XY One independent variable

TableCurve 2D

Give it a column of x and a column of y. TableCurve 2D fits all 3,665 of its built-in equations, sorts them by goodness of fit, and hands you a ranked list with the plot and the statistics for each one. The model you would have reached for is usually in there — along with several you would not have.

Ranked fit with residuals y = ax/(b+x) r² = —
What it does

The whole library, every time

Most fitting tools ask which equation you want. That question caps the result at whatever you happened to think of, and it rewards whoever has the best memory for functional forms rather than whoever has the best data. TableCurve 2D asks nothing and fits everything.

Ranked results

Every candidate is scored on r², adjusted r², the F-statistic and the standard error of the estimate. Sort by any of them and the list reorders.

Graph beside each fit

The review window plots the data, the fitted curve, confidence and prediction bands, and the residuals — so a fit that is numerically good but structurally wrong shows itself.

Code for the winner

Once you pick a model, TableCurve writes it out as source you can paste into your own analysis — coefficients, the function and its derivatives included.

Peak and transition families

Chromatographic and spectroscopic peak shapes, sigmoids and logistic transitions are first-class equation families, not something you have to hand-enter.

Your own equations

Enter a user-defined function in ordinary algebraic notation and TableCurve fits it with the same engine and reports it in the same ranking.

Reports that survive review

Numeric output, graphs and the full statistical summary export to Word and Excel, so what you publish is what the program computed.

From the program

Curves drawn by the fitting engine

Captures from TableCurve 9 itself. The peak and transition families below are built in — nothing here was entered by hand.

Four exponentially modified Gaussian curves plotted together, two leading and two tailing, on a shared axis from 0 to 10
Exponentially modified Gaussian Four parameter sets, leading and tailing
A single Gaussian peak of amplitude 95 centred at x equals 5, plotted from 0 to 10
Gaussian Amplitude 95, centre 5, width 1
The View and Compare Multiple 2D Data Sets window showing seven stacked traces over a rainbow contour field
View and compare multiple 2D datasets Seven nodes stacked over a contour field, with normalisation and identification options
How a fit runs

Four steps from data to model

  1. Bring the data in

    Paste from the clipboard, open a text or Excel file, or type into the data sheet. Columns become x and y.

  2. Run the library

    Choose which equation families to include, then fit. Every selected equation is fitted and scored.

  3. Review the ranking

    Walk the sorted list. Each row carries its plot, its coefficients and its residual pattern.

  4. Take the model out

    Export the report, the graph, or generated source code for the equation you settled on.

Specification

What is in the box

TableCurve 2D is the XY half of the TableCurve 9 program. It installs and licenses on its own, and upgrades to Studio by changing the licence key rather than the installation.

A 2D licence can open and view 3D functions, but surface fitting requires TableCurve 3D or Studio.

Data model
XY pairs, one independent variable
Equations
3,665 built-in, plus user-defined functions
Families
Polynomial, rational, exponential, logarithmic, power, peak, transition, kinetic
Precision
128-bit extended-precision solver for ill-conditioned fits
Statistics
r², adjusted r², F-statistic, standard error, coefficient confidence intervals
AI guidance
Copy for Claude.ai — hands over the equation, statistics and residuals
Graphics
Fitted curve, confidence and prediction bands, residual plots
Export
Word, Excel, clipboard, generated source code
Platform
Windows 10 and 11, 64-bit
Licensing
Annual or perpetual, academic or commercial

Try it on the dataset that is giving you trouble

The trial runs the full equation library. If nothing in 3,665 equations fits your data, that is worth knowing too.