Difference between revisions of "Dynamic Charts"
(→UI) |
(→UI) |
||
Line 13: | Line 13: | ||
==UI== | ==UI== | ||
As mentioned previously it can be difficult to understand outliers, trends etc when the data is presented as just numbers and text as can be seen below. Using Small Multiples with this example below will guide you through creating meaningful information from the data. | As mentioned previously it can be difficult to understand outliers, trends etc when the data is presented as just numbers and text as can be seen below. Using Small Multiples with this example below will guide you through creating meaningful information from the data. | ||
+ | |||
'''Task Pane – building a set of Charts Example Basic''' | '''Task Pane – building a set of Charts Example Basic''' | ||
+ | |||
[[Image:sm1.png|400px|centre]] | [[Image:sm1.png|400px|centre]] | ||
Revision as of 11:33, 16 August 2011
Introduction
Small multiple charts is a term popularised by Edward Tufte. They allow comparing data series as they are repeated for different members.
In XLCubed they are useful for visualising large amounts of data, as hundreds of data points can be plotted and compared easily.
This allows users to find trends and outliers in the data more easily than by trying to study a large table of data.
Video demonstration
For a tutorial, try watching our Small Multiple Videos
UI
As mentioned previously it can be difficult to understand outliers, trends etc when the data is presented as just numbers and text as can be seen below. Using Small Multiples with this example below will guide you through creating meaningful information from the data.
Task Pane – building a set of Charts Example Basic
As you can see the desired end point is a much easier way of viewing the data
Let's work through an example.
Category – defines the Category (X Axis) within each individual Chart, in this case a Trend over Time
The Series defines the number of data series to be charted.
In order to be able to visualise a simple trend this would be easier with Lines rather than Columns. Use the Chart Type on the Options panel of the Task Pane to change the chart style.
By dragging the Region hierarchy onto columns, and then selecting Children of All, we can split data into separate charts at Region level.
We can additionally split this by Channel by dragging the Channel hierarchy onto rows,
Now we have a graph for each Region by Sales Channel. The Scales (Y axis) are shared so that the charts can be easily compared.
More detail is available for each chart by double clicking on the individual chart eg Direct Sales / Australia Pacific, where the data is unclear because of low numbers can be further explored like this. The scale is automatically adjusted to give a clearer picture of the data. Use the back button to return to the higher level.
Manipulating the chart to show Each Product by Channel, simply drag the Products from the Series to the Rows Below you can clearly see the effect of swapping the Product (Rows) and the Channel (Series)
Additionally you can drive the data selection by using Slicers or Excel ranges.
Here we have defined a Slicer on Time at Year level and used it to drive the Categories.
Small Multiple Chart Designer
Insert a Small Multiples chart by using the Small Multiples option on the ribbon, or the Insert -> Small Multiples Menu option.
After inserting a Small Multiples Chart the above windows will be displayed. The Small Multiples Charts - Task Pane is automatically displayed on the right of the sheet. The Task Pane can also be accessed from the right mouse button pop-up menu.
To delete the Small Multiple chart use the right mouse button pop-up menu -> Delete option.
To Move, Size or Close the Task Pane, drag it using the window bar or use the drop down menu on the Task Pane Header.
The data within the Charts can be refreshed (like a grid) by using the right mouse button pop-up menu -> Refresh option
There is also an initial Warning displayed on the empty chart as it is essential for the design of a Small Multiple Chart set that a Category is selected. This warning can also be seen on the Task Pane.
Using some of the other features in the Options section of the Task Panel, charts can be made to display data in order to be able to better view correlations and or outliers
When selecting the X Axis ‘Children of the Geography’ we receive the following message, because there more columns returned than the ‘Max Number of Columns’ defined on the Properties for the Chart. i.e there are 6 values, for ‘Children of Geography’.
Alter the Max Charts on Columns to ‘6’ to display all the Geography – Children.
Note the Warning indicator icon in the bottom left corner of the Chart has now disappeared as all the data is displayed.
As there is nothing selected on the Rows, a better way to display this information, given its long narrow nature would be to turn on Chart wrapping.
The data still has a common scale and is more easily compared.
Using Right Mouse Menu select the position for the Legend, Right, Bottom or None.
If None is selected the Legend can easily be retrieved by using the Right Mouse button anywhere on the Chart.
Using the Legend series on the Charts can be highlighted, this is useful where series overlap. Here the Reseller Order Count appears brighter. Multiple series can be selected by using Cntl Click.
Using the Right Mouse Menu ‘Drill Into’ the data for Canada, you can continue drilling as far as the Hierarchy allows.
Using the Back button returns to the previous level of Charts
Click on the chart you wish to keep and use the Keep Only menu option to remove the other charts
Using this option you can get a closer comparison between two or more Geographies. Here we have removed the United States and Canada, in order to get a better look at how the other Geographies compare to each other. Notice the scales have been adjusted to reflect only the data displayed.
Colour Coding can be applied to the data points in order to highlight differences. Colour Coding can be applied using the plotted Measure, or another selected Measure. Tick the Colour box in the Options area of the Task Pane; this makes the colour gradient active.
A warning will appear that ‘No Values Hierarchy chosen’ next to the Values box in the Options area.
Drag the hierachy to plot, normally the ‘Measures’ to the Values box and select the Measure to be plotted ( Y Axis)
Next select the value to apply the gradient to. This can either be the same Measure that is being charted or another measure altogether.
Above the ‘Reseller Sales Amount is plotted and Chart coloured based on the Reseller Order Quantity.
Above the ‘Reseller Sales Amount is plotted and Chart coloured based on the Discount Amount
The colours used for charting can be changed by using the button to the right of the colour swatch.
Here by looking at the Sales Amount, coloured by different values we can see that although there were large volumes of Sales in the USA they were also the most heavily discounted.
It would be useful to view the Order Numbers vs. the Reseller Sales Amounts; if they are both plotted on the same scale it is almost impossible to see the Order Count values as they are so small by comparison. ‘Move to other axis’ allows you to see two Y Axis scales, one each for Amount and Count.
Using the Right Mouse Menu, whilst positioned on the Legend Entry for the series you wish to move Click ‘Move to other Axis’
Scatter Charts – plotting a correlation
Plotting a correlation, using a scatter chart option - firstly define the Y axis and then the X axis against which it is to be plotted. Here we can see how the Sales Amount and Order Count are related - closely in Australia, but spread out in the United States.