How vahini-1/hsmm-state-detection works
This is a quantitative regime-research framework: market data is turned into return/volatility features, modeled with offline HSMM inference or online particle filtering, then used for regime diagnostics, MS-VAR impulse responses, and strategy backtests. Its C++ engine provides explicit-duration HSMM primitives through a Python binding; Python also provides a Gibbs-sampling alternative and the broader research workflow. The README documents phase relationships, while several implementation details—especially unsampled CLI wiring and offline changepoint algorithms—are not independently verified here.
Main entry points: __init__.py Python HSMM API · backtest_strategy.py Backtest CLI · particle_filter.cpp Particle filter · data.py Data features
Components 21
Market data
- Data features
python/hsmm_regime/data.py - Feature CLI
python/scripts/build_features.py - Fetch CLI
python/scripts/fetch_data.py
Regime inference
- Offline fit CLI
python/scripts/fit_offline_hsmm.py - Online filter CLI
python/scripts/run_online_filter.py - Gibbs fit CLI
python/scripts/fit_gibbs_hsmm.py - Python HSMM API
python/hsmm_regime/__init__.py - Python binding
cpp/src/bindings.cpp - EM and decoding
cpp/src/em_engine.cpp - Forward-backward
cpp/src/forward_backward.cpp - Particle filter
cpp/src/particle_filter.cpp - Duration model
cpp/src/duration_distributions.cpp - Emission model
cpp/src/emissions.cpp - Bayesian sampler
python/hsmm_regime/gibbs.py
Research analysis
- Regime diagnostics
python/hsmm_regime/plotting.py - MS-VAR and IRFs
python/hsmm_regime/ms_var.py - MS-VAR CLI
python/scripts/fit_ms_var.py
Allocation
- Strategy backtest
python/hsmm_regime/strategy.py - Backtest CLI
python/scripts/backtest_strategy.py
Other components
- Researcher
- Market data service
Connections 26
- Researcher → Fetch CLI: requests data
README.md - Market data service → Data features: supplies OHLCV
python/hsmm_regime/data.py - Fetch CLI → Data features: fetches
python/scripts/fetch_data.py - Data features → Feature CLI: builds features
python/scripts/build_features.py - Feature CLI → Offline fit CLI: feeds fitting
README.md - Offline fit CLI → Python HSMM API: fits and decodes
python/scripts/fit_offline_hsmm.py - Python HSMM API → Python binding: calls extension
README.md - Python binding → EM and decoding: exposes fitting
cpp/src/bindings.cpp - Python binding → Particle filter: exposes filtering
cpp/src/bindings.cpp - Online filter CLI → Python HSMM API: streams observations
python/scripts/run_online_filter.py - Python HSMM API → Particle filter: steps filter
python/hsmm_regime/__init__.py - EM and decoding → Duration model: uses durations
cpp/src/em_engine.cpp - EM and decoding → Emission model: fits emissions
cpp/src/em_engine.cpp - Gibbs fit CLI → Bayesian sampler: runs sampler
python/scripts/fit_gibbs_hsmm.py - Offline fit CLI → Regime diagnostics: creates diagnostics
README.md - Online filter CLI → Regime diagnostics: plots posteriors
README.md - MS-VAR CLI → MS-VAR and IRFs: fits MS-VAR
python/scripts/fit_ms_var.py - MS-VAR CLI → Regime diagnostics: plots IRFs
python/scripts/fit_ms_var.py - Backtest CLI → Strategy backtest: runs backtest
python/scripts/backtest_strategy.py - Backtest CLI → Data features: loads features
python/scripts/backtest_strategy.py - Backtest CLI → Python HSMM API: uses regime API
python/scripts/backtest_strategy.py - Backtest CLI → Regime diagnostics: plots performance
python/scripts/backtest_strategy.py - Forward-backward → Duration model: uses durations
cpp/src/forward_backward.cpp - Forward-backward → Emission model: uses emissions
cpp/src/forward_backward.cpp - Particle filter → Duration model: uses durations
cpp/src/particle_filter.cpp - Particle filter → Emission model: uses emissions
cpp/src/particle_filter.cpp