Virtual Machine (JVM) instead of relying on
remote servers. For instance, users can view
a 3-D image of a protein or an oil field, which
would be less robust if the application were
only server-based. “JavaFX delivers the performance we need and provides excellent UI
controls right out of the box,” Egilsson adds.
Agust Egilsson
talks with author
David Baum about
QuantCell.
INDUSTRY APPLICATIONS
QuantCell Research is initially marketing its
solution to finance professionals immersed in
portfolio and risk analysis, because the solution lets those users take advantage of existing
Java-based “quant” libraries—along with other
Java-based libraries, such as the MapReduce
framework for big data analytics. This versatile
architecture lets them easily build advanced
models and financial products faster and
cheaper, and with less reliance on IT.
The company is also expanding into the
life sciences industry, where alpha and beta
users have found QuantCell ideal for analyzing genetic sequences, proteins, and markers.
“QuantCell’s spreadsheet-based environ-
ment is a convenient way to engage non-
experts in complex analytical research,” says
Dr. Styrmir Sigurjonsson, a senior statistician
at Natera, a life science company based in
Redwood City, California, involved in prenatal
genetic testing. “A statistician can develop
algorithms based on lab data and then pass
the QuantCell spreadsheet back to the lab
researcher to continue the analysis. You can
get to a whole new level of complexity with-
out having to bog down the researchers with
that complexity.”
Natera currently uses MATLAB, a popular
numerical computing environment and pro-
gramming language, to develop its prenatal
testing products. Sigurjonsson likes MATLAB’s
flexibility, saying researchers can “stop in the
middle of a program, run other programs, and
manipulate the data any way they like.” He has
been testing an alpha version of QuantCell to
see if it can fulfill this same role in a Java envi-
ronment. So far, he likes what he sees.
BRIGHT FUTURE
JAVA TECH
QuantCell’s
potential.”
ABOUT US
COMMUNITY SUPPORT
As QuantCell gains momentum in anticipation of a production release later in 2012, the
company is banking on continued support
from the Java community to deliver a large and
robust ecosystem of computational solutions.
“We decided to use Java in large part due
to the dynamic nature of Java compilation,
lazy class loading, concurrency in Java, and
optimization in the VM such as just-in-time
compilation,” says Thorleifsson. “The many
available libraries and the Java Community
Process will keep this community thriving for
a long time.”
It’s already a thriving ecosystem, fed largely
by Java experts at universities and open source
communities, which continue to contribute
advanced analytic libraries. Popular examples
include artificial intelligence and machine
learning (Apache Mahout, Java-ML, and
Weka), biotechnology (BioJava), and financial
mathematics (jQuantLib). Unfortunately,
most of these libraries are not immediately
usable by biotech researchers,
financial analysts, and other
domain specialists, who must
depend on skilled programmers
to build custom applications for
each analytical project.
QuantCell changes all that.
All Java APIs, libraries, and tools
work with QuantCell right out of
the box. End users can call upon
these libraries just by adding
them to the classpath, expand-
ing QuantCell’s existing catalog
of libraries, algorithms, and methods.
Applications and APIs created in NetBeans
or Eclipse can be moved into the QuantCell
environment for further work or testing, and
those assembled in QuantCell can be moved
into NetBeans for further development.
blog
David Baum is a freelance business, technology, and
lifestyle writer in Santa Barbara, California.
