Displaying a series of numbers or data graphically as a chart instead of as a list enables users to interpret business data much more quickly and easily. That is the primary purpose of using charts in your Web reports, of course. However, the initial chart that you produce typically isn’t in the best possible format to help business users interpret the data, so you often will need to edit the chart to meet your requirements.
NOTE: If a migration process is required to convert charts from the BW 2.x or 3.x version to the BI 7.0, you will need to upgrade your IGS from the 6.40 version to the 7.0 version. In the previous version of BW, the required Internet Graphics Server to support Web-based graphs was the 6.4 IGS and as we mentioned along with all of the application upgrades for the 7.0 version also comes platform upgrades and this is a significant enhancement. IGS 7.0 is the required environment to use the WAD for BI 7.0.
In the case of the 7.0 BI you can edit a chart via the Web Application Designer or the Report Designer but in either case you have to access the WAD to actually configure the chart type. The following illustration identifies the various elements of a chart. You will encounter these terms as you read about the process of configuring a chart. The information shown on a chart is broken out into four different areas: data information, axis information, chart titles, and the formatted area behind the chart. The data information is referenced by the Data Point and Data Series, the axis information is referenced by the Value Axis (Y), Category Axis (X), and Axis Title, the chart titles are referenced by Chart Title, Legend Text, and Legend Icon, and the formatted area behind the chart is referenced by Drawing Area, Background, and Gridlines. In this section, you will learn how to configure and adjust each of these areas to produce a robust chart for the business user.
To start the process of editing the chart in the WAD, either create a new Web item of the type Chart or select an existing Web item of the type Chart. Then, on the Web Item Parameters tab of the Properties screen area, under Internal Display, click the button for the Edit Chart property to access the Edit Chart dialog box (or Chart Designer), shown in the following illustration.
Alternatively, you can open the Chart Designer by right-clicking the Chart Web item and choosing Edit.
The nice thing about the Chart Designer is that it is set up similarly to other wizards in the SAP system. It offers you a step-by-step process of moving through the configuration of the chart from start to finish. You can also go back, after previewing your chart, and change information and settings until you are happy with the results.
The first step in the process of publishing a chart is to specify which chart type you want to display. You can determine how the data provider is to be built by using the class of the chart type. Now a task that we didn’t really talk about for this process and it’s something that should already be completed and that is creating a query or selecting an existing query that you can use to generate a suitable view of the data. Once this has been identified in the Data Provider section, then you can move through the wizard steps one at a time and validate or change any parameters you need to.
NOTE: You probably have noticed that this is very similar to the Chart Designer in Excel. If you are comfortable with all the functionality of the Chart Designer in Excel, you will have very few issues with this process, perhaps with the exception of figuring out where a particular parameter is located.
As you can see we have exactly six steps to work through to develop a chart with a display of data. If at any time you feel that the setup that you have completed is good enough, then all you need is to skip to the finish. At this point you can choose to refine the format. Once in this particular view of the graph the only other option you have is to finish the chart or, if you require, use the Wizard button to go back to the actual six steps. The following illustration shows these options.
If you choose the Refine button, you move to detailed display level. Here you can configure all of the parameters by clicking on the header in the overview or clicking a particular portion of the chart. We will be working through all of these options later in this chapter. At this point just realizing that you can use this approach to configure rather than the wizard is all we want to reinforce. This information is found in the illustration.
After you click the Refine button, as the preceding illustration shows, your only option to move from this screen is Wizard or Complete. The option to use the refine approach is nice to have and much of the configuration can be accomplished from this but in most cases you have to return to the full Chart Designer to finish your work. One of the lesser-used parameters is the SWAP DISPLAY AXES, which enables you to swap the data providers in the chart. So, if you want to swap automatically the data providers used for the rows and columns, you can use this setting to accomplish the task.
When you edit the chart, the Web Application Designer or Web Application Wizard first shows the default setting, that is, a column chart. Initially, format the chart using the wizard so that the most important settings and the number of displayed data series and data points correspond to your data provider. Then assign the required chart type to the chart and format the chart as required. All chart elements that you can format are displayed in the list of properties on the right of the screen. You can modify the properties as required here based on the list supplied.
