TSF

Summary

Time Series Forecast: fits a least-squares linear regression line over the last N bars and projects it one x-step beyond LINEARREG. Same regression as LINEARREG but evaluated at x=period instead of x=period-1.

Formula

Fit y=b+mx over window (x=0..N-1): m = (NSumXY - SumXSumY)/(SumX^2 - NSumXSqr), b = (SumY - mSumX)/N; output = b + mN. With SumX=N(N-1)/2, SumXSqr=N(N-1)(2N-1)/6.

Inputs

inReal — Input series to regress and forecast

Outputs

outReal — Regression line value projected to x=period (one step past LINEARREG)

Parameters

optInTimePeriod — Number of bars in the regression window

Implementation

TA-Lib Definition: tsf.c · tsf.yaml

Native	File
C	`ta_TSF.c`
Rust	`tsf.rs`
Java	`Core_TSF.java`

TA-Lib is also available for Python, R and more using a wrapper.

Aliases

Time Series Forecast