We usually present results graphically, and XY plots are by far the most common format. Prime's XY plots really are garbage, so bad that they make our best work look amateurish. We users are professionals and this nonsense damages our reputations.
What could PTC have been thinking, to release something like this as a product?
Before I detail the horror, I have to acknowledge that PTC did make a couple of improvements: another line type (-...-...) and the ability to have more than two marker lines on the axes. I was glad to see them. Welcome as they were, though, they were far outweighed by the bad decisions.
Some listed below. Anyone care to add to the list?
* No boxed axes, no grid lines. The point of a graph is to display quantitative relationships within data. Prime XY plots don't support quantitative display because readers can't easily read values from the traces without grid lines or at least boxed, ticked axes on both sides. FIX IT!
* No legends for the traces! This is weird, since it is fundamental to multitrace plots. No other plot package has *ever* omitted legends. What Prime calls a "legend" is just the axis expressions for the traces. Suppose I want to plot columns 1, 2 and 3 of matrix H against column 0. The "legend" will be (on the y axis) H_{i,1}, H_{i,2}, etc and not (for example) "annual GDP of USA", "S&P 500 index" and "US trade balance." FIX IT!
* No legend for the axes or title for the graph. Prime tells us unhelpfully to do it ourselves with text boxes. Not only is this a nuisance, it doesn't rotate the text 90 degrees for the Y axis.
* No secondary Y axis. We use this feature commonly when different traces, or groups of traces, have widely disparate ranges. Prime, again unhelpfully, tells us to scale and shift the data of the traces so they all fit on a common Y axis range. This just passes the garbage on to the end reader, who now has to do mental math or run extra calculations just to interpret the graph! (By the way, does anyone know what "document-centric" means?). FIX IT!
* Densely spaced symbols. We use symbols (e.g., +, *, o) in addition to line types to distinguish one trace from another. But Prime puts a symbol on every point - and, for many data sets, it turns a trace into a blur, hiding the line type and the data values. Mathcad solved it by letting us specify the number of points between symbols. FIX IT!
* Marker lines on the Y axis (i.e., horizontal lines) put the callout box containing its value on the axis end, instead of the free end. That simply obscures the box.
* One marker line may show up darker than the others. That makes the end reader ask, "Why? What did the author mean?"
* Mathcad let us determine the axis span and number of divisions (i.e., where to put the tick marks) by typing the values in a dialog box. Prime does the equivalent by letting us type in, on the graph itself, the numerical values of first, second and last tick mark. It's clever enough - but physically harder to do because the numbers are small, when imaged on the screen, and often close together. It ends up being clumsier than Mathcad.
* The traces have muted colours. But we want bright colours on plots, so that the traces stand out and are easily distinguished from each other. After all, those traces are the results of all our preceding work on calculations.
I just reviewed Forum postings from 2010 and 2011 on plots, and some of the problems above existed at that time and remain unfixed, four years later. I don't have much hope for future versions of Prime.