Structure of the Chart Designer
Working with the Chart Designer
When you edit a chart, you use the Chart Designer to view and edit the preview of the chart. The structure of the Chart Designer window has three areas: chart preview (left side of the screen), list of elements (top-right side of the screen), and property area (bottom-left side of screen), as identified in the Figure. Depending on the Chart Designer feature you are using, these sections can switch sides of the window. If you are going through the wizard process, the chart preview appears on the right side of the window; if you are using the Refine mode, the chart preview appears on the left. The following list describes the three areas:
The lower area of the Chart Designer contains buttons that you can use to save changes and finish editing the chart (Complete), finish editing the chart without saving (Cancel), and call up and navigate in the wizard (Wizard). You can navigate and switch your view in the Chart Designer via several different approaches. My favorite is to just click each of the options in the list of elements and see the different, more-detailed parameters appear in the lower property area of the screen. For example, in the following illustration, Background is selected in the list of elements and its corresponding properties are shown in the property area.
As you can see, the properties include filters and parameters for Color, Texture, Style, and other background elements that can be configured. Also notice that as you choose each of the different elements and its properties are displayed, the affected area is also highlighted in the Chart Designer chart preview. An example of selecting the Columns element is shown in the following illustration.
You can modify the size of areas in the Chart Designer according to your requirements. For example, you may want to increase the size of the property area so that you can display all properties at once. To do so, position the cursor at the intersection of two areas until the cursor changes into two parallel lines. Click and pull the area to the required size. When you have finished editing a chart, or want to terminate editing, close the Chart Designer and return to the WAD. Before you close the Chart Designer, make sure you save your work before exiting, if you want to save it.
Before we start to work our way through the six different steps to develop a chart, let’s look at the chart types so that you are familiar with your choices. The chart type defines how your data is displayed graphically. The chart types can be divided into five classes with respect to processing and complexity. With chart types of the same class, the rows and columns of the underlying table (that is, of the data provider) are processed immediately.
The class to which each chart type belongs is listed in Table. Before you configure or publish a chart, you need to specify which chart type you want to display. You can determine how the data provider is to be built by using the class of the chart type. You can then select an existing query or formulate a new query with which the appropriate view of the data can be created.
Chart Types and Corresponding Classes
I learned the hard way that failing to follow this procedure—that is, specifying the chart type to display before publishing the chart—wastes a lot of time. Instead of identifying the format of the chart first and then building the query to the appropriate view, I would create the query first and then force the chart type to fit my query structure. I do not recommend this approach.
As you can see, the charts that you normally work with are in classes 1 and 2primarily (but I have run into the waterfall chart type in a number of projects). As long as you are comfortable working with class 1 and 2 chart types, you should be able to support over 85 percent of your customers’ needs.
Most of the charts listed in above Table can be displayed with various dimensions, such as 2 dimensions, 2.5 dimensions (limited three-dimensionality), and 3D (regular threedimensionality). Bar and column charts can be converted into pyramid, cone, or cylinder diagrams in 2.5D and 3D mode. The dimensional look and feel of the charts is a good “nice to have” but, depending on the amount of information and data points on the chart, the 3D format may be more distracting than useful. I recommend using the 3D format only if the customer requests it.
Another variant that exists for line, column, bar, profile, and radar charts is a “stacked” variant. These values of the data series are added and displayed on top of one another in a category and normally are differentiated by colors. Stacked charts display the relationship between individual elements and the total of all values.
You can also set 100% variants for line, profile, column, bar, area, and profile charts. The sum of all data series in a category is 100 percent. The values of individual data series are converted into percentage values and displayed accordingly.
a.Class 1 Chart Types
As listed in table, class 1 includes the following chart types: line, profile, column, bar, doughnut, radar, area, profile area, pie, polar, speedometers, and split pie. You build the underlying table of the chart types of class 1 (with the exception of pie charts and speedometers) as follows:
The difference when using a pie chart is that the underlying table has only one row. The values in the columns form the pie segments. If the table contains more than one row, the additional rows are ignored when the table is converted into a chart. However, you can define which row is to be used. When using a speedometer, the difference is that the underlying table has only one data column. Each value in the data column is displayed in the chart as a pointer. The speedometer will be readable if the data column does not contain too many values. If the table contains more than one data column, the additional columns are ignored when the table is converted into a chart. You can define which data column is to be used for the speedometer.
After reviewing these nuisances to building the queries to use for the specific chart types, you should start to get a good idea of what the formatting process for your queries needs to be for each of the different chart types. The following illustration shows the data table that is the basis for line, profile, column, bar, doughnut, and radar charts.
NOTE: You could also generate a split pie chart from this data source, but we will work with another example in order to clarify the possibilities for this chart type.
