Analysis/BI Overview

This article is not a how-to guide or a tutorial on how to build analysis graphs and pivot tables, but a conceptual overview on the various tools and settings that are available. It is recommended that you get a basic acquaintance of the analysis tools first and then use this as a reference to learn about more advanced capabilities. Also see the Visualization Gallery for a visual demonstration of the capabilities.

An analysis graph/table is based on a query. A query is constructed out of entities, and the values for these entities are then arranged in a table or graph for visualization. The types of entities are:

• Indicators
• Dimensions

Filters constitute an additional type of query entity, but they are used for restricting the results rather than for displaying information.

Included in the Dimensions are special timescale Dimensions such as Year, Month, Hour of the Day, Month of the Year, etc. These can only be added once an Indicator or Dimension exists in the query, which defines the query base table, as well as the date field to use for the timescale Dimension.

Both Indicators and Dimensions may be duplicated (right-click/Duplicate) and the same Indicator/Dimension can then be added twice to the same query. Example uses:

• Displaying the same data in multiple ways. For example, as a percentage, as an accumulated value, and as the original value.
• Displaying the same value using multiple visualizations. For example, displaying the same data in both a bar graph and a line, or displaying the same data in both a pie graph and a gauge.
• Filtering each duplicated Indicator a different way. For example, displaying the Sales twice, once filtered for USA and the other filtered for European sales.
• Setting a different date field setting for each Indicator. For example, displaying Sales by Year data once according to the Sales 'Created Date' and again by the 'Closed Date'.

Indicator-Dimension comparison matrix:

An analysis table/graph may consist of:

• A combination of Indicators and Dimensions where the Dimensions group the Indicator aggregated values. For example: Sales by User.
• If an Indicator is sliced by its own base table (e.g. User:Quota by User) this effectively creates a non-aggregated list of records which can then be highlighted using rules and colors, etc.
• Indicators only, for displaying a scorecard, or for plotting two Indicators against each other. For example, display global Sales and Costs for this year.
• Dimensions only, for plotting two fields against each other in a grid. E.g. which Users belong to which Workgroups.

There are many types of calculations that may be performed for different purposes:

• Indicators are always aggregated using an aggregate function: Sum, Average, Min, Max, Count, Standard Deviation, etc. This function is specified in the 'General' Settings for that Indicator.

• If the above aggregation runs on millions of records and is too slow, there is also an option to pre-aggregate values and store them in a temporary table for analysis using Bulk Actions.

• Custom Functions may be used as Indicators to calculate a value using database fields and math, date or even logical functions. Examples: Price+VAT, 100/Price*Discount, DayDiff(StartDate,EndDate), if(Value>100, Value/100, Value), etc. Click on “Customize this list” in the Indicators list to add a function.

• Custom Functions can also use aggregated values so that the calculation is done after the values have been joined and aggregated: Examples: “total Sales-total Costs” (this is not the same as Sales-Costs at the field level because of joins that must be performed between the Sales and the Costs), “minimum Task Start Date - minimum Project Start Date”, etc.

• Custom Function fields can be used as Dimensions to slice Indicators differently. Examples: First letter of the name, concatenation of two values (City + Country), a function that groups together numeric values into discrete values, etc. These must be added as fields first using the Custom Fields tool.

• Custom Functions can use external database code to perform custom complex calculations. See External Functions.

• Custom Functions are always calculated on the fly at the ClayCentral database. If this becomes a performance issue, the calculations should be converted to Automation actions that store the calculated values in fields. In addition, Automation actions can be used to perform a series of complex calculations that must manipulate the values in stages.

• Table Calculations: To perform calculations between two or more values/records (as opposed to calculations between two or more fields), use the Table Calculations feature in the Settings/General/Calculation tab. Some examples: Convert values to a percentage per dimension, calculate an accumulative sum, display a ranking number instead of the value, calculate the difference from previous/next values, show a moving regression slope, etc.

• Trend-lines: Graphs can display a trend-line using any of a dozen pre-defined calculations (regression, exponential, average, etc). Use the Settings/Graph/Trends tab for this. Additionally, you can create your own trend-line using the Table Calculations feature above to convert values, and then show the results as a line in a graph.

• As explained above, Dimensions are always added to every Indicator in the query, effectively slicing all Indicators in a single query by the same Dimensions. Sometimes it is necessary to perform calculations between Indicators that have been sliced differently. For example: “Quota by User” minus “Sales by User by Year”. Use the Analysis Matrix feature for this. Using the Analysis Matrix feature, Indicators can also be combined in multiple stages for complex calculations.

