Statistical Software

NIST users may access a list of statistical software programs available at NIST for performing statistical analyses.

NIST datasets for testing non-linear regression routines are available as package NISTnls for the R language and programming environment for statistical modeling, data analysis, and graphics.

• The package itself is not a NIST product: it was created by Douglas Bates, a member of the core group of R Project Contributors
• R is not a NIST product either. R provides a rich assortment of tools for statistical and graphical analysis; it is highly extensible, and Open Source and free. It is widely regarded as the vehicle of choice for research in statistical methods.

Web Applications

The following software applications are publicly available web applications developed by SED in collaboration with NIST scientists. Click the application names below to access the web-based versions.

• The NIST Uncertainty Machine (NUM) is a web-based software application used to evaluate the measurement uncertainty associated with a scalar or vectorial output quantity that is a known and explicit function of a set of scalar input quantities for which estimates and evaluations of measurement uncertainty are available. For more information see the "About" page provided in the application.
• The NIST Consensus Builder (NICOB) serves to combine measurement results obtained by different laboratories or by application of different measurement methods, into a consensus estimate of the value of a scalar measurand. For more information, see the "About" page provided in the application.
• The NIST ABACUS (App for Bayesian Analysis of Chemical quantities Using Shiny) is an online statistical software package that helps chemical analysts deliver metrologically sound measurement results. More information can be found here.
• The NIST Decision Tree is a web application that aids users in the selection of statistical models and data reduction procedures in key comparisons (KCs). More information about the project can be found here.
• The NIST Charpy Transition Curve Fitting Tool allows users to fit test results obtained from Charpy or toughness tests as a function of test temperature, thereby obtaining so-called transition curves. More details can be found here.
• The NIST COMET application (Counting Method Evaluation Tool), developed alongside scientists from the Biosystems and Biomaterials Division of MML, enables users to carry out statistical analysis as outlined in ISO 20391-2:2019 standard for Cell Counting. More information can be found here.
• The Atom Probe Peak Fitting application allows users to estimate isotopic abundances from Atom Probe Mass Spec data using a novel, nonparametric peak fitting algorithm. 

Software for Download

The following is a list of SED developed software programs. These are freely downloadable for both NIST and non-NIST users.

• Dataplot - a graphical data analysis program.
• Limits of Detection - limits of detection based on ASTM E-2677
• Recipe - a Fortran program for computing regression based tolerance limits.
• Extreme Winds - the extreme winds web site contains several custom Fortran programs and several MATLAB programs for analyzing extreme winds and the effect of winds on building structures.

The following software is available for download. However, it is not supported and is provided on an "as is" basis. There may not be any documentation available (other than that provided in the source code) for this software.

• Omnitab - the ancestor to MINITAB. Note that although we will provide this software on request, OMNITAB is no longer actively developed or supported.
• STSPAC - Fortran subroutines written by former SED staff member Charlie Reeve. In particular, these subroutines include functions for computing the cumulative distribution function and generating random numbers for the doubly non-central F and doubly non-central t distributions.
• DATAPAC - Fortran subroutines written by James Filliben. Most of these subroutines are incorporated into the Dataplot software listed above. DATAPAC has not been actively developed for some time. Of primary interest are some of the probability distribution functions.