Most of the examples mentioned here were createdusing the chart attribute Switch Axes to Display(SWITCHMATRIX='X').
Line Chart: Data trends are shown in a line chart. The data is entered at regular intervals. Categories such as items groups are groups of similar characteristic values such as material types, product types, customer attributes, etc.and are normally entered on the X axis and values such as revenue on the Y axis. This can be seen in the illustration shown here.
Completing this initial basic line chart simply required adding some minor text to the chart. Most of the time was spent creating the query to support this chart. With the appropriate data available, a reasonable chart can be created almost immediately. To create this chart, I adjusted the default chart type from column to line by right-clicking on the Chart Web item and choosing Edit from the context menu and I then chose basic lines rather than the dimensional options, as shown in the following illustration.
In step 2 of the Chart Designer, I added text for the titles of the chart and the axes, as shown next. As you can see, when you enter text in the fields on the left, it appears in the chart by default. At this point, if you click the Complete button, you have just built a line chart.
For line charts, you can set the manner in which the lines pass between data points:direct, as in the preceding example; as curved lines; or in varying increments. To do so, go to the next step in the Chart Wizard, choose Data Series Format | <name of series> | Line Type. In all the other examples in this section, we will only add some text to the formatting process, so they all can be developed in a matter of seconds.
Profile Chart: In a profile chart, the lines are arranged vertically, as shown here, and not horizontally as in a line chart. Otherwise, the profile chart corresponds to the line chart. Basically, the axes are switched for a different view of the data. Notice that the text also automatically adjusts to the appropriate axis.
Profile charts have the same options as line charts for setting the manner in which the lines pass between data points. Rather than a direct line that passes between the data points, as shown in the previous example, you can also have the lines curve by different levels. Again, you choose Data Series Format | <name of series> | Line Type from the next step in the Chart Wizard. Another option to allow this same formatting option is that you can find these parameters if you go to the Refine screen and choose the Data Series Format. This can be seen in the following illustration.
And the results of this are in this illustration. As in this situation, not the best display but there are many other options within that area. Before you attempt zigzag lines, I recommend experimenting with the dimensional profiles and lines first to see if this helps produce a dynamic chart.
Column Chart: The column chart seems to be one of the most popular chart types, probably because it can be easily understood and assimilated within a very short period of time. Depending on your audience, you can make a column chart very robust or very basic. As its name suggests, in a column chart, the comparisons between individual elements are shown in a column. Categories are arranged horizontally and values are arranged vertically. To show changes within a certain time interval, you can use either column charts or XY scatter charts (described in the “Class 2 Chart Types” section). The following illustration shows the numerous different column options that are available.
An example of the look and feel of the Stacked Columns option is shown in the following illustration. You can see that this would also be a very direct chart to offer for analysis.
The result of choosing the basic Columns option is shown here.In a 3D column chart, you can also depict the columns as cylinders, cones, or pyramids, which can really be a great visual enhancement. In a 2D column chart, you can depict the columns as triangles (2D pyramids). To do this, choose Columns | BlockStyle. If you are using a 2D column chart, there is no difference between the setting Pyramid and the setting Cone.
Bar Chart: The next most popular chart type might be the bar chart, in which comparisons between individual elements are shown in a bar chart. Categories are arranged vertically and values are arranged horizontally. The emphasis is on the comparison of values and not on displaying a change during a period of time. An example bar chart is shown in the illustration.
In cases such as this example, you will need to make some additional changes to the chart. Here, the data points for Sales are probably not sufficiently detailed. Therefore, going into the Chart Type and identifying the types of values would be required. For example, in this case we have a comparison between Yearly Sales of 2007 and 2008 and we would need to identify the types of values rather than just showing the title of Sales.
As in a 3D column chart, in a 3D bar chart you can depict the bars as cylinders, cones, or pyramids. In a 2D bar chart, you can depict the bars as triangles (2D pyramids). To do this, choose Bars |BlockStyle.If you are using a 2D bar chart, there is no difference between the Pyramid and Cone settings.
Doughnut Chart: In a doughnut chart, the relationship between parts of a whole are displayed in a doughnut. This is similar to a pie chart, discussed a bit later in this section. In contrast to the pie chart, however, the doughnut can represent more than one data series,where each ring corresponds to a data series, as shown here. I do not favor this chart type, but in the correct situation it can work well.
