ml_forecast Probabilistic forecasting toolkit
The “emit P(event), don’t assert yes/no” pillar: an analytical volatility cone, an empirical session-open reach-probability counter, a robust (Theil-Sen) trendline, a pretrained logistic hazard head with its survival curve, total-variation distance for scoring a predicted distribution, and empirical-Bayes trust shrinkage.
Design: Pine has no RNG, so the cone is ANALYTICAL (z·σ·√h) — deterministic and non-repainting, not a random Monte-Carlo. The hazard head is a FORWARD PASS over weights you train OFFLINE (logistic regression on your labeled events) and pass frozen — the library does the inference, not the training. Stateful pieces (reach counters) keep their state HOST-side; this library owns the math.
Concept credit: volatility cones, empirical reach/touch probability, Theil-Sen robust regression, discrete-time survival hazards, total-variation distance, and empirical-Bayes shrinkage are standard statistical techniques — original, dependency-free Pine v6 packaging. Educational; not financial advice.
Library "ml_forecast"
volCone(price, sigmaPerBar, horizon, z)
Parameters:
price (float)
sigmaPerBar (float)
horizon (simple int)
z (simple float)
reachRecord(hits, maxExcursionPct, stepPct)
Parameters:
hits (array<float>)
maxExcursionPct (float)
stepPct (simple float)
reachProb(hits, k, sessions)
Parameters:
hits (array<float>)
k (int)
sessions (float)
theilSen(src, len)
Parameters:
src (float)
len (simple int)
logisticHazard(features, weights, bias)
Parameters:
features (array<float>)
weights (array<float>)
bias (simple float)
survivalProb(p0, h, k)
Parameters:
p0 (float)
h (float)
k (simple int)
survLen(p0, h, kmax)
Parameters:
p0 (float)
h (float)
kmax (simple int)
tvd(p, q)
Parameters:
p (array<float>)
q (array<float>)
trustWeight(trustCx, trustCc)
Parameters:
trustCx (float)
trustCc (float)
trustShrink(forecast, realized, trust)
Parameters:
forecast (float)
realized (float)
trust (float)
Design: Pine has no RNG, so the cone is ANALYTICAL (z·σ·√h) — deterministic and non-repainting, not a random Monte-Carlo. The hazard head is a FORWARD PASS over weights you train OFFLINE (logistic regression on your labeled events) and pass frozen — the library does the inference, not the training. Stateful pieces (reach counters) keep their state HOST-side; this library owns the math.
Concept credit: volatility cones, empirical reach/touch probability, Theil-Sen robust regression, discrete-time survival hazards, total-variation distance, and empirical-Bayes shrinkage are standard statistical techniques — original, dependency-free Pine v6 packaging. Educational; not financial advice.
Library "ml_forecast"
volCone(price, sigmaPerBar, horizon, z)
Parameters:
price (float)
sigmaPerBar (float)
horizon (simple int)
z (simple float)
reachRecord(hits, maxExcursionPct, stepPct)
Parameters:
hits (array<float>)
maxExcursionPct (float)
stepPct (simple float)
reachProb(hits, k, sessions)
Parameters:
hits (array<float>)
k (int)
sessions (float)
theilSen(src, len)
Parameters:
src (float)
len (simple int)
logisticHazard(features, weights, bias)
Parameters:
features (array<float>)
weights (array<float>)
bias (simple float)
survivalProb(p0, h, k)
Parameters:
p0 (float)
h (float)
k (simple int)
survLen(p0, h, kmax)
Parameters:
p0 (float)
h (float)
kmax (simple int)
tvd(p, q)
Parameters:
p (array<float>)
q (array<float>)
trustWeight(trustCx, trustCc)
Parameters:
trustCx (float)
trustCc (float)
trustShrink(forecast, realized, trust)
Parameters:
forecast (float)
realized (float)
trust (float)
Disclaimer
The information and publications are not meant to be, and do not constitute, financial, investment, trading, or other types of advice or recommendations supplied or endorsed by TradingView. Read more in the Terms of Use.
Disclaimer
The information and publications are not meant to be, and do not constitute, financial, investment, trading, or other types of advice or recommendations supplied or endorsed by TradingView. Read more in the Terms of Use.