515391_en-US

liliconcom

Graph, Plot, Map, Chart and Gram

Hi,
I recently got a similar question from two customers
Combining the two people's questions is about the differences between Graph, Plot, Map, Chart, and Gram.
I searched several places including JMP Community but don't understand enough the difference between the above five.
Anyway, if list-up the names inside JMP, it is roughly as follows.
Thanks
Ikju

-Graph : Graph builder
-Plot : Bubble, Scatter, Parallel, Cell, Contour, Ternary, Surface, Overlay, Mosaic, Box
-Map : Heatmap, Tree map
-Chart : Pie, Bar, Variability, Control
-Gram : Histogram, Diagram

Re: Graph, Plot, Map, Chart and Gram

I think the short answer is that a lot of it is historical and there isn't really a rhyme or reason to why one term is used over another other.  There's also a lot of interchangeability in the terms.  You'd probably have to look at when each one entered into the literature to understand why they are named as they are.

The histogram is a great example.  The version we use today is a bar chart (you could call it a bar graph or bar plot, too), but it has its own name because Pearson assigned special meaning to how it was constructed (bins and counts). There are also different flavors of histograms that are completely different types of visualizations (line charts, smoothed curves, dot plots, for example). 

 

Re: Graph, Plot, Map, Chart and Gram

There are several views on the difference, but the semantic difference does not seem to be MECE.
(Mutually Exclusive and Collective Exhaustive)
Like Mike's opinion, I'd like to understand historical or practical habit, original namer's just naming or the meaning s(he) gave.

Re: Graph, Plot, Map, Chart and Gram

To me the most clarifying thing is a series of calculations in JMP, on a simple example, either in column formula(s) or in JSL, to show what JMP is doing.  The names are often misleading in statistics for me personally.  The more we can show through simple and powerful examples, "what a histogram is," "what a normality test is", "what a CDF or PDF is," the more empowered our users are going to be.  Remember that they have to defend what they are presenting, often to senior management in PowerPoint slides with sufficient explanation.  To be able to adequately explain and defend an analysis I personally feel that you have to have some deeper understanding of "how to build it."  This is kind of the paradigm of a good scientist or engineer.