// 05 · regime detection

REGIME DETECTION

Hidden Markov Models on your own data.

Upload a CSV. Viterbi parses it, infers regimes via Baum-Welch, decodes the most-likely state path, and shows you per-regime statistics. No data leaves your browser.

// drop a csv here, or click to choose

UPLOAD

TradingView, NinjaTrader, Tradovate, Quantower, or anything that saves to .csv with a header row. Trade logs and time-series both supported.

CHOOSE FILE

// time-series

Returns or prices per period

One row per day, hour, or bar. A column named return, close, or equity is detected automatically. Use this mode for HMM, GARCH, ARIMA, and Monte Carlo on equity curves.

// trade-log

One row per trade

Entry time, exit time, side, and P&L per trade. Detected when the CSV has trade-shape columns. Suited for expectancy, win-rate confidence intervals, and prop firm simulations. Native exports from NinjaTrader, Tradovate, and TradingView all fit here.

// requires a header row · supports csv, tsv, txt

// fin