There are several types of filters that can be added to an analysis query (see also Advanced Filters and Custom Functions for general filtering features):

• Global filters (the Filters button next to the graph/table): This filters all the Indicators and Dimensions in the query and the filters are joined to the query using the base table. For example, Sales and Costs by User filtered by Department=“European”. In this case Department will be joined to the Sales(Opportunity) and Costs(Tasks) tables.

• Indicator filters (Settings/General/Filters tab): Applied only to a single Indicator. For example: Sales for 2008, and Sales for 2007 can be added as two Indicators, each with a different Year filter value.

• Filters on aggregated Indicator values (Settings/General/Filters tab): Filter on the results of the aggregation. For example, Sales by User where the total Sales for that User exceeds $10000.

Note that the date filter above the graph area is a global filter, available there for convenience due to frequent use. It filters on the date field defined as the Indicator date field in Settings/General/General.

A query entity (Dimension, Indicator or Group) can be placed in various 'locations' for displaying the same results in different arrangements. This includes pivoting functionality as well as other features.

Note that since results are displayed simultaneously as both a table and a graph, each of these locations may mean different things depending on whether you are currently viewing the graph or the table.

• Rows (first level): This appears on the y-axis for graphs, and as cell values/rows in the table ('Net Sales' in the example). Note that if there are multiple row values per column, the graph will display multiple plot points on the same column, and the table will display multiple rows. For special handling by other types of graphs, see Graph Types below.

• Rows (second level and up): This appears as a pivoted row in both the table and graph with a separate row of graphs per value, or a pivoted row section per value in the table (User in the example).

• Columns (first level): If the 'Multi-Column'/'First column as pivot' setting is toggled, the first level column will be displayed as a pivot (see next). Otherwise: The values will be displayed along the x-axis in the graph, and as an additional non-pivoted column in the table ('Quarter' in the example). For special handling by other types of graphs, see Graph Types below.

• Columns (second level and up): This appears as a pivoted column in both the table and graph with a separate column of graphs per value, or a pivoted column section per value in the table (Year in the example).

• Series: For the graph, this appears as a new graph series per value. If multiple entities are placed in the series, then each unique combination of all the series values will create a new series. For the table, these will appear as additional non-pivoted columns in the table. I.e. this can be used to display a flat list of many table columns without pivoting. The 'level' setting in this location will determine the order of the columns in the table, as well as the priority for Settings (see below). For special handling by other types of graphs, see Graph Types below.

• Pages: Only one entity is allowed in this location. A navigation bar will appear under the graph/table for scrolling through the page values. For example, Sales by User, with Year as the page, will allow the viewer to see one year at a time or even view all of the years as an animated sequence over time.

• Grouped (see below on Grouping): This means that the Indicator or Dimension appears as part of a group of Indicators/Dimensions and its location is determined by the location of the group. Note that setting the level in this case determines the order of the Indicators/Dimensions within the group.

• Hidden: This is allowed currently only for Indicators. This means that the Indicator is not shown but it can still be used to affect the visualization of the graph/table. For example, a Cost Indicator can cause the Sales values to grow in size or change in color.

• Remove/Aggregate: This is not the same as Hidden. It means that the Indicator or Dimension is removed at the server end before returning results. These entities cannot be used for affecting the visualization of the graph/table at the client end. For Indicators, this is useful when the Indicator is only needed as part of a Custom Function. For Dimensions, the Dimension is re-aggregated and removed. For example, if the query is Sum(Sales) by Year, the Year can then be removed/re-aggregated using an Average function which results in: “Average Sum of Sales per Year”. Note that this is not the same as “Average Sales”.

Note that the boxes for Colors/Shapes/Sizes are not actually locations, but shortcuts for auto-assigning colors/shapes/sizes based on values in an Indicator or Dimension. For example, dragging 'User' into 'Colors' will change the values of the graph/table, assigning a different color per user. An Indicator can be dragged into any of these boxes and removed from their proper locations in order to set them as 'Hidden'.

Methods for changing Locations and Levels of query entities:

• Check/Uncheck Indicators or Dimensions in the list of available Indicators/Dimensions. New entities will be auto-assigned a location.
• Drag-n-drop entities from one location to another, or from the list of Indicators/Dimensions to a location.
• Change the Location/Level settings in the Settings dialog.
• Use the drill-down menu.