You can change the width of the rings by setting the size of the hole. To do so, choose Doughnut | Hole Size | <value as percentage of ring size>.
Radar Chart: Another chart that is a bit conceptual in nature is the radar chart, shown next. In this case, each category has its own value axis emanating from the middle. The values of a data series are linked with lines. Radar charts can be used to compare data series: The data series with the highest values occupies the most space. The radar chart is a bit more direct than the doughnut chart, but still I can’t think of many situations in which to use it. There are definitely more types of charts that are easier to explain and understand.
You can also depict the chart areas filled in. Choose Radar | Filled to do so. Depending on the size of the individual data series, the areas may overlap.
For the following series of chart types, understanding what types of data show up in specific chart types becomes very important. For this series of chart types, we will be using a different data set. The following illustration is the basis for these charts. As you can see, the data provider has been changed to only one key figure for Material Group (note that in the initial query this is listed as the Product Group information and we have adjusted for reporting purposes to show as Material Group).
Area Chart :The area chart can be used in a number of situations. It’s very easy to read and interpret and has a similar look and feel as a bar or line chart that is filled in. The area between the axes and the data series is filled in an area chart. A stacked area chart depicts the sum of the applied values, thereby illustrating the relationships of parts to each other. An example is shown in the following illustration. Just to mix things up a bit, I used the 2.5D view of the area chart. In a situation where you had two sets of data (two key figures), you would see this chart as having another area behind the first area in a different color. This can get difficult to read due to the fact that some of the information may be hidden behind the initial key figure area.
You can determine the line type in an area chart in the same way as for a line chart. The only difference is that the line type Curve is not supported for an area chart. I would definitely not highlight that as a big deal since the direct line type is the most popular.
Profile Area Chart: In a profile area chart, the areas are arranged vertically and not horizontally as in an area chart. Otherwise, the profile chart corresponds to the area chart. As the following illustration shows, this is not the easiest type of chart to read.
You can determine the line type in a profile area chart in the same way as for a profile chart.
Pie Chart: The pie chart is definitely one that you can use frequently—it’s basic, straightforward, and can have a dramatic effect if formatted correctly. In a pie chart, the proportional part of the elements of a data series are displayed in a whole. This chart type has only a single data series and is used primarily to highlight a particularly important element. As the following illustration shows, adding some depth to this chart can make it much more interesting. I also added another option, Explosion Offset, which separates the pie pieces. You can find this option if you use Refine and choose PIE properties in the overview section, and then increase the Explosion Offset in the properties section and separate the pieces to whatever spacing looks good to you. The preview will show you the adjustment dynamically as you change the settings. Note in the illustration that you can scroll across the screen and see the details for each of the pie pieces.
A pie chart only reads a single data series from a data source and ignores all others. You can determine which series is to be used by choosing Pie | Series Index | <number oftable series>. Another parameter you can easily set is how much of the drawing area your pie chart should take up. To do so, choose Pie | UsedSpace | <value as percentage of drawing area>.
Polar Chart: The polar chart is another specially formatted chart type that can be used in certain situations, but it doesn’t really offer any advantage over and above the other, more readable chart types. In any case, it’s an option and each category has its own value axis. The values of a data series are depicted as areas, as shown in the following illustration. As you can see, this chart type would have to be explained carefully to make sure that the business user is correctly interpreting the information. Looking at the results of this chart type, you would be hard-pressed to come up with the information from the report.
A polar chart only reads a single data series from a data source and ignores all others. You can determine which series is to be used by choosing Polar | Series Index | <number of series>.
Speedometer Chart: Another one of my favorites is the speedometer chart type. A speedometer displays one or more key figures in the form of a pointer. The speedometer is divided into several value ranges, and the user immediately sees the value range in which the pointer is currently positioned.
A speedometer only displays one data column of the data provider. This is the basic speedometer with no color groups, but it is dramatic enough to catch your interest. In the case of the speedometer chart, I typically turn off the parameter to Swap Display Axis since a swap of pointers with other values in this situation would not show correctly or make any sense. The Swap Display Axes setting in the Web Item Parameters tab of the WAD is shown here.
The results of these parameter and settings changes are shown in the following illustration.
