Ugrás a fő tartalomhoz

Rezim Lab

Regime Lab

Regime detection and visualization on your backtest history — identify the trending and sideways periods for the era Strategy.

Open cBots → Regime Lab (/quant/regime).

What it does​

Enter a backtest instance ID or paste a return series and it identifies:

  • Trending vs. ranging periods using a rolling volatility/zscore filter.
  • Average length of each regime.
  • Return distribution per regime (did your strategy actually make money in trends, or only in range-bounces?).
POST /api/quant/regime
{ "returns": [0.006, 0.004, -0.002, ...] }

Why it is reliable​

Pure deterministic domain code (Core.Quant.RegimeLab) with no infrastructure dependency — unit-tested for regime boundary detection, edge cases (all-trending, all-range, single data point), and output consistency.