
• segment_analyzer/indicators.py: OHLC normalization and MA5/MA34 calculation.
• segment_analyzer/analyzer.py: core segment, pivot, turn candle, endpoint extension, and post-processing rules.
• segment_analyzer/agent.py: orchestration, batch tasks, JSON output, and summaries.
• segment_analyzer/gui.py: Tkinter GUI.
• segment_analyzer/models.py: output data structures.

Output

Each JSON segment report includes fields such as:

• segment_no, segment_direction
• confirmed_top, confirmed_bottom
• candidate_top, candidate_bottom
• unconfirmed_candidates
• pivot_confirm_k, pivot_line, pivot_lower, pivot_upper, pivot_type
• turn_k, initial_cross, refresh_status
• decision_status

Catalog

• Segment identify
• Segment trend analysis
• Buying and selling points

For complete field explanations and rule details, use the Chinese README: README.zh-CN.md.