You can also determine which data column is to be displayed by choosing Speedometer| Data Index | <number of data column>. In terms of the display of the pointers, you can determine whether to display the categories as pointers rather than the data column by choosing Speedometer | Use Categories. You can also define the look and feel of the pointers—whether you want the arrow displayed or something else to take its place by using Speedometer | Show Arrows. Some of the other options in a speedometer are to add some color formatting, font changes, and background colors. In some cases with the different properties that you can use within the chart types some don’t really work such as with speedometer chart types. You can change the dimensionality of the chart type but with the speedometer chart type nothing will happen since 2.5D and 3D are not possible. These settings are found in the Global Settings under Refine. An example of the color formatting is illustrated next.
Split Pie Chart: The split pie chart is another favorite of mine. In a split pie chart, several data series are displayed per category as a pie segment, and are depicted in proportion to each other. For this example, we will go back to the data provider we used previously for the column chart type and have two key figures by year and see what this offers us. I like the display of the different sections, and if we add some dimensionality to this, it immediately looks close to being a finished chart. The next illustration shows an example of the end result.
One of the other options of all the different free-flowing chart types, such as the pie and polar charts, is that they can be moved or rotated into different views. This is very useful if you are going to work with 2.5D or 3D shapes.
At this point, you likely are starting to get a feel for what the initial dashboard process would look like. If we were to incorporate several of these charts together into one dashboard, we could start blueprinting what we might want to present to the customer. I’ve taken some of these chart types and used a simple table to hold their format a bit more. This takes about 15–20 minutes and lays the foundation for a dashboard. An example of the initial configuration and the result is shown in the next illustrations. Although this example is limited to six charts, it is sufficient to help you understand the overall dashboard setup and architecture.
b.Class 2 Chart Types
Class 2 chart types include scatter charts, time scatter charts, histograms, and heatmaps. They are not necessarily more complex than the class 1 chart types but are more specific to certain situations than the class 1 chart types. Thus, in the case of class 2 chart types, we have a more defined set of data to use for building these charts. In this case, you build the table on which a chart type of class 2 is based as follows:
The X value of a data point is always from the first data row. The Y value of a data point is from one of the remaining data rows, depending on the data series to which the data point belongs. The data providers for histograms and heatmaps need a different structure; the requirements for these are described in the corresponding sections for these chart types.
Scatter Chart: In a scatter chart, either the relationship between numeric values is displayed in several data series or two groups of numbers are entered as a row of XY coordinates. This chart type displays irregular intervals (clusters) and is normally used for scientific data so using these types of charts for business purposes requires a unique situation. Both axes of a scatter chart are value axes, so we would not necessarily use the scatter chart type to display something with a characteristic like Product Groups down the rows since this would not look correct once displayed. In other chart types, the X axis is used to display categories or groups of characteristics. The data that is to be used for those chart types is organized by month and key figures. The information used to build a scatter chart type is shown in the following illustration.
The resulting chart is more of a flexible line chart type. Since both axes are values, it’s important to be aware of the appropriate scaling factor for the particular values. The final result is shown in the next illustration.
You can fill the areas between two points of a data series, almost as if the chart were an area chart. To do so, choose Scatter | Filled.
Time Scatter Chart: A subset of the scatter chart is the time scatter chart. The X value can be a date or time. The basic Chart Designer configuration screen for this chart type, shown next, gives you a hint that it is not meant to be used to show information by products or customers; rather, it is used for tracking and data analysis in a more value-driven approach. Notice the formatted Date approach to the X axis.
As you can see, the axes are both value driven and not category driven. Therefore, the data needs to be similar to the data used in the scatter chart type; the data provider for this example is shown in the following illustration.
The results of applying the time scatter chart type are shown in the following illustration. The X axis is the time in hours and the Y axis is the number of batches processed.
These types of charts are great for displays with date or time values. You can set up to three different time axes; for example, one for years, one for quarters, and one for months. To do this, choose Time Axis | Line | Line Type1 to Line Type3. You can also use Line Format1 to Line Format3 to specify the format in which the time values are to be displayed. You can use the properties Line Step1 to Line Step3 to set the intervals between time units. The following abbreviations are used for the time specifications:
D = day, Mon = month, Y = year, W = week, Q = quarter, H = hour, Min = minute, S = second
As in the former scatter chart type, you can fill the area beneath a data series by choosing Time Scatter | Filled. This might offer a more dramatic display of the information.
Histogram Chart: The histogram is used primarily for the graphic display of a series of activities. The frequency of a characteristic is displayed in a histogram (for example, the sales revenue for a product group). The frequencies are divided into classes, where each class corresponds to a column in the histogram. In a histogram, categories (classes) are entered on the X axis and the corresponding values are entered on the Y axis. The information for the histogram chart type is shown in the following illustration.
