|
Special Relativity in Financial Modeling 1.0.0
Lorentz transforms, spacetime classification, and geodesic price paths for quantitative finance
|
Rolling-window z-score normaliser for tick close prices. More...
#include <algorithm>#include <cassert>#include <cmath>#include <cstddef>#include <stdexcept>#include <string>#include <vector>Go to the source code of this file.
Classes | |
| class | srfm::stream::CoordinateNormalizer |
| Rolling z-score normaliser using Welford's online algorithm. More... | |
Namespaces | |
| namespace | srfm |
| namespace | srfm::stream |
Rolling-window z-score normaliser for tick close prices.
Module: include/srfm/stream/ Owner: AGT-10 (Builder) — 2026-03-01
Map raw close prices into a normalised coordinate space using a rolling Welford online mean/variance estimator over a configurable window.
normalised = (close − μ_window) / σ_window
This dimensionless coordinate feeds into BetaCalculator and LorentzTransform, ensuring the relativistic machinery operates on unit-free quantities.
• Online: O(1) per tick, no recomputation over the full window. • Bounded: memory is exactly window * sizeof(double) + a few scalars. • Zero-allocation: circular buffer is pre-allocated at construction. • Non-throwing: update() and normalise() are noexcept. • Pre-warm: returns 0.0 until window is full (conservative).
• Cross-field normalisation (high/low/volume normalised separately if needed) • Persistence or serialisation of the rolling state • Thread safety — single-threaded use only
Definition in file coordinate_normalizer.hpp.