tqsdk

天勤量化TqSdk - 开源Python期货/期权交易SDK,提供实时行情、回测和实盘交易功能。

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "tqsdk" with this command: npx skills add coderwpf/tqsdk

TqSdk(天勤量化SDK)

TqSdk 是信易科技开发的开源Python期货/期权量化交易SDK。通过统一的异步API提供实时行情、历史数据、回测和实盘交易功能。

文档:https://doc.shinnytech.com/tqsdk/latest/ 免费版提供延时行情(15分钟延迟),专业版支持实时数据。

安装

pip install tqsdk

快速入门

from tqsdk import TqApi, TqAuth, TqBacktest
from datetime import date

# 实盘模式(免费账户提供延时行情)
api = TqApi(auth=TqAuth("your_username", "your_password"))

# 获取实时行情
quote = api.get_quote("SHFE.cu2401")  # Shanghai copper futures
print(f"最新价: {quote.last_price}, 成交量: {quote.volume}")

# 获取K线数据
klines = api.get_kline_serial("SHFE.cu2401", duration_seconds=60)  # 1-min bars
print(klines.tail())

# 关闭API
api.close()

代码格式

EXCHANGE.CONTRACT
交易所代码示例
上海期货交易所SHFESHFE.cu2401(铜)
大连商品交易所DCEDCE.m2405(豆粕)
郑州商品交易所CZCECZCE.CF405(棉花)
中国金融期货交易所CFFEXCFFEX.IF2401(沪深300期货)
上海国际能源交易中心INEINE.sc2407(原油)
广州期货交易所GFEXGFEX.si2407(工业硅)
上交所期权SSESSE.10004816(50ETF期权)
深交所期权SZSESZSE.90000001(300ETF期权)

行情数据

实时行情

from tqsdk import TqApi, TqAuth

api = TqApi(auth=TqAuth("user", "pass"))

quote = api.get_quote("CFFEX.IF2401")
# 关键字段:
# quote.last_price      — 最新价
# quote.bid_price1      — 买一价
# quote.ask_price1      — 卖一价
# quote.bid_volume1     — 买一量
# quote.ask_volume1     — 卖一量
# quote.highest         — 日最高价
# quote.lowest          — 日最低价
# quote.open            — 开盘价
# quote.close           — 昨收价
# quote.volume          — 总成交量
# quote.amount          — 总成交额
# quote.open_interest   — 持仓量
# quote.upper_limit     — 涨停价
# quote.lower_limit     — 跌停价
# quote.pre_settlement  — 昨结算价
# quote.settlement      — 今结算价

# 等待行情更新
while True:
    api.wait_update()
    if api.is_changing(quote, "last_price"):
        print(f"Price update: {quote.last_price}")

K线数据

# 获取K线序列(返回pandas DataFrame)
klines = api.get_kline_serial(
    "SHFE.cu2401",
    duration_seconds=60,     # K线周期:60=1分钟, 300=5分钟, 3600=1小时, 86400=日线
    data_length=200          # 获取K线数量
)
# 列:datetime, open, high, low, close, volume, open_oi, close_oi

# 多周期
klines_1m = api.get_kline_serial("SHFE.cu2401", 60)
klines_5m = api.get_kline_serial("SHFE.cu2401", 300)
klines_1d = api.get_kline_serial("SHFE.cu2401", 86400)

Tick数据

ticks = api.get_tick_serial("SHFE.cu2401", data_length=500)
# 列:datetime, last_price, highest, lowest, bid_price1, ask_price1,
#          bid_volume1, ask_volume1, volume, amount, open_interest

交易

下单

from tqsdk import TqApi, TqAuth

api = TqApi(auth=TqAuth("user", "pass"))

# 限价单 — buy open 2 lots
order = api.insert_order(
    symbol="SHFE.cu2401",
    direction="BUY",           # "BUY" or "SELL"
    offset="OPEN",             # "OPEN", "CLOSE", "CLOSETODAY"
    volume=2,                  # Number of lots
    limit_price=68000.0        # Limit price (None for market order)
)

# 市价单(FAK — 即成剩撤)
order = api.insert_order(
    symbol="SHFE.cu2401",
    direction="BUY",
    offset="OPEN",
    volume=2
)

# 撤单
api.cancel_order(order)

# 检查委托状态
while True:
    api.wait_update()
    if order.status == "FINISHED":
        print(f"Order finished: filled={order.volume_orign - order.volume_left}")
        break

