Regression Prediction Calculator

Fit a regression model to your data points and predict y from a given x, or x from y. Supports linear, quadratic, logarithmic, exponential and power models with R squared and a scatter plot.

Input

Enter your (x, y) data points, choose a regression model, and specify the value to predict.

Data points (x, y)

One pair per line, comma or space separated (for example 1, 2.1).

Regression model

Prediction direction

x to predict from

Finds the y for this x.

Result

Predicted y at x = 7

13.9867y

y = 0.0467 + 1.9914 x

Model

Linear y = a + b x

R squared

0.9985

Correlation r

0.9992

Data points

Scatter plot and regression curve

Data points

No.	x	y
1	1	2.1
2	2	3.9
3	3	6.2
4	4	7.8
5	5	10.1
6	6	12

How it works

Enter one (x, y) pair per line, separated by a comma or whitespace. Blank lines and invalid rows are ignored.
Models are fitted by least squares. Logarithmic, exponential and power models are linearized with logarithms and the coefficients are transformed back to the original scale, while the quadratic model solves the normal equations (a 3 by 3 linear system).
The coefficient of determination R squared comes from the residual sum of squares of observed versus predicted values. Values closer to 1 indicate a better fit. The correlation coefficient is not shown for the quadratic model.
Switch the prediction direction to choose forward prediction of y from x, or inverse prediction of x from y. The quadratic model does not support inverse prediction.
Logarithmic and power models require positive x, while exponential and power models require positive y. Change the model if your data does not meet these conditions.
In the scatter plot, data points are blue, the regression curve is a blue line, and the predicted point is red. The plot is a guide only, and extrapolation beyond the data range can carry large error.

Reviews

Tell us what you think of this calculator.