I've been very impressed by Tableau, it seems to be a very well thought out product - and definately a refreshing change from Office or SPSS (which we have been using for INFO 470). Excel doesnt even seem like it is capable of doing visualizations for anything but bivariate or trivariate data(in hindsight), and so that was cool to be able to do. I've definately found though my experience with it however that the hard part is really figuring out what variables will be interesting to look at and what wont. I also found that I enjoyed putting together visualizations multivariate data much more, it was more interesting to do I think.
One of the things that took me a little while to get used to was that the order of the variables in the shelves mattered, which I wasnt expecting and caught me off guard. Also, I found it difficult to get into the mentality where I would really rely on the undo feature. It's very powerful in what it lets you do to explore data, but I was always reluctant to do anything in the application that would "destroy" what I had constructed so far. I never did figure out how to effectively use the level of detail shelf, I found that it always seemed to make the display much less revealing because I ended up with a lot of unlabeled data.
I was a little curious why the export as image featured a few of the less common format, and not gif or jpeg. I am guessing this is for licensing reasons? It seems to me that gif would be very well suited for this kind of data, because it does not require too many colors and thus lends itself to an indexed color format. All in all, I think this is a really cool program to have, but it is obviously aimed at the corporate world - $1000+ is a little too expensive unless you were going to use it as a core part of your job. This restricts it to use to being data exploration tool, and not something that you would buy so you can make graphics for presentations and the like.


This was the section I had the most fun with, I spent a while just messing around with stuff trying to see if I could come up with something useful, and in the process found a lot of things that were definately not.