Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: How to graph this.
From
Nick Cox <[email protected]>
To
"[email protected]" <[email protected]>
Subject
Re: st: How to graph this.
Date
Wed, 5 Feb 2014 14:18:48 +0000
The thread on Stack Overflow
http://stats.stackexchange.com/questions/69263/is-there-a-name-for-this-chart-sort-of-a-cross-between-a-pie-chart-and-a-mekko
raises similar issues. The question was about a treemap; a keen Stata
user showed a horizontal bar chart for the same data.
The thread has pictures, links, several good points, etc.
Nick
[email protected]
On 3 February 2014 18:54, David Hoaglin <[email protected]> wrote:
> Dear Amadou,
>
> A fair amount of literature in statistical graphics supports the
> position that stacked bars are not an effective form of display. The
> display on that MIT site is cute, but it also is not effective, for
> similar reasons. It's clever that, by hovering the cursor over each
> of the unlabeled little rectangles, one can get its name and size, but
> those do not allow useful graphical comparisons among all the little
> pieces (with their labels in view). Beyond a general comparison among
> the larger areas shown with different colors, the main message of that
> display is more like "See how clever we are." For a more-effective
> set of displays, one could start with the large pieces, as separate
> (undivided) bars and then use further displays of the same type to
> show the individual components in each of those large pieces.
>
> Your work on African products may be part analysis and part display.
> The analysis should get at the main regularities and patterns, and
> then the display can emphasize those (and important departures from
> them).
>
> David Hoaglin
>
> On Mon, Feb 3, 2014 at 12:47 PM, Amadou DIALLO <[email protected]> wrote:
>> Dear Nick,
>> Indeed I've seen some java code for tree maps on google. But
>> translating this into stata/mata is way beyond my limited skills. I
>> think this is something for StataCorp or you, if you plan to develop
>> such tool (:-)).
>> Nonetheless, this thread has directed me to other important tools and
>> I want to thank the group for its help.
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/faqs/resources/statalist-faq/
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/