LINEST in WPS Spreadsheet: Fit Linear Coefficients and an Intercept

With x values 1,2,3 and y values 5,8,11, =LINEST(B2:B4,A2:A4,TRUE,FALSE) is independently expected to return the coefficient row 3, 2: slope first, then intercept. The model is y=3x+2.

LINEST(known_y's,known_x's,const,stats)

• known_y's: required numeric response observations.
• known_x's: optional numeric predictors. Omission uses 1,2,3,… matching the y dataset.
• const: optional logical flag. TRUE or omission estimates b in y=mx+b. FALSE fixes b=0 and refits the slopes.
• stats: optional logical flag. FALSE or omission returns coefficients and intercept; TRUE requests additional regression statistics.

With one predictor, known x and y must have the same dimensions. With multiple predictors, known y must be one row or column. Each x column is one predictor when y is a column; each x row is one predictor when y is a row. WPS documents the coefficient ordering as {mn,mn−1,…,m1,b}, so multiple predictor coefficients appear in reverse predictor order, followed by the intercept.

Enter 1,2,3 in A2:A4 and 5,8,11 in B2:B4. Reserve D2:E2 for two coefficient cells and enter =LINEST(B2:B4,A2:A4,TRUE,FALSE) using the array-output method supported by your build. Independently expect D2=3 and E2=2.

The historical WPS Windows walkthrough selects the two-cell output, presses F2, then Ctrl+Shift+Enter to expose both coefficients. That is evidence for its demonstrated workflow, not a blanket claim about dynamic spilling in every current version. If only 3 is visible, do not assume the intercept is missing from the mathematical result.

Independently, the centered least-squares slope is Σ(x−x̄)(y−ȳ)/Σ(x−x̄)² = 6/2 = 3, and the intercept is ȳ−3x̄=8−6=2. Use both coefficients to predict: at x=4, 3×4+2=14.

=LINEST(B2:B4,A2:A4,FALSE,FALSE) fits y=mx. The independent slope is Σxy/Σx²=54/14=27/7, approximately 3.857142857143; the intercept is fixed at 0. It is not the original slope 3 with the old intercept discarded. A physical zero-origin requirement may justify this model, but choosing it merely to simplify the output can bias a fit.

With stats=TRUE, the first-party text lists standard errors of coefficients and intercept, r², standard error of the y estimate, F, residual degrees of freedom, regression sum of squares and residual sum of squares. The coefficient row remains the starting point. This guide does not supply a cell-by-cell full-statistics extraction recipe: the source’s illustrated layout and current WPS array behavior have not been visually verified.

A perfect-fit teaching dataset has zero residual error and can make some inferential statistics degenerate. Use an appropriate noisy dataset, adequate residual degrees of freedom and the installed function help before interpreting F or coefficient uncertainty. A high r² is not proof of causation or of accurate extrapolation.

Check predictor/response orientation and observation alignment. Identical x values cannot identify a unique single-predictor slope with an intercept. Redundant predictor columns can make coefficients nonunique. Do not treat these as formatting problems. Keep array output cells available and inspect local separator requirements.

If predictions rather than coefficients are the goal, TREND is the related operation. Keep model assumptions, estimation uncertainty and out-of-range prediction risk separate from whether a formula accepts its inputs.

The official WPS LINEST lesson was checked on October 19, 2026. Its edition banner identifies WPS Office for Windows Personal. This Chinese first-party documentation is not proof of universal English-global, Mac, Linux, web or mobile support. Introduction and retirement versions remain unknown. All numerical results in this guide are independently calculated expectations, not native WPS test results or borrowed Excel outputs. Current GUI access was unavailable; no current-application screenshots or native execution are claimed.

INTERCEPT, RSQ. Browse the WPS function library.