This post discusses functionality in Chart FX for WPF 8.1. For more information click here.
Chart FX for WPF 8.0 includes support for data transforms including OtherTransform and CrosstabTransform. These transforms work with the data as a whole and they will typically affect the shape of the data i.e. they will affect the number of series or points plotted in your chart. Sometimes to analyze the data more effectively you also need minor tweaks on how the data is plotted, to accomplish this we will be supporting a new concept called Series Transforms.
A series transform is a class that affects the values plotted for a single series, a very simple yet useful example would be to plot the percentage of change from the previous value, e.g. let’s plot the US gross domestic product since 1990
Even though this post is all about data analysis, we also want to show our visual capabilities, this look was heavily “inspired” on a chart about a certain golfer yearly earnings I saw some time ago on a news site. Note that all the visual attributes used here (a couple of templates plus some property manipulation) are supported in 8.0 as well, additionally we are using the “Helvetica Neue” font created by Linotype.
If you are analyzing this chart it is hard to draw any conclusions besides the fact that the GDP seems to grow linearly on the 90’s and then dips a little in the early 2000’s before retaking its previous almost linear trend. Let’s add the Rate of Change Transform and see what happens
<cfx:SeriesAttributes.Transform>
<cfxData:RateOfChangeTransform/>
</cfx:SeriesAttributes.Transform>
You can see what a different picture is painted when we plot the percentage change year-over-year, note that in the 90’s the GDP actually go up and down between 4.6 and 6% and the dip in the early 200’s is considerable as it goes down from 6% to 3%. Also note that at the end of the plot there is a clear down trend that was almost impossible to detect on the original chart.
The transform also gets a chance to influence the axis labels (we are manually adjusting the title in our sample) and the tooltips as shown in the next figure.
This transform would also be very useful if you are plotting more than one series when the data range on the series are slightly off, for example if you are plotting sales for 2 products where product A sales range between 1 and 7 millions and product B sales range between 20 and 50 millions, it is very hard to compare growth but the rate of change as a percentage will clearly give you a normalized view of your data.
JuanC