In this case, I had to do some adjustments to the X and Y axes to allow the data to be shown consistently across the chart. Even after these changes, understanding the results of this chart type is difficult, as shown next. Typically, you’ll know when you need to use a histogram chart type because the data that you are trying to chart will not fit any of the other chart types and the histogram will be the only one that makes sense.
Analyzing this information based on the data provider, we see that the sequence is on the Y axis and the sales data is on the X axis. The numbers within the chart are the number of sequences that are within that range, so the preceding histogram shows one sequence in the 198,000 to 300,000 range, one sequence in the 300,000 to 400,000 range, two sequences in the 400,000 to 500,000 range, and the remaining eight sequences from 500,000 to over 600,000. This is very similar to the clustering process in data mining, where items that are within a certain range are grouped together. To be displayed correctly, a histogram needs one data source with exactly the structure shown earlier. The first data column contains unique numeric values only; these do not need to be sorted. The second data column contains the values that are sorted into the classes of the histogram. Again, notice we are talking about two data columns and one category column. You can control the number of classes by choosing Histogram | Classes | <required number of classes>.
Heatmap Chart: The last class 2 chart type is the heatmap. This chart type is very unique in nature and with all the requirements needed in terms of a specific set of data we will talk through this one and show a sample of data required as well as the resulting graph. Heatmaps allow you to display large volumes of data compactly in a diagram. As in the histogram, the heatmap chart type groups information together in ranges of values, so this chart type is appropriate if you are not really looking for details but for patterns in the data.
You can display the values of two key figures compactly and independently of each other for a number of data series. The display is two dimensional:
You can identify unusual values and trends easily and answer business questions such as “How do the sales figures in various distribution channels and product groups compare to each other?” An example of the data required for this chart type is shown in the next illustration.
You build the report on which a heatmap is based with a table that must contain the following:
So, this is probably the most restrictive chart type we’ve encountered so far. The required query is very specific and a bit more involved to work with. The basic Chart Designer configuration diagram itself is intense, as shown next, and if your information or your requirements are not specific to the heatmap process, the chart will make very little sense.
A diagram of what is happening in this chart is shown in the next illustration. You can see that you define a driver key figure, Billed Quantity in this example, that serves as the lead value for sizing the rectangles.
As previously mentioned, the two display dimensions are area and color. In the area dimension, the characteristic value Fax (Distribution Channel characteristic) results in the large square to the upper left of the heatmap, since its categories (Product Groups) have the largest Billed Quantity. The three categories are within this rectangle; each is represented as a rectangle proportional to its Billed Quantity. Therefore, your three Product Groups are defined by the three color sections within each of the groups. In the color dimension, the three product groups Bag & Outdoor, Accessories, and Office are differentiated by color based on the Net Sales for each Distribution Channel. For the Distribution Channel, the colors for the Bag & Outdoor and Accessories rectangles are similar, whereas the Office category, on the other hand, is easy to distinguish (compare to data provider). The good thing about this is that when you render the template, you can scroll over each of the rectangles to identify the values and, in this case, the Distribution Channels.
c.Class 3 Chart Type: Portfolio
The only chart type included in this class is the portfolio. Unlike the last chart type discussed, heatmap, the portfolio is a chart type you see all the time, particularly when dealing with financial information. For example, the Wall Street Journal and USA Today often use a portfolio chart to depict the ebb and flow of the different stock groups or industries for that week. These chart types identify groups that are growing and/or declining over the past X time frame. An example of the basic configuration screen for the portfolio is shown in the Chart Designer in the following illustration.
A portfolio displays the position of an object (enterprise, product, and so on) in a fourfield matrix. The position of the object is defined using two dimensions and the X and Y axes depending on the movement of the overall amounts or values. So, the base of zero is moving to correct for some variance that is defined. Portfolios are used mainly in enterprise and product comparisons. A portfolio can be sorted, for example, like the products of an enterprise by their dimensions economics and strategic significance.
The most important aspect of this chart type is the construction of the underlying query to support it. Once you get this set up correctly, the final build of the chart itself is routine. You build the underlying table for a portfolio as follows:
The X value of a data point is always defined from the first data column. The Y value of a data point is defined from the second data column, depending on the data column to which the data point belongs. The X and Y values together give you the center of the bubble. In the following example, the X axis is the Net Sales, Y is the Invoiced Quantity, and the bubble size as the average value—Sales/Quantity.
