Regression Calculator: Linear, Polynomial, Curve & Logistic Regression
A free regression calculator that fits simple and multiple linear regression, quadratic and cubic polynomial regression, exponential/power/logarithmic curves, and binary logistic regression. Paste in your data and it returns the regression equation, coefficients, and R² so you don't have to solve the normal equations by hand.
By Amanda Reeds, Content Researcher ·
This regression calculator fits the models covered in a typical statistics or data-analysis course, all in one page. Pick a mode above (Simple Linear, Multiple Linear, Polynomial, Exponential/Power/Log, or Logistic), paste in your x,y data, and the calculator returns the fitted equation, the coefficients, and a goodness-of-fit statistic.
It works as a general regression analysis calculator for the problems that come up most often: fitting a least squares regression line through two-variable data, running a multiple linear regression calculator with two or three predictors, fitting a quadratic regression calculator or cubic curve, fitting exponential, power, or logarithmic growth curves, and fitting a binary logistic regression calculator for 0/1 outcome data.
Five focused modes cover most of what a regression formula calculator needs to do without turning into an unfocused catch-all. Each one shows the intermediate statistics, such as the sums used in the normal equations or the number of fitting iterations, next to the final coefficients, so the output reads like a worked answer rather than a single unexplained figure.
What You Can Do With This Regression Calculator
Simple Linear Regression
A full linear regression calculator and least squares regression line calculator in one. Paste in x,y pairs and it fits y = b0 + b1x by least squares, reporting the slope, intercept, correlation coefficient, and R².
- Returns the regression equation, slope (b1), and intercept (b0)
- Reports the correlation coefficient r and R²
- Optionally predicts y for a new x-value
Multiple Linear Regression
A dedicated multiple linear regression calculator for 2 or 3 predictor variables. Enter rows of x1, x2, (x3,) y and it solves the normal equations for every coefficient at once.
- Handles 2-predictor and 3-predictor multiple regression equations
- Returns the intercept and each coefficient (b0, b1, b2, b3)
- Reports R² for the fitted multiple regression model
Polynomial Regression
A combined quadratic regression calculator and cubic regression calculator. Choose degree 2 or degree 3 and it fits the curve through your x,y data by least squares.
- Degree-2 (quadratic) and degree-3 (cubic) polynomial fits
- Returns every coefficient in the polynomial equation
- Reports R² so you can judge how well the curve fits
Exponential, Power & Logarithmic Regression
An exponential regression calculator, power regression calculator, and logarithmic regression calculator in a single submode selector, for data that grows, decays, or levels off rather than following a straight line.
- Exponential: y = a·b^x for growth and decay data
- Power: y = a·x^b for proportional-scaling data
- Logarithmic: y = a + b·ln(x) for data that levels off
Binary Logistic Regression
A logistic regression calculator for a 0/1 outcome. Enter x,y pairs where y is binary and it fits the logistic curve p = 1 / (1 + e^-(b0 + b1x)) using iteratively reweighted least squares.
- Fits binary logistic regression from raw x,y data
- Returns b0, b1, McFadden's pseudo-R², and training accuracy
- Optionally predicts a probability for a new x-value
How To Use The Regression Calculator
Pick the mode that matches your data, not just the shape of the scatterplot. If you have one predictor and a straight-line trend, that's Simple Linear even if you're tempted to reach for Polynomial; if you have two or three input columns feeding one output column, that's Multiple Linear even though each column alone could technically be run through Simple Linear.
- Choose a mode from the row above the calculator screen.
- In Simple Linear, Polynomial, Exponential/Power/Log, or Logistic mode, paste one x,y pair per line, for example
1,2. - In Multiple Linear mode, first pick 2 or 3 predictors, then paste one row per line as x1,x2,y or x1,x2,x3,y.
- In Exponential/Power/Log mode, pick the curve shape first, since each needs y > 0 and/or x > 0.
- Press the mode's Calculate button. The result panel lists the regression equation, every coefficient, and a goodness-of-fit statistic.
Quick Action Steps
- Pick the mode that matches your data's shape and number of predictors.
- Paste in every row you have; more rows generally means a more reliable fit.
- Press Calculate and read R² (or accuracy, for Logistic) before trusting the equation.