持仓与账户

# 获取账户信息
account = api.get_account()
# account.balance        — 账户余额
# account.available      — 可用资金
# account.margin         — 已用保证金
# account.float_profit   — 浮动盈亏
# account.position_profit — 持仓盈亏
# account.commission     — 今日手续费

# 获取持仓
position = api.get_position("SHFE.cu2401")
# position.pos_long      — 多头持仓量
# position.pos_short     — 空头持仓量
# position.pos_long_today — 今多仓
# position.float_profit_long  — 多头浮动盈亏
# position.float_profit_short — 空头浮动盈亏
# position.open_price_long    — 多头平均开仓价
# position.open_price_short   — 空头平均开仓价

回测

from tqsdk import TqApi, TqAuth, TqBacktest, TqSim
from datetime import date

# 创建回测API
api = TqApi(
    backtest=TqBacktest(
        start_dt=date(2024, 1, 1),
        end_dt=date(2024, 6, 30)
    ),
    account=TqSim(init_balance=1000000),  # 模拟账户,初始资金100万
    auth=TqAuth("user", "pass")
)

# 策略逻辑(同一套代码适用于实盘和回测)
klines = api.get_kline_serial("CFFEX.IF2401", 60 * 60)  # 1-hour bars
position = api.get_position("CFFEX.IF2401")

while True:
    api.wait_update()
    if api.is_changing(klines.iloc[-1], "close"):
        ma5 = klines["close"].iloc[-5:].mean()
        ma20 = klines["close"].iloc[-20:].mean()

        if ma5 > ma20 and position.pos_long == 0:
            api.insert_order("CFFEX.IF2401", "BUY", "OPEN", 1, klines.iloc[-1]["close"])
        elif ma5 < ma20 and position.pos_long > 0:
            api.insert_order("CFFEX.IF2401", "SELL", "CLOSE", 1, klines.iloc[-1]["close"])

api.close()

进阶示例

双合约价差交易

from tqsdk import TqApi, TqAuth

api = TqApi(auth=TqAuth("user", "pass"))

quote_near = api.get_quote("SHFE.rb2401")   # Near-month rebar
quote_far = api.get_quote("SHFE.rb2405")    # Far-month rebar
pos_near = api.get_position("SHFE.rb2401")
pos_far = api.get_position("SHFE.rb2405")

SPREAD_OPEN = 100    # 开仓价差阈值
SPREAD_CLOSE = 20    # 平仓价差阈值

while True:
    api.wait_update()
    spread = quote_near.last_price - quote_far.last_price

    if spread > SPREAD_OPEN and pos_near.pos_short == 0:
        # 价差过大:卖近月,买远月
        api.insert_order("SHFE.rb2401", "SELL", "OPEN", 1, quote_near.bid_price1)
        api.insert_order("SHFE.rb2405", "BUY", "OPEN", 1, quote_far.ask_price1)
        print(f"Open spread trade: spread={spread:.0f}")

    elif spread < SPREAD_CLOSE and pos_near.pos_short > 0:
        # 价差收敛:双腿平仓
        api.insert_order("SHFE.rb2401", "BUY", "CLOSE", 1, quote_near.ask_price1)
        api.insert_order("SHFE.rb2405", "SELL", "CLOSE", 1, quote_far.bid_price1)
        print(f"Close spread trade: spread={spread:.0f}")

基于ATR的止损策略

from tqsdk import TqApi, TqAuth
import numpy as np

api = TqApi(auth=TqAuth("user", "pass"))

symbol = "CFFEX.IF2401"
klines = api.get_kline_serial(symbol, 86400, data_length=50)  # 日线
position = api.get_position(symbol)

ATR_PERIOD = 14
ATR_MULTIPLIER = 2.0
entry_price = 0.0

while True:
    api.wait_update()
    if not api.is_changing(klines.iloc[-1], "close"):
        continue