Once we have identified the structure of the information, application of it to the chart would be similar to the following illustration.
There are additional parameters that you can readjust, such as the form of the markers from a bubble or hexagon, by choosing Data Series Format | Default Data Series | Area Properties | MarkerShape. The field MarkerSize has no function for this chart type. You can change the marker size using the following approaches:
If you are formatting a single portfolio with a fixed size, use automatic calculation of the value range and deactivate the property Portfolio | Size in Percent. If you are formatting a portfolio whose size can change, use automatic calculation of the value range as well as the property Portfolio | Size in Percent. If you want to display multiple portfolios on a Web site with comparable scaling, enter a minimum and maximum value (for all portfolios). Finally, you can use the property Portfolio | Minimum Value to ensure that small markers remain visible.
d.Class 4 Chart Types
Class 4 chart types are useful for tracking and analysis of activities such as projects and processes. We are all familiar with the concepts of the two class 4 chart types, which are the familiar Gantt chart and the milestone trend analysis (MTA). These charts are designed to depict information dealing with dates and times as they relate to the process and progress through a project and either hitting milestones and/or tracking variances from the milestones required. One of the other areas in which I have some experience, the Finance module in SAP, has a fairly new component called the Financial Closing Cockpit. One of its major charting components is the use of a Gantt chart to track the period closing process in finance or controlling. The next illustration displays the initial configuration screen for a Gantt chart in the Chart Designer.
Gantt Chart: Specific to the Gantt chart, you can illustrate the time progression of projects and their substeps. You can also group these substeps in categories. The time is displayed on the X axis and the substeps and categories are displayed on the Y axis. If you created categories, all substeps are displayed separately for each category on the Y axis. You can display the time in days or you can use start and end times:
All combinations of start and end points are supported:
If you select Start Time as the start point, the system uses the current data as the Start Date. If you select Start Date or End Date as the start point or end point, the system fills Start Time and End Time with the value 0:00. This is particularly important for the interpretation of end points in the Gantt chart.
As with class 3 chart types, the construction of the data provider is the most important part of the process. You build the table on which a Gantt chart is based as follows:
An example of this type of query is displayed in the following illustration.
The example illustrates a table for a Gantt chart; it contains three categories and no data for some start and end times. The substeps are the Project Steps: Chart, Object Services, and Web Runtime. The categories are the Work Packages: Concept Creation, Specification, and Solution Validation. Once this information is applied to the Gantt chart type, the final result would be something similar to the display in the following illustration.
In the example, the categories Concept Creation, Specification, and Solution Validation are displayed from bottom to top on the Y axis. The substeps are included in accordance with the data in the underlying table, that is, only Chart and Web Runtime are included for the Concept Creation category because no data is specified for the Object Services substep. The start point and end points for the substeps are displayed on the X axis (month, day).
Milestone Trend Analysis Chart (MTA): The other class 4 chart type is the milestone trend analysis, which enables you to monitor the contents of the project progression. Based on thedata, you define the milestones and then can schedule appointments and display anydeviations. The Y axis is defined as the target time axis with the scheduled milestones; the Xaxis represents the actual time axis. The appointments for project meetings (reporting times)are recorded on this X axis. In these project meetings, for each milestone, each owner isasked about the upcoming fulfillment date. The fulfillment dates named are entered into thechart using the meeting time. A forecast curve is produced for each milestone. If target andactual times coincide, the milestone for the scheduled time has been reached and theforecast curve runs horizontally. In the other cases, if the milestone is moved during theproject meetings to a later time or earlier time, the forecast curve rises or drops, respectively.
You build the underlying data provider (query) for a milestone trend analysis in this manner:
Therefore, as your format, you would have within the columns specific projects— say, with two columns each, the first being the reported time and the second being the milestone time—and down the rows would be the actual milestone. This would give you the formatting required to fit the information required by the MTA template, shown here in the Chart Designer.
The MTA can also be created without categories, which means it contains data series only.
e.Class 5 Chart Types
The final group, the class 5 chart types, comprises the delta chart and its variant waterfall chart. These chart types use the “flowing” concept to display information. You probably have seen or worked with this type of chart when the goal is to group together multiple sets of information, normally to offer some sort of range of information and display it graphically.