⚠ Watch Out For This
A low R² is not a calculator error; it means the chosen model does not explain much of the variation in your data. If Simple Linear gives a poor R² but the scatter clearly curves, try Polynomial or the Exponential/Power/Log mode instead. If Multiple Linear or Polynomial reports a fitting error, it usually means the data has too few rows for the number of coefficients, or the x-values repeat without enough variation to solve the normal equations.
How This Compares To Other Regression Tools
A Casio fx-991 series or TI-84 regression mode, an Excel LINEST/trendline formula, or an R or Python lm()/scikit-learn script all cover pieces of the same ground: fitting a line or curve through data. AceCalculator's Regression Calculator exists for the moments you want a straight numeric answer with the equation shown, without a physical calculator, spreadsheet formulas, or a coding environment.
- ✓ No download, install, or device needed, unlike a TI-84 or Casio fx-991
- ✓ No spreadsheet formulas or coding syntax to remember, unlike Excel, R, or Python
- ✓ Five focused modes in one page instead of one general-purpose input box
- ✓ Reports R², r, or pseudo-R² and accuracy alongside every fitted equation
- ✓ Handles linear, polynomial, curve-fit, and logistic regression in one place
- ✓ Runs entirely in your browser, nothing is stored or sent to a server
More Tools By AceCalculator.com
- Algebra Calculator: full algebra step-by-step solver
- Graphing Calculator: plot and visualize 2D functions
- Geometry Calculator: triangles, angles & coordinates
- Probability Distribution Calculator: binomial, normal & more
- Matrix Calculator: inverse, determinant, eigenvalues
- Derivative Calculator: basic, implicit & nth-order derivatives
- 3D Graphing Calculator: plot surfaces in three dimensions
- Math Calculators Hub: all maths tools in one place
Content Researcher · AceCalculator
Amanda Reeds researches and builds AceCalculator's math tools, with a focus on matching each calculator's output to standard classroom and statistics-course notation. She has written and verified the working behind AceCalculator's algebra, geometry, and equation-solving calculators.
Frequently Asked Questions
Yes. Every mode, including simple and multiple linear regression, polynomial regression, exponential/power/logarithmic curve fitting, and logistic regression, is free with no account, signup, or download required.
Enter your x,y pairs into the Simple Linear mode, one pair per line, and press Calculate. The calculator fits the least squares regression line and returns the slope (b1), intercept (b0), the equation y = b0 + b1x, the correlation coefficient r, and R².
Yes. The Multiple Linear mode fits a model with 2 or 3 predictor variables using ordinary least squares and returns the intercept, each coefficient, the full regression equation, and R².
Yes. The Polynomial mode fits a degree-2 (quadratic) or degree-3 (cubic) curve through your x,y data using least squares and returns every coefficient along with R².
Yes. The Exponential/Power/Log mode linearizes each model, for example ln y = ln a + x ln b for exponential regression, to solve for a and b, then reports R² measured against your original y-values.
The Logistic mode takes x,y pairs where y is 0 or 1 and fits a binary logistic regression model using iteratively reweighted least squares. It returns the coefficients b0 and b1, the logistic equation, McFadden's pseudo-R², and training accuracy, plus an optional predicted probability for a given x.
R², the coefficient of determination, is the share of the variation in y explained by the fitted model, from 0 (no fit) to 1 (a perfect fit). The calculator reports it for every mode except logistic regression, which uses McFadden's pseudo-R² instead since ordinary R² does not apply to a probability model.
It solves the same categories of problems as the regression functions on a Casio fx-991 series or a TI-84, such as linear, quadratic, and logistic regression, but it is an independent AceCalculator tool and is not affiliated with Casio or Texas Instruments.
Each mode displays the intermediate statistics, such as the sums used in the normal equations or the number of fitting iterations for logistic regression, alongside the final coefficients, rather than only a single output number.
Yes. Simple Linear and Logistic modes include an optional field to enter an x-value once the model is fitted, and the result panel adds the corresponding predicted y-value or predicted probability.
Important Disclaimer
AceCalculator's Regression Calculator is provided for informational and educational purposes only. It is an independent tool and is not affiliated with, endorsed by, or a product of Texas Instruments, Casio, Microsoft Excel, Desmos, or any statistical software vendor. Results are for reference and study and are not a substitute for verification by a qualified statistician or instructor where accuracy is critical. See our full Disclaimer and Terms and Conditions.