srfm::portfolio::MinkowskiCovariance Class Reference

#include <portfolio_manifold.hpp>

Public Member Functions
	MinkowskiCovariance (double c_market=constants::SPEED_OF_INFORMATION) noexcept
void	add_asset (AssetEvent event)
std::size_t	size () const noexcept
	Return the number of asset events currently stored.
std::optional< Eigen::MatrixXd >	compute_spacetime_covariance () const noexcept
std::optional< double >	interval_correlation (std::size_t i, std::size_t j) const noexcept
std::optional< manifold::IntervalType >	classify_pair (std::size_t i, std::size_t j) const noexcept
const std::vector< AssetEvent > &	events () const noexcept
	Read-only access to the stored asset events.
void	clear () noexcept
	Clear all stored events.

Detailed Description

Accumulates a set of AssetEvents and computes an NxN covariance matrix whose (i,j) entry is derived from the spacetime interval between asset i and asset j.

Algorithm

For each ordered pair (i, j): ds²(i,j) = −c²·Δt² + ΔP² + ΔV² + ΔM² where Δx = event_j − event_i in each coordinate.

The covariance entry C(i,j) is defined as: C(i,j) = exp(−|ds²(i,j)|) (Gaussian kernel over interval)

This maps:

• TIMELIKE pairs (ds² ≪ 0) → near 0 (strongly causal, large |ds²|)
• LIGHTLIKE pairs (ds² ≈ 0) → 1.0 (maximally correlated at light cone)
• SPACELIKE pairs (ds² ≫ 0) → near 0 (stochastic, acausal)

The diagonal is set to 1.0 (each asset is perfectly correlated with itself).

Usage

MinkowskiCovariance mc;
mc.add_asset(AssetEvent{"AAPL", 1.0, 150.0, 1e8, 2.4e12});
mc.add_asset(AssetEvent{"MSFT", 1.0, 290.0, 8e7, 2.1e12});
auto cov = mc.compute_spacetime_covariance();
// cov is a 2x2 Eigen::MatrixXd

Definition at line 97 of file portfolio_manifold.hpp.

Constructor & Destructor Documentation

MinkowskiCovariance()

srfm::portfolio::MinkowskiCovariance::MinkowskiCovariance ( double  c_market = constants::SPEED_OF_INFORMATION )

explicitnoexcept

Construct with optional speed-of-information parameter.

Parameters

c_market

Speed-of-information constant (default: 1.0).

Definition at line 43 of file portfolio_manifold.cpp.

Member Function Documentation

add_asset()

void srfm::portfolio::MinkowskiCovariance::add_asset

(

AssetEvent

event

)

Add an asset event to the manifold.

Events are stored in insertion order. The i-th added event becomes row/column i of the output covariance matrix.

Parameters

event

Asset event to add.

Definition at line 46 of file portfolio_manifold.cpp.

classify_pair()

std::optional< manifold::IntervalType > srfm::portfolio::MinkowskiCovariance::classify_pair ( std::size_t  i, std::size_t  j  ) const

noexcept

Classify the spacetime interval between asset pair (i, j).

Parameters

i	Index of first asset (0-based).
j	Index of second asset (0-based).

Returns

IntervalType (TIMELIKE / LIGHTLIKE / SPACELIKE), or nullopt if indices are out of range.

Definition at line 89 of file portfolio_manifold.cpp.

clear()

void srfm::portfolio::MinkowskiCovariance::clear ( )

noexcept

Clear all stored events.

Definition at line 58 of file portfolio_manifold.cpp.

compute_spacetime_covariance()

std::optional< Eigen::MatrixXd > srfm::portfolio::MinkowskiCovariance::compute_spacetime_covariance ( ) const

noexcept

Compute the NxN Minkowski covariance matrix.

Requires at least 2 assets; returns nullopt if fewer are stored.

Returns

NxN Eigen::MatrixXd where N = size(), or nullopt on failure.

Definition at line 102 of file portfolio_manifold.cpp.

events()

const std::vector< AssetEvent > & srfm::portfolio::MinkowskiCovariance::events ( ) const

noexcept

Read-only access to the stored asset events.

Definition at line 54 of file portfolio_manifold.cpp.

interval_correlation()

std::optional< double > srfm::portfolio::MinkowskiCovariance::interval_correlation ( std::size_t  i, std::size_t  j  ) const

noexcept

Compute the raw spacetime interval ds²(i, j) between asset pair (i, j).

Parameters

i	Index of first asset (0-based).
j	Index of second asset (0-based).

Returns

ds²(i, j), or nullopt if indices are out of range.

Definition at line 80 of file portfolio_manifold.cpp.

size()

std::size_t srfm::portfolio::MinkowskiCovariance::size ( ) const

noexcept

Return the number of asset events currently stored.

Definition at line 50 of file portfolio_manifold.cpp.

---

The documentation for this class was generated from the following files:

• include/portfolio_manifold.hpp
• src/portfolio_manifold.cpp