"There is pretty much no decent scripting language available on Windows, so if you do not have immediate access to some unix (or at least cygwin), you're out of luck."<p>He seems to be fully uniformed about what he writes about. The graphs he produced in Excel can be produced without Excel and just with open-source tools on practically any platform. Python and Gnuplot can be installed anywhere:<p><a href="http://www.python.org/" rel="nofollow">http://www.python.org/</a><p>"Python runs on Windows, Linux/Unix, Mac OS X, and has been ported to the Java and .NET virtual machines."<p><a href="http://www.gnuplot.info/" rel="nofollow">http://www.gnuplot.info/</a><p>"Gnuplot is a portable command-line driven graphing utility for Linux, OS/2, MS Windows, OSX, VMS, and many other platforms."
I'm all in favor of better python interfaces for excel, but you need a better example, computing rolling time series statistics, and visualizing them? That's a job for pandas.<p>If you want excel as the front end (I understand why you would want that) you could use pandas to process the data, and use excel as the frontend, instead of writing all that code.
I've played with gltail (can be found at <a href="http://www.fudgie.org/" rel="nofollow">http://www.fudgie.org/</a>) since yesterday. It should do what you want to achieve and run under Windows as well, though it's written in openGL and Ruby.
I don't understand why the author is using 'IronSpread' which is available on Windows only. For python there are modules like xlwt (www.python-excel.org) and openpyxl (<a href="https://bitbucket.org/ericgazoni/openpyxl/" rel="nofollow">https://bitbucket.org/ericgazoni/openpyxl/</a>) using which one
can generate excel files from python.