visualise_grid_search_equity_curves#

API documentation for tradeexecutor.visual.grid_search.visualise_grid_search_equity_curves Python function.

visualise_grid_search_equity_curves(results, name=None, benchmark_indexes=None, height=1200, colour=None, log_y=False, alpha=0.7)[source]#

Draw multiple equity curves in the same chart.

  • See how all grid searched strategies work

  • Benchmark against buy and hold of various assets

  • Benchmark against hold all cash

Note

Only good up to ~hundreds results. If more than thousand result, rendering takes too long time.

Example that draws equity curves of a grid search results.

from tradeexecutor.visual.grid_search import visualise_grid_search_equity_curves
from tradeexecutor.analysis.multi_asset_benchmark import get_benchmark_data

# Automatically create BTC and ETH buy and hold benchmark if present
# in the trading universe
benchmark_indexes = get_benchmark_data(
    strategy_universe,
    cumulative_with_initial_cash=ShiftedStrategyParameters.initial_cash,
)

fig = visualise_grid_search_equity_curves(
    grid_search_results,
    name="8h clock shift, stop loss added and adjusted momentum",
    benchmark_indexes=benchmark_indexes,
    log_y=False,
)
fig.show()
Parameters:

  • results (List[GridSearchResult]) – Results from the grid search.

  • benchmark_indexes (pandas.core.frame.DataFrame | None) –

    List of other asset price series displayed on the timeline besides equity curve.

    DataFrame containing multiple series.

    • Asset name is the series name.

    • Setting colour for pd.Series.attrs allows you to override the colour of the index

  • height – Chart height in pixels

  • colour – Colour of the equity curve e.g. “rgba(160, 160, 160, 0.5)”. If provided, all equity curves will be drawn with this colour.

  • start_at – When the backtest started

  • end_at – When the backtest ended

  • additional_indicators –

    Additional technical indicators drawn on this chart.

    List of indicator names.

    The indicators must be plotted earlier using state.visualisation.plot_indicator().

    Note: Currently not very useful due to Y axis scale

  • log_y –

    Use logarithmic Y-axis.

    Because we accumulate larger treasury over time, the swings in the value will be higher later. We need to use a logarithmic Y axis so that we can compare the performance early in the strateg and late in the strategy.

  • name (str | None) –

Return type:

Figure