Delta Chart: A delta chart outlines the development of a total value by displaying various interim values. Important to note is that these interim values are not displayed as subtotals, but as deltas. The delta chart only deals with flat data tables; this means the interim values are individual items. If you have a hierarchical data table—that is, if the interim values are made up of multiple single items—the system automatically selects the waterfall chart variant to display the data. The differences between the chart types delta chart and waterfall chart are outlined.
Differences Between the Chart Types Delta Chart and Water fall Chart
An example of a table for a delta chart.
The critical concern here is that the table does not contain a hierarchy (hierarchy nodes), so any further drilldown is not possible. The totals used in the chart are the items Cost of Goods Sold (start value), which is displayed on the debit side, and Revenue (end value), which is displayed on the credit side. All remaining values are interim values, which are displayed in the chart as deltas. The start and end values (including all interim values) are used to calculate the total sum (Overall Result). The results are displayed in the following illustration.
This example includes the items Cost of Goods Sold and Revenue as totals (displayed at the two ends of the graph). The four interim values are displayed as deltas: Customer Rebates is therefore the difference between the previous value Cost of Goods Sold and the following value Overhead Costs, and so on.
Waterfall Chart: As I mentioned, a variant of this chart type is the waterfall chart and it is probably the one that we recognize as the more commonly used chart type. In terms of the data structure for this type of chart, we can have a number of different formats:
You can see that this information would fit well into the configuration format of the delta chart type. The following illustration shows the initial configuration screen used for the delta chart type. I’ve chosen the 2.5D display to give this view a bit of depth.
In some cases where the waterfall format is required, some additional wrinkles may come into play. For example, if the report requirement is to include not only these start and end values but also averages and actuals, this doesn’t quite fit into the chart type delta because of the additional requirements. So, to facilitate this, we use another chart type to mirror image the waterfall format and also support the addition of averages and actuals. An example of the type of data contemplated here is shown in the following illustration.
In this data set, we have Customer Contracts with the Minimum, Maximum, Average, and Actual information. With this set of data, we can develop a waterfall chart either with the delta chart type (making some minor adjustments) or by using a more basic approach and starting with a line chart type. If you work with the delta chart type, the approach is straightforward and you will see that the values for each of the groups fit nicely into the requirements. Going a step further and using the line chart type to format a waterfall chart, you would start in the Chart Designer with this type and in Step 3 of 6 you would change the individual series into Stacked Columns to generate the range format for the difference between the minimum and maximum, as shown in the following illustration.
The trick here is that once you are finished setting up your stacked columns, you can then go back and hide the sections of the stack that you don’t want to see. This generates a waterfall view of the data using the line and stacked chart types. I’ve used this approach to show that even the more complex chart types can be imitated by using either two or more basic chart types that are available. Once you’ve configured these ranges, you can still keep the Actual and Average to display as the line chart. The following illustration shows the end result of this format.
As you can see, you have the floating ranges for Minimum (the lower edge of each of the floating boxes) and Maximum (the sum of the minimum and maximum numbers is the size of the box), the Average and Actual displayed as a line chart. This is a complex chart type but results in a very straightforward display of the data.
Formatting Charts Using the Wizard
This section explains the series of wizard steps that you go through to set up any of the chart types previously discussed. As with any of the objects within BI, there are some parameters and functions that are consistent across all the different chart types. If we look at those consistent parameters, we can cover about 75 percent of all the activities required to develop a chart. As I mentioned earlier, the great thing about this process is that it’s very close to or the same as using the familiar chart options with Excel, and we have a wizard to help us move through the steps. The wizard helps you to format a chart in just a few steps. We’ve already reviewed all the chart types so th
SAP BI Related Interview Questions
|SAP BO Interview Questions||SAP ABAP Interview Questions|
|SAP BW Interview Questions||SAP BPC Interview Questions|
|SAP BODS Interview Questions||SAP BDC Interview Questions|
|SAP BW on HANA Interview Questions||Sap Bapi Interview Questions|
|Sap Business One Interview Questions|
Sap Bi Tutorial
Bex Web Analyzer Reporting Functionality
Getting A Fast Start With Bi Patterns In The Bex Web Template
Basics Of The Web Application Designer
Advanced Configuration Using The Web Application Designer
Advanced Functionality Of The Report Designer
Developing Effective Web Reports
Developing High-impact Dashboards
Migrating 3.x To 7.0 Bex Web Reports And The Wad
Integration Of Sap Businessobjects Components Into The Bi Environment
All rights reserved © 2020 Wisdom IT Services India Pvt. Ltd
Wisdomjobs.com is one of the best job search sites in India.