Appearance

Personalize your experience

Mode

Accent Color

Direction

C++ SDK Reference

C++ SDK

Header-only C++20 (packages/quatick-backtest-cpp): the position simulator, an event-driven engine with the same order rules as Python, analytics, a parallel optimizer, walk-forward, Monte Carlo and portfolios.

Headers

Header	Contents
bars.hpp	Bars, Params (name= arguments, QUATICK_PARAMS), read_csv, load_engine_bars
ta.hpp	sma, , rma, , highest, lowest, change, roc, , rsi, tr, atr, macd, bbands, crossover, crossunder
backtest.hpp	Execution (all cost and slippage models), simulate, compute_stats, backtest
engine.hpp	, Context, orders (market, limit, stop, stop-limit, trailing, brackets, OCO), run<S>()
.hpp	risk_metrics, trade_metrics, drawdown_periods, monthly_returns
.hpp	grid_search / search on a pool, walk_forward, deflated_sharpe
montecarlo.hpp	trade_monte_carlo ( / shuffle), stationary_bootstrap
portfolio.hpp	portfolio_backtest (target weights, shared capital, attribution, turnover)
event_main.hpp	QUATICK_EVENT_STRATEGY(S): the main that runs S on the Quatick engine

An event-driven strategy

breakout.cpp

1#include <quatick/quatick.hpp>23struct Breakout : quatick::Strategy {4  quatick::Series hi, atr;5  void init(quatick::Context& c) override {6    hi = quatick::highest(c.bars.high, p.geti("n", 20));7    atr = quatick::atr(c.bars, 14);8  }9  void on_bar(quatick::Context& c) override {10    const size_t i = c.i;11    if (c.position() != 0 || !c.open_orders().empty() || i == 0 || quatick::is_na(hi[i - 1])) return;12    if (c.close() > hi[i - 1]) {13      const double stop = c.close() - 2 * atr[i];14      c.buy(c.size_by_risk(1.0, c.close() - stop), {.tag = "breakout", .sl = stop, .tp_pct = 6});15    }16  }17};1819QUATICK_EVENT_STRATEGY(Breakout)   // upload as-is: the server supplies bars, params and settings

OrderOpts fields are designated initializers in declaration order: type, limit, stop, trail_pct, trail_amount, tif, expire_bars, tag, reduce_only, oco, sl, tp, sl_pct, tp_pct.

Local research

research.cpp

auto bars = quatick::read_csv("btc_1h.csv", "1h");
quatick::Execution ex;
ex.cost_kind = quatick::CostKind::CryptoMakerTaker; ex.cost_value = 0.02; ex.taker_pct = 0.05;
ex.max_drawdown_stop_pct = 25;

auto r = quatick::run<Breakout>(bars, {}, ex);
auto risk = quatick::risk_metrics(r.equity);          // var95, cvar95, omega, ...
auto tm = quatick::trade_metrics(r.trades);           // sqn, kelly, by_exit, by_side

auto eval = [&](const quatick::Bars& b, const quatick::Params& p) { return quatick::run<Breakout>(b, p, ex); };
auto runs = quatick::grid_search(eval, bars, {{"n", {10, 20, 40}}}, "sharpe", 8);   // 8 threads
auto wf = quatick::walk_forward(eval, bars, {{"n", {10, 20, 40}}}, 3000, 1000);

std::vector<double> pnl; for (auto& t : r.trades) pnl.push_back(t.pnl);
auto mc = quatick::trade_monte_carlo(pnl, ex.capital, 10000, quatick::McMethod::Bootstrap);

build

c++ -std=c++20 -O2 -pthread -I packages/quatick-backtest-cpp/include research.cpp -o research
# CMake: add_subdirectory(packages/quatick-backtest-cpp) and link quatick::backtest

Three ways to upload C++

File contains	The server
std::<double> (const quatick::Bars&, const quatick::Params&)	adds its own main and simulates the positions
a quatick:: and QUATICK_EVENT_STRATEGY(S)	runs the event engine with the run’s costs and settings
its own int main()	runs it as a program; it must write {"positions"\|"signals"\|""} to QUATICK_OUTPUT_JSON

ℹ

Note:The engine compiles every translation unit of a project with the and every project folder on the include , and caches the binary by source hash, so sweeps and compile once.

Appearance

C++ SDK Reference

Headers

An event-driven strategy

Local research

Three ways to upload C++

See also

Related articles

Appearance

Headers

An event-driven strategy

Local research

Three ways to upload C++

See also

Related articles

Related to C++ SDK Reference