    # 计算ATR
    highs = klines["high"].iloc[-ATR_PERIOD-1:]
    lows = klines["low"].iloc[-ATR_PERIOD-1:]
    closes = klines["close"].iloc[-ATR_PERIOD-1:]
    tr = np.maximum(highs.values[1:] - lows.values[1:],
                    np.abs(highs.values[1:] - closes.values[:-1]),
                    np.abs(lows.values[1:] - closes.values[:-1]))
    atr = np.mean(tr[-ATR_PERIOD:])

    current_price = klines.iloc[-1]["close"]
    ma20 = klines["close"].iloc[-20:].mean()

    if position.pos_long == 0:
        # 入场:价格在20日均线之上
        if current_price > ma20:
            api.insert_order(symbol, "BUY", "OPEN", 1, current_price)
            entry_price = current_price
            print(f"Entry: price={current_price:.2f}, ATR={atr:.2f}")
    else:
        # ATR跟踪止损
        stop_price = entry_price - ATR_MULTIPLIER * atr
        if current_price < stop_price:
            api.insert_order(symbol, "SELL", "CLOSE", position.pos_long, current_price)
            print(f"Stop loss: price={current_price:.2f}, stop={stop_price:.2f}")

api.close()

使用技巧

  • 免费版提供延时行情(15分钟延迟),专业版支持实时数据。
  • 同一套代码适用于回测和实盘,只需修改API初始化。
  • api.wait_update() 是核心事件循环,所有数据更新通过它接收。
  • 使用 api.is_changing() 检查特定数据是否已更新。
  • 支持国内所有主要交易所的期货和期权。
  • 文档:https://doc.shinnytech.com/tqsdk/latest/

目标持仓助手(TargetPosTask)

TqSdk提供便捷的目标持仓管理工具,自动完成开平仓操作:

from tqsdk import TqApi, TqAuth, TargetPosTask

api = TqApi(auth=TqAuth("user", "pass"))
symbol = "SHFE.cu2401"
target = TargetPosTask(api, symbol)
klines = api.get_kline_serial(symbol, 86400, data_length=30)

while True:
    api.wait_update()
    if api.is_changing(klines.iloc[-1], "close"):
        ma5 = klines["close"].iloc[-5:].mean()
        ma20 = klines["close"].iloc[-20:].mean()
        if ma5 > ma20:
            target.set_target_volume(3)     # 自动调整到多头3手
        elif ma5 < ma20:
            target.set_target_volume(-3)    # 自动调整到空头3手
        else:
            target.set_target_volume(0)     # 自动平仓
api.close()

期权交易

from tqsdk import TqApi, TqAuth

api = TqApi(auth=TqAuth("user", "pass"))

# 获取期权合约列表
options = api.query_options("SSE.510050")  # 50ETF期权
print(f"共 {len(options)} 个期权合约")

# 获取特定期权行情
quote = api.get_quote("SSE.10004816")
print(f"最新价: {quote.last_price}")
print(f"隐含波动率: {quote.implied_volatility}")
print(f"Delta: {quote.delta}, Gamma: {quote.gamma}")
print(f"Theta: {quote.theta}, Vega: {quote.vega}")
api.close()

常见错误处理

错误原因解决方法
Authentication failed账户密码错误检查TqAuth的用户名密码
Insufficient margin保证金不足检查 account.available
Symbol not found合约代码错误或已过期检查代码格式和合约月份
Connection lost网络断开TqSdk会自动重连
Backtest data not available历史数据不可用检查回测日期范围

支持的K线周期

周期duration_seconds说明
3秒3超短线
5秒5超短线
10秒10短线
15秒15短线
30秒30短线
1分钟60最常用短线周期
5分钟300常用日内周期
15分钟900日内波段
30分钟1800日内波段
1小时3600中线周期
2小时7200中线周期
4小时14400中线周期
日线86400趋势跟踪

进阶示例:海龟交易策略

from tqsdk import TqApi, TqAuth, TargetPosTask
import numpy as np

api = TqApi(auth=TqAuth("user", "pass"))
symbol = "CFFEX.IF2401"
klines = api.get_kline_serial(symbol, 86400, data_length=60)
target = TargetPosTask(api, symbol)

ENTRY_PERIOD = 20    # 唐奇安通道入场周期
EXIT_PERIOD = 10     # 唐奇安通道出场周期
ATR_PERIOD = 20      # ATR周期
RISK_RATIO = 0.01    # 单笔风险比例

while True:
    api.wait_update()
    if not api.is_changing(klines.iloc[-1], "close"):
        continue

    highs = klines["high"].values
    lows = klines["low"].values
    closes = klines["close"].values

    if len(closes) < ENTRY_PERIOD + 1:
        continue

    # 唐奇安通道
    entry_high = np.max(highs[-ENTRY_PERIOD-1:-1])
    entry_low = np.min(lows[-ENTRY_PERIOD-1:-1])
    exit_high = np.max(highs[-EXIT_PERIOD-1:-1])
    exit_low = np.min(lows[-EXIT_PERIOD-1:-1])

