Browse

STEYX Function (LibreOffice Calc)

Math Intermediate LibreOffice Calc LibreOffice 3.0

The STEYX function in LibreOffice Calc returns the standard error of the predicted Y-values in a linear regression. This guide explains syntax, interpretation, examples, errors, and best practices.

The **STEYX** function returns the **standard error of the predicted Y-values** in a linear regression. It measures how far data points typically fall from the regression line — a key indicator of model accuracy and reliability.

This guide provides syntax, interpretation rules, examples, common errors, and best practices, making it one of the most complete STEYX references available for LibreOffice Calc.

What the STEYX Function Does

  • Calculates the standard error of the estimate
  • Measures the typical vertical distance between data points and the regression line
  • Works with numeric X/Y pairs
  • Useful for regression diagnostics and model evaluation
  • Works across sheets

A smaller STEYX means a better‑fitting model.

Syntax

STEYX(known_y; known_x)

Where:

  • known_y — dependent variable (Y values)
  • known_x — independent variable (X values)
STEYX is equivalent to the “standard error of Y estimate” returned by LINEST (row 3, column 2).

Interpretation of STEYX

STEYX Value Meaning
Near 0 Excellent fit — points lie close to the regression line
Small Good fit
Large Poor fit — high scatter around the line

STEYX is measured in the same units as Y.

Basic Examples

Standard error of regression

=STEYX(B1:B10; A1:A10)

Across sheets

=STEYX(Sheet1.B1:B50; Sheet2.A1:A50)

Using named ranges

=STEYX(Sales; MarketingSpend)

With dates as X-values

=STEYX(B1:B100; A1:A100)

(Calc converts dates to serial numbers.)

Advanced Examples

STEYX ignoring errors

=STEYX(IF(ISNUMBER(B1:B100); B1:B100); IF(ISNUMBER(A1:A100); A1:A100))

(Confirm with Ctrl+Shift+Enter in older Calc.)

STEYX using filtered (visible) data only

Use SUBTOTAL helper column to filter X/Y before passing to STEYX.

STEYX after removing outliers

=STEYX(FILTER(B1:B100; B1:B100<1000); FILTER(A1:A100; B1:B100<1000))

STEYX for log-transformed regression

=STEYX(LN(B1:B10); LN(A1:A10))

STEYX for exponential regression (manual)

Exponential model:

y = b * m^x

Linearized:

=STEYX(LN(Y1:Y20); X1:X20)

STEYX from LINEST (full regression)

=INDEX(LINEST(B1:B10; A1:A10; TRUE; TRUE); 3; 2)

How STEYX Calculates Standard Error of Estimate

The formula is:

[ SE = \sqrt{\frac{\sum (y_i - \hat{y}_i)^2}{n - 2}} ]

Where:

  • ( y_i ) = actual Y
  • ( \hat{y}_i ) = predicted Y from regression
  • ( n ) = number of data points

This is the standard deviation of the residuals.

Common Errors and Fixes

Err:502 — Invalid argument

Occurs when:

  • X and Y ranges have different sizes
  • One or both arrays contain no numeric values
  • Fewer than 3 data points (n - 2 must be > 0)

Err:504 — Parameter error

Occurs when:

  • Semicolons are incorrect
  • Range references malformed

STEYX returns unexpectedly large value

Possible causes:

  • Poor linear fit
  • High scatter in data
  • Outliers
  • Non-linear relationship

STEYX differs from LINEST

They are identical — LINEST simply returns more statistics.

Best Practices

  • Use STEYX to evaluate regression accuracy
  • Use RSQ to measure explained variance
  • Use LINEST for full diagnostics (F-statistic, df, etc.)
  • Remove outliers before modeling
  • Plot residuals to check for non-linearity
  • Use named ranges for cleaner formulas
STEYX tells you how “tight” your regression line is — a low value means your model is reliable, predictable, and statistically meaningful.