Results

Experiment.run returns an ExperimentResult. It contains experiment-level status and warnings plus one RunResult for each strategy (and the benchmark, when configured).

from backtide import DataExpConfig, Experiment, ExperimentConfig, GeneralExpConfig
from backtide.strategies import BuyAndHold

config = ExperimentConfig(
    general=GeneralExpConfig(name="Inspect Apple results"),
    data=DataExpConfig(symbols=["AAPL"]),
)
result = Experiment(config, strategies=[BuyAndHold()]).run()

print(result.status, result.warnings)
for run in result.strategies:
    print(run.strategy_name, run.metrics.get("total_return"), run.error)

Extract the useful parts

Each strategy result exposes four main collections:

• metrics is a dict[str, float] for quick ranking and reporting.
• equity_curve contains chronological EquitySample objects for equity and drawdown analysis.
• trades contains closed round trips, including entry, exit, quantity, and PnL.
• orders contains every processed order, including fills, cancellations, and rejections.

Use ordinary Python to select the run you need and convert records to tabular data:

import pandas as pd

successful = [run for run in result.strategies if run.error is None and not run.is_benchmark]
best = max(successful, key=lambda run: run.metrics.get("sharpe", float("-inf")))

metric_row = {"strategy": best.strategy_name, **best.metrics}
trades = pd.DataFrame(
    {
        "symbol": trade.symbol,
        "quantity": trade.quantity,
        "entry_ts": trade.entry_ts,
        "exit_ts": trade.exit_ts,
        "pnl": trade.pnl,
    }
    for trade in best.trades
)
equity = pd.DataFrame(
    {
        "timestamp": sample.timestamp,
        "equity": sample.equity,
        "drawdown": sample.drawdown,
    }
    for sample in best.equity_curve
)

Check result.status, result.warnings, and every run.error before comparing metrics. A partial experiment may still contain valid strategy results, but failed runs should not silently enter a ranking. Use Plots when the sequence and shape of results matters more than a scalar.