Drag-and-drop actions:

• Drag an unchecked entity in the Indicators/Dimensions lists to a location to add it to the query.
• Drag an entity from the Indicators/Dimensions lists on top of an existing entity in the location boxes in order to replace it.
• Drag an entity from a location box back to the Indicators/Dimensions lists to remove it from the query. If the Indicator is also in a Colors/Size/Shape box, it will only be set to Hidden and not removed.
• Drag an entity from any list/location to a new location.
• Drag within the same location box to a different row to change its level/order.
• Drag an existing grouped entity from the Indicators/Dimensions lists into a location box to ungroup it.
• Drag an ungrouped entity onto a Group entity to group it.
• Drag an empty area in the Rows list into the Columns list to switch all entities between the two locations, and vice versa.

Grouping is a powerful but abstract concept that allows grouping multiple Indicators or Dimensions onto the same pivot level and location.

By default, new Indicators are added grouped, and Dimensions are not.

Consider the following example, displayed in four progressive screenshots:

1. We start with a simple graph that displays 'Amount of Opportunities' (an Indicator) by Year. The Amount is displayed on the y-axis, and the year on the x-axis.

2. Now we want to add 'Amount of Tasks' as a second Indicator and display it also in the Rows location/axis. Without grouping, the screenshot shows what happens. The new entity is added as a new Row pivot, and one row of graphs is displayed for every unique Amount of Tasks value.

3. If we group 'Amount of Opportunities' and 'Amount of Tasks' together, we get two entities to position instead of the two Indicators: 'Indicator Names' and 'Indicator Values'. The third screenshot shows what happens when the 'Indicator Values' is displayed as the first Row entity, and the 'Indicator Names' as the second Row entity. Note that both Indicators are arranged on the same pivot level.

4. The fourth screenshot shows what happens when we move the 'Indicator Names' to Series, and the 'Indicator Values' stays in the first level Row. Each Indicator entity creates a separate graph series instead of its values, which would not be possible without grouping.

To dig into this concept deeper, consider the following:

• The 'Names' entity is a collection of the names in the group, not the values. In the above example this contains two values: 'Amount of Opportunities' and 'Amount of Tasks'.
• The 'Values' entity is actually a multi-dimensional collection containing multiple sets of values, one for each group member. In the above example, it contains two sets of values for the two amounts.
• If the 'Names' entity is on a higher level than the 'Values', each time the 'Values' are shown, it will only show the values for the current group member. See example #3 - each graph displays a different Indicator on its y-axis. In this case it can be said that the 'Values' will only contain one set of values at a time.
• Similarly, when this concept is applied to a graph and the 'Names' is positioned in a graph series or column, then each set of row values will contain a different set of 'Values'. With pivoted levels this is no problem because each set of values is shown in a different row/column. But inside a graph, this means that multiple values must be shown against the same x or y axis even though each set belongs to a different Indicator/Dimension. For this reason, 'Indicator Names' can be used as the graph series or columns because a graph can show multiple numeric values on the same axis, but 'Dimension Names' must be on a second level or higher because a graph cannot display multiple Dimension sets on the same axis.

Limitations: Based on the above and due to limitations in tables and graphs, there are a few limitations on where the group 'Names' and 'Values' entities can be placed. For example, 'Names' must be on the same level or higher than 'Values'. The editor will automatically correct illegal placements of the grouped entities as necessary.

Uses:

• Showing multiple Indicators on the same graph axis, as demonstrated above.
• Comparing multiple Indicators side by side in rows/columns of graphs or table pivot sections, as demonstrated above.
• Comparing two variations on the same Indicator side-by-side, for example Min Sale vs Max Sale.
• Comparing the same values sliced by different Dimensions side-by-side for visual scanning. For example, Sales can be sliced by Category, Country, User, Industry, etc to quickly and visually identify sales patterns.
• If you are displaying multiple indicators in a graph and need to display the same data in two different graphs or visualizations side-by-side with automatic scrolling, you can use both grouped Indicators and grouped Dimensions (duplicate the Dimension and group it). See the Visualization Gallery for examples.

There are several axes in analysis tables and graphs:

• Graph axes: The x-axis and y-axis in graphs themselves mapped to first level Columns and Rows respectively. For other special graph types, see below.
• Graph pivot axes: when multiple entities are added to Rows/Columns.
• Table pivot axes: when multiple entities are added to Rows/Columns.

