REVIEWS MATHSTATICA 2.5
mathStatica sees impressive developments with this upgrade – both in its own right and in its utilisation of new features in Wolfram Mathematica 8 (which it requires). According to the documentation, these are built on a 60 per cent expansion in the code base; I have to take that on trust, but the resulting benefits are empirically verifiable without any special effort. Starting the program, the visible changes
are a new and well-designed scrolling palette design to accommodate expanded content (with distributions now organised by domain) and a permanent entry in Mathematica’s palettes menu. Both, in themselves, make the product more productive. In use, the most immediately noticeable thing
is speed. Leaving aside the inherent advantage of mathStatica’s conceptually distinct symbolic approach to statistical calculations, which in itself makes everything considerably faster than Mathematica alone, this is down to improved efficiency and transparent parallel processing. Multiple cores are picked up and used without any user action or even awareness. I didn’t attempt to conduct scientific benchmarking, but my quick and dirty experiments with combinations of functions I use regularly
SIGMAPLOT 12
One of the most popular technical graphics products used in scientific research, SigmaPlot benefitted in its last release from wholesale import of analytic tools provided by its stable mate, SigmaStat. This decision pushed it firmly into the exploratory statistics segment of the market, which it had already penetrated on a de- facto basis for many users. The strategy continues in the current release, with the inclusion of enzyme kinematics, of which more is below.
Looking first of all at SigmaPlot’s core purpose, there are of course additions spread
between different visualisation types and functions. Macros are, by definition, things we could build for ourselves; but, in the real world, we often do not. One of the new prewritten examples in this release, well designed and serving as an example of what can be done, produces a dot density graph of unusual visual quality. Other new macros provide rich series exploration facilities and comparison vehicles for multiple methods, fitted lines or distributions. Curve fitting gains added subtlety with details such as parameter function weight variables. Subtyping adds sophistication to radar plots, while generic gradation and transparency controls extend the range of visual control across graphic types.
The interface sees
developments to keep SigmaPlot spruced up. Ribbons, tabs,
docking, MS Office style main button (where the file menu used to be), properties handling, and so on, keep the look and feel (and therefore productivity) in line with what is happening elsewhere in the productivity software universe. The addition of an enzyme kinematics module is a useful development which works well, providing appropriate new analyses and graphics that significantly extend the products reach, though it is an add-in rather than an inherent component. Compared to the overall image of SigmaPlot, it has, to my eye at least, the appearance of having been a late decision which will take another release version to fully bed in. It can be found on the right hand end of the Toolbox menu ribbon, though, as with some other imported components, no obvious reference to it is made in the help file.
The module operates, as do some other imported components, as a bolt-on facility with its own way of doing things rather than seamlessly flowing from the instincts which a SigmaPlot user has learnt in the package as a whole. On the other hand, it must be said in fairness that this is a small criticism of a big addition to a big product.
Analyses of enzyme assay are, in any case, specific to a particular user group, not part of the average statistical tool kit. The add-in appears next to the pharmacology group and those who want a kinematics tool will probably already know what they want to do with it and how to use it. The history of SigmaPlot development gives confidence of smoother integration to come. Reviewed in March 2012
www.sigmaplot.co.uk/products/ sigmaplot/
sigmaplot-details.php
indicator bars which, while they make no actual difference to anything, do subjectively make waiting times seem shorter! Moving on to the work itself, my favourite
bits were in the handling of multivariate problems, particularly the beautiful handling of multivariate discrete distributions. MDD usage is integrated into the new and improved territories of the Prob function, along with the improved piecewise handling which in many circumstances further enhances mathStatica’s separation from Mathematica in terms of quality and elegance. Non-rectangular domains are a potential
suggest that a dual core machine cuts delivery time for results by anything up to 40 per cent compared to a single core, with subsequent doubling of cores cutting the time by roughly the same proportion. Comparing mathStatica to raw Mathematica
over the same functions saw improvement by factors of up to 30, though somewhere between five and 10 was the average figure across all trials. On a dual core machine, a large shopping basket of varied work selected to be representative of my own real day to day activity was dealt with in roughly a third of the time required by the previous release. A welcome cosmetic touch is the presence of progress
headache to which mathStatica extends a new air of calm and simplicity, using Wolfram’s own piecewise provision in combination with standard domain expressions. Tere are, of course, as in most upgrades,
numerous other improvements, additions, tweaks and so on, plus those in the underlying Mathematica system from which mathStatica can cherry pick at will. Over all, taking both the major extensions and the incremental details together as a whole, the result is a thoroughly comprehensive upgrade to what was already a very impressive product that is effectively unique in its market niche. Reviewed in January 2012
www.mathstatica.com
16 BEYOND THE NUMBERS A STATISTICS SPECIAL
SCIENTIFIC COMPUTING WORLD
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52