Dear R users, I have been using R for 10 years, and I love it very much. But in my daily job for drug discovery, some people use Spotfire. I tried Spotfire on couple of data sets. It sounds I still need do some data manipulation before plot figures. For example, I can not plot figure with data arranged in rows (is this true, or I am stupid?). So far I don't feel any benefit Spotfire can provide over R. I am just wondering whether it just because I am new to Spotfire, or it's true that Spotfire is not a good tool for statistician. Also could anyone give me any suggestion how to learn Spotfire? Thanks John [[alternative HTML version deleted]]
On 12-01-11 10:13 AM, John Smith wrote:> Dear R users, > > I have been using R for 10 years, and I love it very much. But in my daily > job for drug discovery, some people use Spotfire. I tried Spotfire on > couple of data sets. It sounds I still need do some data manipulation > before plot figures. For example, I can not plot figure with data arranged > in rows (is this true, or I am stupid?). So far I don't feel any benefit > Spotfire can provide over R. I am just wondering whether it just because I > am new to Spotfire, or it's true that Spotfire is not a good tool for > statistician. > > Also could anyone give me any suggestion how to learn Spotfire?Shouldn't you be asking this question to Spotfire users? Duncan Murdoch
John, Spotfire is a menu driven data exploration tool, very popular here with biologists who found that their previous Excel based approach doesn't cut it for large data sets. When TIBCO wanted to expand the tool with further quantitative features they made (I think) a bright decision to purchase S-plus and integrate it as a back end, rather than try to write dozens of new modules in house. The "Splus vs R" aspect of the list responses misses the main point, however. Spotfire is designed to let you nose around in a data set, quickly plotting various aspects, zoom in on subsets (imagine a mouse based version of the "pinch" metafor used on the iphone), etc. It is a useful and very well designed tool; one demo was enough to make the sale and early growth here was explosive. But if you already know R you can do those graphs already, albeit quite a bit slower. I decided not to persue proficiency in Spotfire, but that was partly because it's Windows based and I prefer Unix. Also most of my work is at the post-exploration phase, and I would have flipped back to straight R for that anyway. Terry Therneau