M3a toolchain: R 4.x + jsonlite only, all stats hand-rolled

Add R + jsonlite to SessionStart hook (minimal deps — no CRAN
statistical packages). Update M3a spec §3: GARCH, Heston, HMM, DCC,
copula, jump-diffusion, fBM all hand-rolled with base R primitives.
This commit is contained in:
Claude 2026-07-18 04:17:45 +00:00
parent e408409b77
commit 31b29cab32
No known key found for this signature in database
2 changed files with 15 additions and 4 deletions

View File

@ -74,6 +74,16 @@ else
log "fpm $(fpm --version 2>&1 | head -1)" log "fpm $(fpm --version 2>&1 | head -1)"
fi fi
# ---------------------------------------------------------------------------
# R + jsonlite (economy organ: M3a statistical sims)
# ---------------------------------------------------------------------------
if command -v Rscript >/dev/null 2>&1 && Rscript -e 'library(jsonlite)' >/dev/null 2>&1; then
log "R + jsonlite already present; skipping."
else
log "Installing r-base-core r-cran-jsonlite via apt-get ..."
sudo apt-get install -y r-base-core r-cran-jsonlite || die "apt-get install of r-base-core/r-cran-jsonlite failed."
fi
# --------------------------------------------------------------------------- # ---------------------------------------------------------------------------
# Tcl (economy organ: M3 sim hub) # Tcl (economy organ: M3 sim hub)
# --------------------------------------------------------------------------- # ---------------------------------------------------------------------------

View File

@ -16,10 +16,11 @@ execution C5 (industry standard since 2001). Jump-diffusion C5 (Merton 1976). Pa
for crypto markets C1. for crypto markets C1.
## 3. Language & location ## 3. Language & location
TBD · `src/economy/sims/statistical/`. **R** — native statistical distribution ecosystem, **R 4.x** (apt `r-base-core`) · `src/economy/sims/statistical/`. Minimal dependencies:
matrix operations, and time-series libraries (GARCH, ARIMA, HMM) without wrapping external `r-base-core` + `jsonlite` (Hub JSON protocol). All statistical models — GARCH, Heston SDE,
solvers. Fractional Brownian motion generation uses spectral methods (Hosking 1984, Wood & Chan HMM, DCC, copula, jump-diffusion, fBM — are hand-rolled using base R primitives (`optim`,
1994) or Cholesky decomposition of the covariance matrix. `fft`, `arima`, matrix ops). No CRAN statistical packages. Fractional Brownian motion via
spectral methods (Hosking 1984 / Wood & Chan 1994) uses base R `fft()`.
## 4. Does / does-not ## 4. Does / does-not
- **Does:** run Monte Carlo price simulations (GBM, Merton jump-diffusion, Heston stochastic - **Does:** run Monte Carlo price simulations (GBM, Merton jump-diffusion, Heston stochastic