All of the above are considered axes, and most settings and concepts for axes such as “Equal Axes”, “Reverse”, “Show Titles”, “Axis Font”, etc. apply to them all. Only some axis settings such as “Swap Scales”, “Show Line”, “Ticks”, etc. apply to graph axes only.

Axes may display any of three things, and settings may be applied separately to each of these:

• Titles: The title of the axis or pivot, e.g. “Sales” or “User”.
• Values: The values in that axis, e.g. “$1000” or “James”.
• Totals: In an analysis table only, pivots are by default totalled and sub-totalled, adding another row or column per pivot in order to display these.

By default, each axis will only display the values that appear in that pivot section. For example, “John” and “Mary” may have made sales amounting to $500 and $1000 in 2006, and the graph will display these, but if, in 2007, only “John” has sales of $50, the graph for 2007 will show a Sales axis of 0-50 only, and only one user.

The same behaviour applies to pivots in tables. Only the relevant data is displayed in each pivot section to save space, and this allows viewers to determine which dimensions/values are relevant to which pivot section.

But, sometimes, especially with multiple graphs, you would want to be able to quickly and visually compare the graph results without reading and comparing the values. To do this, the “Equal Axes” setting must be enabled for the Sales Indicator, the User Dimension, or both. This ensures that each graph will have the same range of values in its axis, and that the graphs can be visually compared to each other according to the size and position of the plot points.

Pivoted and table axes are always discrete.

A graph axis can be either discrete or continuous. A continuous axis means that:

• The distance between values are always linear. For example, values of 1, 2000, 2001 means a large gap will be shown between 1 and 2000, whereas a discrete axis would only plot three equal distances.
• …Which means that mumeric values can be compared according to their position/size in the graph.
• Additionally, timescale and date values are plotted in their proper timescale distance.

The following rules apply:

• Indicators are always displayed in continuous graph axes. To show field values as discrete values, you must add the field as a Dimension.
• Non numeric/date/duration Dimensions can only be displayed as discrete values.
• Timescale Dimensions are displayed in continuous axes by default but can be changed in the Settings/General/General tab.
• Numeric/date/duration Dimensions are shown in discrete axes by default but can also be changed.
• Text values can be converted to numeric values and vice versa using Custom Functions.
• An upcoming Discretize function will allow converting numeric data into discrete grouped ranges.

The Sort button allows sorting by up to six entities, either Indicators or Dimensions. The sort can be defined as ascending or descending.

The sort will affect how values are shown in axes and pivots.

There is also a “Reverse” setting in Settings/Graph/Axes which affects the sorting, but is not quite the same:

• If defined, it overrides the ascending/descending order defined in the Sort options even for the analysis table.
• For graphs, this is the only way to reverse a continuous axis.
• For certain graph types, this reverses how the plot points or graphics are drawn. For example, bars would be drawn from top to bottom instead of starting at the bottom.

Tables by default display totals and pivot subtotals for all numeric values and Indicators.

For a non-pivoted column to display totals, both the 'Total' and 'Subtotal' settings must be on.

For a pivot to show subtotals, the 'Subtotal' setting must be on for that entity.

Example:

Sales & Costs (non-pivoted Series and Rows) pivoted on a 'User' dimension.

If Sales has 'Total' checked and Costs has it unchecked, each series column will only total the Sales column.

Each User pivot section will subtotal only the Sales column as long as the User dimension has 'Subtotal' checked.

When pivoted Rows/Columns are added to a query, multiple graphs are displayed in a grid (as shown in examples above).

By default the grid size is limited to 5 rows and 5 columns of graphs in order to make the graphs readable.

This grid can be adjusted using the settings under the graph area. There are three settings:

• Maximum amount of graphs per row
• Maximum amount of graphs per column
• Whether to wrap a long row or column to the next row/column. This setting only has an effect when there is only one row or one column, and the amount of graphs exceeds the grid size settings.

If the grid of graphs does not display all available graphs, scrollbars will automatically appear for displaying the next/previous batch of graphs.

Note that there is another setting which may cause the vertical scrollbar to appear: “Columns Per Page” in Settings/Graph/Graph. This is useful when there are too many columns in the graph x-axis and you want to be able to scroll through them rather than display them all at once. However, if the grid cannot show all of the graphs, the grid scrollbar will override this scrollbar.

Graph Types are grouped under two categories: Graphs that can show multiple columns (or that have at least two axes and an ability to show series), and graphs that can only show one set of values.

