Computing & Software

Prediction MAE, RMSE, Bias & R-Squared Calculator

Evaluate existing numeric predictions against observations without fitting a new model.

How the calculation works

Mean absolute error = prediction_mae(actual,predicted) input units. Root mean squared error = prediction_rmse(actual,predicted) input units. Mean signed prediction bias = prediction_bias(actual,predicted) input units. R-squared against observed-mean baseline = prediction_r2(actual,predicted) R².

Worked example

Example inputs — Actual values: 1, 2, 3; Predicted values in matching order: 1, 2, 4. Expected results — Mean absolute error: 0.3333333333 input units; Root mean squared error: 0.5773502692 input units; Mean signed prediction bias: 0.3333333333 input units; R-squared against observed-mean baseline: 0.5 R².

Assumptions and limits

Analytical software/hardware scenario, not a benchmark or a live provider quote. Use the actual architecture, workload and billing definitions; overheads not named in the model are excluded. Equal-length paired lists with nonconstant actual values for R². Negative R² is possible for poor predictions. This differs from fitting a regression line to the same observations.

Reference and further reading

Read the underlying subject reference. Calculith is an independent implementation; the reference does not endorse or certify this tool.

Related calculators

Browse all 1000 calculators →

Example values are editable. Results are estimates where assumptions apply. Calculations run in your browser. Updated 2026-11-16.