    # 计算ATR
    tr = np.maximum(highs[1:] - lows[1:],
                    np.abs(highs[1:] - closes[:-1]),
                    np.abs(lows[1:] - closes[:-1]))
    atr = np.mean(tr[-ATR_PERIOD:])

    current = closes[-1]
    account = api.get_account()
    position = api.get_position(symbol)
    unit_size = max(1, int(account.balance * RISK_RATIO / atr))

    if position.pos_long == 0 and position.pos_short == 0:
        if current > entry_high:
            target.set_target_volume(unit_size)      # 突码20日高点做多
        elif current < entry_low:
            target.set_target_volume(-unit_size)     # 突码20日低点做空
    elif position.pos_long > 0:
        if current < exit_low:
            target.set_target_volume(0)              # 跌码10日低点平多
    elif position.pos_short > 0:
        if current > exit_high:
            target.set_target_volume(0)              # 突码10日高点平空
api.close()

进阶示例:多品种强弱对冲

from tqsdk import TqApi, TqAuth, TargetPosTask

api = TqApi(auth=TqAuth("user", "pass"))

# 黑色系品种:螺纹钢 vs 铁矿石
symbol_a = "SHFE.rb2405"
symbol_b = "DCE.i2405"
klines_a = api.get_kline_serial(symbol_a, 86400, data_length=30)
klines_b = api.get_kline_serial(symbol_b, 86400, data_length=30)
target_a = TargetPosTask(api, symbol_a)
target_b = TargetPosTask(api, symbol_b)
LOOKBACK = 20

while True:
    api.wait_update()
    if len(klines_a) < LOOKBACK or len(klines_b) < LOOKBACK:
        continue

    ret_a = (klines_a.iloc[-1]["close"] / klines_a.iloc[-LOOKBACK]["close"]) - 1
    ret_b = (klines_b.iloc[-1]["close"] / klines_b.iloc[-LOOKBACK]["close"]) - 1

    if ret_a > ret_b + 0.02:       # 强弱差大于2%
        target_a.set_target_volume(1)    # 做多强势品种
        target_b.set_target_volume(-1)   # 做空弱势品种
    elif ret_a < ret_b - 0.02:     # 强弱反转
        target_a.set_target_volume(0)
        target_b.set_target_volume(0)
api.close()


🤖 AI Agent 高阶使用指南

对于 AI Agent,在使用该量化/数据工具时应遵循以下高阶策略和最佳实践,以确保任务的高效完成:

1. 数据校验与错误处理

在获取数据或执行操作后,AI 应当主动检查返回的结果格式是否符合预期,以及是否存在缺失值(NaN)或空数据。

  • 示例策略:在通过 API 获取数据框(DataFrame)后,使用 if df.empty: 进行校验;捕获 Exception 以防网络或接口错误导致进程崩溃。

2. 多步组合分析

AI 经常需要进行宏观经济分析或跨市场对比。应善于将当前接口与其他数据源或工具组合使用。

  • 示例策略:先获取板块或指数的宏观数据,再筛选成分股,最后对具体标的进行深入的财务或技术面分析,形成完整的决策链条。

3. 构建动态监控与日志

对于交易和策略类任务,AI 可以定期拉取数据并建立监控机制。

  • 示例策略:使用循环或定时任务检查特定标的的异动(如涨跌停、放量),并在发现满足条件的信号时输出结构化日志或触发预警。

社区与支持

大佬量化 维护 — 量化交易教学与策略研发团队。

微信客服: bossquant1 · Bilibili · 搜索 大佬量化 — 微信公众号 / Bilibili / 抖音

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Coding

Iphone Add Music To Video

Turn a 30-second iPhone video clip and an MP3 song file into 1080p music-backed videos just by typing what you need. Whether it's adding background music to...

Registry SourceRecently Updated
Coding

Linkedin Video Maker

Get professional LinkedIn videos ready to post, without touching a single slider. Upload your video clips or images (MP4, MOV, AVI, WebM, up to 500MB), say s...

Registry SourceRecently Updated
Coding

Ps Tool

Display information about active processes. Use for monitoring running programs, checking resource usage, and system diagnostics.

Registry SourceRecently Updated
Coding

Serena

Use Serena-backed semantic code navigation and editing when working in existing software projects with non-trivial structure, especially multi-file repos whe...

Registry SourceRecently Updated