Introduction
The visual OLAP-analysis is an advanced technology that creates graphical OLAP-data slices suitable for visual manipulation of data to reveal "hotspots", hidden interrelation between different kinds of data and form graphical reports.
The screenshot below illustrates two ways of presenting the same data: textual and graphical, so that you can see the difference between them.
 |  |
| Textual presentation of data is more precise, but seriously lacks clearness | The approximation of graphical presentation is easily compensated by its clearness |
Graphical and textual OLAP-analysis is based on the same technology: presenting data in the form of OLAP-slice that can be pivoted (pivoting is rotation of the Cube), drilled, filtered, sorted or grouped.
In addition to that, the graphical OLAP has its unique technologies, such as:
- Visual analysis and comparison of measures, using an adequate presentation of data;
- Revealing interconnections between measures in consideration of influential factors;
- Revealing clusters.
 | Visual comparison of two measures - Order Count and Order Quantity – in view of product categories by financial years. |
 | Revealing correlation between Tax Amount and Gross Profit measures in view of the products divided into sales channels. As you can see, there is a clear correlation between these two measures for the products sold through the Internet, but there is none for those sold by resellers |
 | Selection of the products that bring the most significant gross profit. Later we can apply a filter that will leave out only the elements of the selected area, or view and, if necessary, copy the information about the selected objects |
Now, let’s discuss these visual analysis technologies in detail.
Visual analysis and comparison of measures
This is the easiest way of revealing hotspots in the analyzed data. With the correct way of graphic representation, an experienced analyst will evaluate the situation at a glance and make the necessary decision.
 | Diagrams of product categories sales, summed up by half-years. Moving the mouse cursor over the point in the diagram, you can get the detailed information about it. |
 | The summed up sales volume by regions. The tree most significant regions are highlighted. |
 | Density of distribution of the sales’ summed results by region. |
Using all the facilities of the graphic OLAP-analysis implies that you understand the methods of processing data OLAP-slices. We’ll discuss them below.
Revealing the interconnection between the measures in view of the inflectional factors
Placing different measure to the X and the Y axes, lets you detect the correlation between the measures even with different detaining conditions. For such analysis of data, place one of the measures into the Rows area, and the second – into Columns. The detailing hierarchies are situated in the Color, Shape and Details areas. For example:
 | This chart presents a clear correlation between the Sales Amount and the Tax Amount measures in view of products (the Details axis), where the Tax Amount is directly proportional to the Sales Amount. |
Analysis may reveal other correlation types:
 |  |  |
| A fuzzy direct proportional relation | A fuzzy inversely proportional relation | No correlation |
Revealing Clusters
This type of analysis reveals groups of the detailed positions and their influence on the measure values. For example, for the Internet Gross Profit and the Internet Average Unit Price measures in view of products, we get the following chart:
As we can see, the points are grouped in clusters, within which they have approximately the same corresponding axis values. Select one of the groups and watch the information about the selected points (click the chart, after the area is selected, click the right button of the mouse and choose "Show Underlying data"):
As we see, all the selected points have the same value in the "Product Category" and "Product Subcategory". Let’s suppose all the other products are grouped according to the same principle. To confirm our idea, let’s place the "Product Subcategory" to the color modifier and the "Product Category" – to the shape modifier.
Thus, our idea was correct: the clusters consist of the products of the "Bikes" category (circles on the chart), and the most dense once belong to the same subcategory (circles of the same color).
Using modifiers for more efficient visual OLAP reports
If you need to display the values of several meausres on the same Cube axes, then using series or even a few charts is not your only option, neither, in most cases it is in fact effective.
If among the displayed measures we can point out the main measure, more important than the secondary ones, whose values are not crucial, then it makes sence to use color or size modifiers for displaying those secondary measures:
This chart shows the changes of the “Sales” measure with time. The color of lines and points of the chart depends on the “Order Count” measure values and changes from red to green. The same as in the previous chart. The difference is: the value of the “Order Count” measure, placed to the size modifier as well, influences the size of the point.
Not only measures can be placed to modifiers area, but hierarchies as well. In this case, each hierarchy member is represented on the chart by its own series.
