CurveExpert Professional 2.7.3 documentation

What is CurveExpert Professional?

CurveExpert Professional is a cross-platform solution for curve fitting and data analysis. Data can be modelled using a toolbox of linear regression models, nonlinear regression models, smoothing methods, or various kinds of splines.

Over 60 models are built-in, but custom regression models may also be defined by the user. Full-featured publication-quality graphing capability allows thorough examination and presentation of the curve fit. The process of finding the best fit can be automated by letting CurveExpert Professional compare your data to each model to choose the best curve. The software is designed with the purpose of generating high quality results and output while saving your time in the process.

CurveExpert Professional compares very favorably to the competition and does so at a lower price; for an interactive comparison, visit http://www.curveexpert.net/products/compare-to-the-competition. The major features of the software are enumerated below:

  • Multiplatform: runs on Windows, Mac, and Linux. Saved files are portable between platforms.

  • Easy-to-use User Interface: most mathematically-intensive applications are very difficult to use. CurveExpert Professional has a very intuitive interface, which allows you to import your data, generate results, and create publication-quality plots with very minimal effort. In fact, to import a file takes only four clicks, and generating a battery of results with associated graphs takes two more.

  • Robust file import: data files come in many shapes and sizes, and CurveExpert Professional makes importing your data files very easy. The smart file reader avoids non-data areas of your file dynamically, and attempts to find labels for each column of data in your file.

  • Publication quality graphs: an arbitrary number of graphs can be created and saved. The rendering of the plots is of publication quality, with full antialiasing support and the ability to extensively customize each graph. Graphs can be saved to a variety of graphics file formats, and they may be directly copied and pasted into another application. For the purposes of 3D plotting, scatterplots, surfaces, and contour plots are supported. Graph themes allow you to customize a look that you prefer, and reuse it. Graphs are interactive, with zooming, panning, autoscaling, and view stack capability.

  • Multivariate: linear and nonlinear regressions can be computed for datasets with multiple independent variables, without any limits.

  • Multicore: put your extra CPU cores to work and compute results more quickly. A typical speedup is 3.5X on a 4 core system.

  • Built-in models: over 60 built-in nonlinear models, with high-quality automatic initial guesses, are available for use. The provided models cover all of the major families.

  • Detailed Result Query: parameter values and uncertainties, covariance matrix, parameter histories, residual plots. Forward and backward evaluation of the model’s value, differentiation, integration, and table generation. Residual runs test, F-Tests, Akaike’s Information Criterion Test. All results formatted beautifully in HTML reports, with fully customizable numerical formatting.

  • Custom models: you can also define models yourself, using a very large library of built-in mathematical functions, and parameters in your models can take any name that you like. For more advanced users, a model can be as complex as necessary, as long as it can be expressed as a Python function. This allows the creation of extremely complicated models, instead of the normal one-line-equation models limitation in other software. Further, the number of parameters is unlimited.

  • Functions: much like models, a suite of built-in functions (over 40) can be evaluated, and you can define your own custom functions as well.

  • Ranking of results: results are automatically ranked by your choice of score, correlation coefficient, standard error, or coefficient of determination.

  • Weighting: Support for six different weighting schemes for weighted linear and nonlinear regression.

  • Validated: validated against the Statistical Reference Datasets Project of the National Institute of Standards and Technology. These datasets can be downloaded directly at http://www.itl.nist.gov/div898/strd/general/dataarchive.html, but are also included verbatim in the CurveExpert Professional distribution for you to use yourself.

  • Quality spreadsheet: the built-in spreadsheet allows you to manually enter data and/or modify it with a suite of data transformation tools. Data entry and cutting and pasting capabilities are as easy as Excel.

  • Localization: Importing data or interoperating in European-style environments (which use a comma as a decimal) is extremely easy; regional settings are automatically obeyed, or can be selectively enabled in order to match your particular workflow.

  • Logging: a log of actions is kept across sessions of the software, in case you need to recreate a particular result. A messages pane keeps you informed of the status of every computed result.

  • Documentation: Extensive documentation in HTML and PDF format, available both directly from the software and online at https://www.curveexpert.net/support/documentation/.