BaseBot

class MyBot(qx.BaseBot):
    def __init__(self):
        self.tune = { ... }
        self.clamps = { ... }

Methods to override

`indicators(data)`

def indicators(self, data: Data) -> dict

Must override. Takes a Data object, returns a dict of indicator arrays keyed by name. Values are numpy arrays, one element per candle, all the same length.

`strategy(state, indicators)`

def strategy(self, state: dict, indicators: dict) -> Buy | Sell | Thresholds | Hold | None

Called every tick. state contains: last_trade, unix, wallet, and the current candle OHLCV fields (open, high, low, close, volume). indicators is the single-tick indicator values (scalars, not arrays).

Return a signal or None (no action). Default returns None.

`reset()`

def reset(self) -> None

Called between backtest runs during optimization. Override to clear any internal state so results don't bleed between runs. Default is no-op.

`execution(signal, indicators, wallet)`

def execution(self, signal: Signal, indicators: dict, wallet: WalletBase) -> Signal

Post-strategy hook. Called with the signal your strategy returned, the full indicator arrays, and the wallet. Modify the signal (e.g., override the execution price) or return None to cancel the trade. Default returns the signal unchanged.

`fitness(states, raw_states, asset, currency)`

def fitness(self, states, raw_states, asset, currency) -> tuple[list, dict]

Controls which metrics are computed and whether to add custom metrics. Returns a list of metric keys and an optional dict of custom metrics. Default returns (["roi", "cagr", "trade_win_rate"], {}).

Available metric keys: roi, cagr, sharpe_ratio, sortino_ratio, maximum_drawdown, calmar_ratio, omega_ratio, beta, alpha, info_ratio, profit_factor, trade_win_rate, payoff_ratio, skewness, kurtosis, efficiency_ratio, drawdown_duration, hurst_exponent, dpt.

Custom metrics are included as ind:name:low~high in results.

`plot(data, states, indicators, block)`

def plot(self, data, states, indicators, block) -> None

Override to customize the chart. Default calls qx.plot() with an empty indicator format. The axes list from qx.plot() gives you direct access to matplotlib Axes objects for custom drawing.

`autorange()`

def autorange(self) -> int

Returns the warmup period in days. The engine skips this many candles before calling strategy(). Auto-computed from tune keys ending in _period — takes the maximum value and rounds up. Returns 0 if no _period keys exist. Override when your custom indicator needs more warmup than any single tune parameter suggests.

Attributes your bot defines

Attribute	Type	Description
`self.tune`	`dict`	Tune parameters. Keys ending in `_period` are auto-scaled to candle size (treated as days). Values can be floats or numpy arrays.
`self.clamps`	`dict`	Clamp ranges in format `{param: [min, midpoint, max, clamp_flag]}`. Set `clamp_flag=0` on a parameter to skip it during optimization. You can also use the short form `{param: [min, max]}` — the engine expands it to 4 elements automatically.

Attributes the framework sets

Attribute	Type	Description
`self.info`	`Info`	Read-only dict. Keys: `mode` (`"backtest"`/`"papertrade"`/`"live"`), `start`, `live_data`, `live_trades`.
`self.gravitas`	`float`	Optional — used by the Gravitas optimizer. Not framework-enforced.

`Info`

Read-only dict attached to bot.info. Attempting to write raises TypeError.

Keys: mode (the current execution mode), start (start timestamp), live_data (high-resolution candle data during papertrade/live), live_trades (live fill records).