Plotting

Tradetropy renders interactive Bokeh charts for both finished backtests (static) and live/replay sessions. Every indicator draws through one generic renderer, so built-in and custom indicators plot the same way with no plotting-package code.

Backtest charts

After a backtest, bt.plot() opens an interactive chart with the candles, your indicators (overlays on price and own-panel studies), trade markers and the equity curve:

from tradetropy import BacktestEngine
from tradetropy.datasets import load_btcusd_1m

bt = BacktestEngine.by_klines(SmaCross(), data=(load_btcusd_1m(),))
bt.run()
bt.plot()                      # opens in the browser
bt.plot(theme='dark')          # dark theme

PlotConfig controls the chart chrome (theme, sizing, ...):

from tradetropy import PlotConfig
bt.plot(config=PlotConfig(theme='dark'))

Themes and indicator colors

The theme styles only the chrome (background, grid, candles, axes). Indicator colors are fixed object attributes and are independent of the theme, so an indicator looks the same in light and dark mode. Set an indicator's colors on its object (constructor args or IndicatorPlotConfig), not via the theme.

Candle alignment

By default an indicator's drawn geometry (bubbles, zones, labels) is snapped to the candle grid so everything shares the candle time unit. Disable it for sub-candle microstructure precision:

bt.plot(align_indicators_to_candle=False)

Level of detail on large charts

A backtest chart with tens of thousands of candles automatically switches to a level-of-detail (LOD) view once the series exceeds 6000 candles (and no footprint is enabled, which is already a zoomed-in detail view). The full candle data stays intact; only the currently visible window is drawn, capped at 4000 candles. When the visible window is small enough, every real candle is shown 1:1; when it is not, the window is aggregated into buckets and each is reduced to one synthetic candle (open=first, close=last, high=max, low=min, volume=sum). Because each bucket keeps its true high/low, the Y-axis autoscale still frames the panel correctly - the reduction is purely a drawing optimization, refined automatically as you zoom in.

There is nothing to configure: LOD kicks in transparently based on the candle count, and the HoverTool on a bucketed candle describes the aggregated OHLCV for that time span rather than a single original candle.

Live and replay charts

Live, replay and paper-trading sessions use LiveChart, which streams the same glyphs in real time and adds trade overlays (closed-trade markers, average-price line, take-profit/stop-loss lines):

from tradetropy.plotting import PlotConfig
from tradetropy.plotting.live import LiveChart

chart = LiveChart(config=PlotConfig(theme='dark'), max_candles=200)
engine.run(live_chart=chart)

See Replay and paper trading and Live trading.