# Linear Regression

ICE Data Services -

upperLinearReg(), middleLinearReg(), lowerLinearReg() Built-in Study Functions

Linear Regression is a statistical tool used to predict future values from past values. In the case of security prices, it is commonly used to determine when prices are overextended. A linear regression trend line uses the least squares method to plot a straight line through prices so as to minimize the distances between the prices and resulting trend line.

Syntax
upperLinearReg(  nLength , nStdev [, Series | sym() | inv() ][, nBarIndex ] )
middleLinearReg( nLength , nStdev [, Series | sym() | inv() ][, nBarIndex ] )
lowerLinearReg(  nLength , nStdev [, Series | sym() | inv() ][, nBarIndex ] )

Parameters
Parameter: Description: Default:
nLength Required. Number of periods to use for calculation.

n/a nStdev Required. Number of standard deviations to use for calculation.

n/a Series or
sym() or
inv()
[Optional]   Series Object or function of sym() or inv() to determine symbol/interval source for the study.

Chart's sym/inv nBarIndex [Optional]  Bar index of series to retrieve.

n/a

Return Value(s)

Returns a Series Object when nBarIndex is not specified.

Returns a single value when nBarIndex is specified.

Notes
Only available in versions 7.9 or later.

Code Examples
Single Line Indicator:
 `function main() { return new Array(upperLinearReg(20, 1), middleLinearReg(20, 1), lowerLinearReg(20, 1) );}`

Single Line Study-on-Study Indicator:
 `function main() { return new Array(upperLinearReg(20, 1, hl2()), middleLinearReg(20, 1, hl2()), lowerLinearReg(20, 1, hl2()) );}`

Retrieve single values:
 ``function main() { var nValue_0 = middleLinearReg(20, 1, 0); // Current Bar Index var nValue_1 = middleLinearReg(20, 1, -1); // Prior Bar Index return new Array(nValue_0, nValue_1);}``

Initialize Series Objects:
 ``//global Series Object variablesvar xStudy1 = null;var xStudy2 = null;var xStudy3 = null;function main() { if (xStudy1 == null) xStudy1 = upperLinearReg(20, 1); if (xStudy2 == null) xStudy2 = middleLinearReg(20, 1); if (xStudy3 == null) xStudy3 = lowerLinearReg(20, 1); // retrieve single values for conditional statements var nValue1_0 = xStudy1.getValue(0); // Current Bar Index value var nValue2_0 = xStudy2.getValue(0); // Current Bar Index value var nValue3_0 = xStudy3.getValue(0); // Current Bar Index value var nValue1_1 = xStudy1.getValue(-1); // Prior Bar Index value // Plot Current Bar Index Value return new Array(nValue1_0, nValue2_0, nValue3_0); }``

Initialize a Series Object based on an external interval:
 ``// global Series Object variablesvar xStudy1 = null;var xStudy2 = null;var xStudy3 = null;var bInit = false; // Initialization flagfunction main() { if (bInit == false) { xStudy1 = upperLinearReg(20, 1, inv(20)); // 20-min interval xStudy2 = middleLinearReg(20, 1, inv(20)); // 20-min interval xStudy3 = lowerLinearReg(20, 1, inv(20)); // 20-min interval bInit = true; } // retrieve single values for conditional statements var nValue1_0 = xStudy1.getValue(0); // Current Bar Index value var nValue1_1 = xStudy1.getValue(-1); // Prior Bar Index value // Synchronized Series plot return new Array(getSeries(xStudy1), getSeries(xStudy2), getSeries(xStudy3) ); }``

Initialize a Series Object based on an external symbol and interval:
 ``// global Series Object variablesvar xStudy1 = null;var xStudy2 = null;var xStudy3 = null;var bInit = false; // Initialization flagfunction main() { if (bInit == false) { xStudy1 = upperLinearReg(20, 1, sym("IBM,20")); // IBM 20-min interval xStudy2 = middleLinearReg(``20, 1, sym("IBM,20")``); // ``IBM ``20-min interval xStudy3 = lowerLinearReg(``20, 1, sym("IBM,20")``); // ``IBM ``20-min interval bInit = true; } // retrieve single values for conditional statements var nValue1_0 = xStudy1.getValue(0); // Current Bar Index value var nValue1_1 = xStudy1.getValue(-1); // Prior Bar Index value // Synchronized Series plot return new Array(getSeries(xStudy1), getSeries(xStudy2), getSeries(xStudy3) ); }``