There are plenty of third-party Excel add-ons available (we’ll
look at some of these next month) and we’ve already looked
at some of the leading graphics software. However for heavyweight
applications and exploratory data analysis, design engineers are
likely to turn to one of the leading-edge mathematical programs
that takeadvantage of some serious maths routines to manipulate
critical data.
Mathcad is one of the most popular maths-based
applications for engineers, and that popularity is likely to increase
now that they have introduced an add-on Data Analysis Extension
Pack. Compatible with the latest Mathcad version 11 and alsowith
Mathcad Enterprise, this offers a wide range of techniques for
managing data in Mathcad. It handles multiple datasets in many
different file formats, which you can evaluate visually and qualitatively
to determine the best course of analysis, and boasts a pretty comprehensive
library of functions, including utilities, statistics, outliers,
regression and splines capabilities and matrix utility functions.
One of the most impressive features of this new Extension Pack
is its ability to handle problem datasets, such as very small or
very large scale data which are usually difficult to fit. Data
with inconsistent column formatting, data with hundreds of measurements
that must be reduced to a more compact representation, and data
with missing or suspected outliers are all within the Pack’s
capabilities. It lets you perform principle component analyses
for multivariate data; generate probability plots, including normal
and Weibull plots; use statistical functions for exploratory data
analysis; detect outlier and missing values; and determine confidence
limits.
Matlab (www.mathworks.com) claims to support the entire data analysis
process, and dedicated Matlab users will certainly find a great
deal of data analysis capability without having to buy an addon
toolbox. An Import Wizard supports most standard file formats as
well as direct access to data from serial ports and TCP/IP connections.
An add-on Data Acquisition Toolbox lets you stream live, measured
data directly into Matlab for analysis andvisualisation; other
Toolboxes are available to communicate with GPIB devices and with
databases.
Commonly used methods for 1-D, 2-D and generalised multi-dimensional
data analysis are built directly into Matlab.Among the functions
you can perform are correlations, basic statistics and curvefitting,
Fourier analysis and filtering and matrix analysis. If you need
more advanced statistics and curve-fitting, or more advanced mathematical
functions such as statistics and partial differential equations,
you can buy Toolboxes to handle these. Add-ons are also available
for both Matlab and Mathcad for signal processing, image processing
and wavelets.
The programming features of Mathematica (www.wolfram.com)
allow you to augment this symbolic computing package with a variety
of data analysistools. There are many dozens of such packages available,
both from Wolfram Research and from third-party developers, ranging
from control and optimisation to wavelet theory, neural networks
and fuzzy logic. The Experimental Data Analyst package, for example,
provides extensive data fitting, visualisation and error analysis
capabilities.
Maple (maple.adeptscience.co.uk) takes a different approach, showing
its academic origins by offering a large selection of add-on packages
free to users of the software. The Web site offers a huge range
of downloadable PowerTools and Application Packs, and while the
majority of these are aimed at the education market, there’s
a broad selection of tools that design engineers will find useful,
covering areas from finite element analysis to mathematical modelling,
from fluid dynamics to signal processing, and covering mechanical,
structural, chemical and electronic engineering.
Dr Know's recommended download is the Mathcad and Data Analysis Extension Pack brochure - download yours today.