Fwd: [rCharts] Learning rCharts (#491)
Agustin Lobo
From: Eugene Geis <notifi...@github.com>
Date: Tue, Aug 12, 2014 at 4:00 PM
Subject: [rCharts] Learning rCharts (#491)
To: ramnathv/rCharts <rCh...@noreply.github.com>
I'm curious about why half of the linked plots from these discussions do not work, or in other words, show up completely blank on my Chrome browser? A great deal of the examples that are linked on rcharts.io do not appear at all. For example, http://rcharts.io/viewer/?6644957#.U-k7IeNdXAQ
The association of the commands rPlot, nPlot, rCharts::Highcharts$new(), mPlot, hPlot, xPlot, Leaflet$new(), etc, with their respective javascript libraries, doesn't necessarily make sense as it seems like rPlot works with several JS libraries, nPlot works with NVD3, others work with only the respective Highcharts, MorrisJS, Leaflet, or... where does dimple fit in, or ggvis, or rVega??? Or when you call these objects in shiny, you need a second argument call for the library? And who would know unless you troll stackoverflow!?
The explanations are so ambiguous and horribly laid out around this software... worse than any other package I've seen in R. Has anyone here seen the "Shiny" tutorial? Can we try to be 20% as good as that? Especially because it seems that rCharts might turn out to be the most flexible graphics package yet, but the learning curve seems impossible as nearly everything is explained on a case by case basis... which is NOT helpful when you have to troll 80 webpages to find the example that can finally be built upon for the project an individual is currently working on.
When I run 'str(rPlotObject)', the output includes almost 40 attributable objects to the rPlot object, oh "and 32 methods, of which 20 are possibly relevant"... Are you kidding? Can there please be ANYTHING documented that could help parse out the 40+20 properties of this thing? Every single "tutorial" seems to use one type of Plot object, exploits one or two of these 60 attributes, and occasionally has observable output. On this page: http://ramnathv.github.io/posts/rcharts-nvd3/index.html and on this page: http://ramnathv.github.io/rCharts/ , every plot is two or three lines of code with little to ZERO explanation! As stupid as this may sound (probably cause I don't know any javascript), I can't even understand what is meant by a "tooltip". Is it the "mouseover" event? Is it the "hover" event? Is it the "click" event? Is it all three?? Does it only work with cer tain libraries or plot objects? And is it setup differently for each library? Sometimes I see a "name" object that gives the tooltip the content, other times is seems that the popup that shows up upon "clicking" is being called the tooltip; making the use of the "tooltip" term much more ambiguous. As a visual analytics guy, the mouseover interactivity is the most important reason to use a library like this and yet it is the most ambiguous piece of this entire project! It is completely unclear how the javascript content references the R data and vice versa. How much javascript do I need to know to get going with this?
Moreover, the rCharts pdf tutorial found here: https://media.readthedocs.org/pdf/rcharts/latest/rcharts.pdf is more than half empty and has plots that are completely missing or haphazardly cropped. This document is so shoddy in giving explicit reference to the attributes of each type of chart... and who cares anyway when 80% of the cited charts are completely missing?!
My apologies if this is a bit scathing... but like I said, the most versatile R package should NOT be the worst documented.
—
Reply to this email directly or view it on GitHub.
--
Agustin Lobo
alob...@gmail.com