MathGroup Archive 2007

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Developing Applications using Mathematica

  • To: mathgroup at
  • Subject: [mg77412] Re: Developing Applications using Mathematica
  • From: "atul" <atulksharma at>
  • Date: Fri, 8 Jun 2007 05:28:05 -0400 (EDT)
  • References: <f3r96l$t5f$>

Mark: There is another option that I'm surprised hasn't been mentioned. 
Although I can't claim to have applied it in a large scale setting, I have 
had occasion to develop a GUI front-end for colleagues who are not seasoned 
Mathematica users (nursing and technical staff in a children's hospital 
dialysis unit). This application uses Mathematica to  perform the background 
calculations for modeling of hemodialysis  quality control, 
what-if-scenarios, and individualized prescription development. Details of 
our model, if relevant to the scale of your project, may be found in our 
published references, which I would be glad to send you. They include 
Pediatr Nephrol (2001) 16:383-390 (a review of urea kinetic modeling in 
dialysis) and Kidney International, Vol. 58 (2000), pp. 2138-2146 (our 
specific physiologic model):

Since familiarity with Excel is widespread (it is the hospital standard for 
ordering, inventory, and routine data management and reporting), using 
Mathematica Link for Excel makes for a very familiar interface, and allows 
the results from Mathematica to be extracted and manipulated by users who 
might otherwise balk at working directly with Mathematica output (eg 
administrative staff making year-end reports).

On another occasion, I had to develop a pre-test evaluation for a novel test 
of kidney function that we perform in our Clinical Investigation Unit 
(again, model details - if required - may be found in our published 
description). Our CIU nurses were required to input baseline data to 
estimate pharmacokinetic profiles needed to perform the test, with modeling 
done behind the scenes and output in the form of loading doses and infusion 
rates for the individual studies. In this case, broad familiarity with Excel 
again made it a logical choice for interface/ data reporting. However, 
because the CIU represented a different cost-center and our university does 
not have a site license with WRI, I was obliged to adopt a 'rival CAS' with 
an existing site license and - as it turns out - built-in Excel 
connectivity. Although I grumbled about it at the time,  from the 
perspective of the end-users, the underlying calculation engine was 
obviously of no great concern, and in both cases, my work was immensely 
simpler because I had access to the built-in instruction sets.

Atul Sharma MD,
Pediatric Nephrologist,
Montreal Children's Hospital/McGill University

"Coleman, Mark" <Mark.Coleman at> wrote in message 
news:f3r96l$t5f$1 at
> Greetings,
> I'm curious about the feasibility of using Mathematica as part of 
> large-scale
> business application within my company. I know there are various ways
> one can make use of Mathematica kernel functionality using non-Mathematica 
> user
> interface elements, e.g., via MathLink or WebMathematica. I was wondering 
> if
> anyone could share experiences they have in building applications --
> particularly in a business rather than an academic environment -- using
> Mathematica. For my project, we need to distribute an application to over 
> 1000
> users across the U.S. If I could use Mathematica as the calculation 
> engine, the
> development process would be orders of magnitude simpler. I know that in
> theory this sort of this is possible, but I'm very curious as to the how
> it might be effectively done.
> And a follow-up question. Given the interesting new dynamic interface
> features in Mathematica v6, could one build the entire application in
> Mathematica, user interface and all, and then deploy it on a large scale
> to non-Mathematica users?
> Thanks,
> -Mark

  • Prev by Date: Pattern Matching Mathematica 6 versus 5.2?
  • Next by Date: Graphics3D exported to pdf
  • Previous by thread: Re: Developing Applications using Mathematica
  • Next by thread: possible mistake in FullGraphics documentation