Multi-Column graphs include:

• XY: Bar, Line, Scatter, Bubble, Area, Box, Whisker, Open-Close
• Radar: Line, Scatter, Bubble, Area
• Contour graph (which includes also a z-axis)
• Surface graph (which includes also a z-axis)

Non multi-column graphs include:

• Pie and Donut
• Pyramid and Cone
• Gauges: Semi, Wide, Circle, etc.

Using series, some graphs can be combined in a single graph. For example, Sales can be displayed as bars, and Costs as lines. Combinations allowed are:

• All XY graphs as well as Contour graphs.
• All Radar graphs.

The rest of the graphs can only be combined by using pivots and multiple graphs. This allows comparing the values side by side using different graph types. But multi-column graphs cannot be combined with non multi-column graphs even using this method unless you add two separate analysis components to a dashboard or document template.

A Surface graph uses the Series as its y-axis and the first Row entity as its z-axis.

A Contour graph uses the Series as its z-axis and the first Row entity as its y-axis. This is to allow combinations with the rest of the XY graphs.

An Interline graph collects sets of two series, draws two curved lines for them, and fills the area in between them using the colors for both lines depending on where the line is relative to the other line.

Radar graphs show the first Row entity on the radial axis.

Pie, Donut, Pyramid, and Cone draw one section per series value and the first Column entity is shown as a pivot.

Gauges display one pointer per series value, supporting multiple pointers, and the first Column entity is shown as a pivot.

Special notes for colors, shapes and sizes per graph type:

• All graph types support multiple colors.
• The surface/contour graphs can only change color on its z-axis. If colors change based on another axis, it makes its best effort to approximate the color against the z-axis.
• Currently, lines and areas can only display up to 256 colors on a single line or area series. But lines also show colors using marks.
• Lines by default show marks on top of the lines in order to support unlimited colors, sizes and shapes. These can be switched off using the Settings, or changed to Bubbles which don't support shapes.
• The following graph types do not support varying sizes: Area, Line without marks, Surface, Contour, Pyramid/Cone.
• Pie graphs support sizes in 3D only.
• Only the following graph types support varying shapes: Bars in 3D, Lines with marks (XY and Radar), Scatter, Gauges (pointer shapes).
• The Pie graph shows varying shapes as the 'exploding' effect.
• Analysis tables support colors and sizes, but not shapes.

Settings for analysis visualizations are applied using an advanced matrix of defaults, query entities, rules, values, and priorities. This section lists the basic concepts for applying settings, as well as some special settings.

There are three groups of settings:

• General. This is only enabled for query items, not the [Default] entity. These settings are mostly settings that affect the query or the values rather than the visualizations. For example, Indicator filters, calculations, the Equal Axes setting, query entity location and level, etc.
• Graph: All the graph options, separated by the graph entity (points, sets or series, axes, the graph itself, etc).
• Table: Settings that only affect the table. Note that some settings are duplicated so that they don't affect both the table and graph together. For example, the graph axis font may need to be changed for readability, but the table axis font is fine as is.

Settings can be defined at:

• The [Default] level.
• Per query entity. For example, each Indicator can have its own font.
• Per query entity value. For example, Indicator values of 1000 and up can change color, or Dimensions in position 4 can have its own font, or the existence of a Dimension with the value 'Europe' can change the graph background.

Settings applied to a specific query entity can affect other query entities depending on what is set in the 'Apply To' drop down:

• [Self] only affects itself.
• [Self and below] affects all query entities on the same level or below in the pivot structure. For example, if Year is set as Row level 2, anything in Row/Column level 2 and under will be affected, as well as anything in the Series location.
• [All]. Note that choosing this effectively makes it the same as a [Default] setting for all query items, except that it gets a higher priority than the [Default] settings.
• A specific query entity other than itself. This is self explanatory.
• In addition, the table options include an Apply To drop-down that determines whether the settings affect the original values, the total values, or both.

A special note regarding graph options: All graph settings except Axis and Zone settings always apply to [All]. This is because it would be difficult to figure out which entity is shown as a graph, which as a set, and which as a plot point. So, for example, setting the graph background in a query entity shown as a line will affect the graph that it is in. Axis and Zone settings do not work this way for the simple reason that there are multiple axes per graph, and a user needs the control to decide which specific axis to apply the settings to.

Note that, using the Value drop-down, multiple rules can be applied. Examples:

• Different fonts or colors per value or range of values.
• Add two [Default] values, one that applies green to table values, and another that applies blue to table totals.
• Add two of the same value to apply settings to two specific query entities. Each value would have a different Apply To setting in this case.

All visualization related settings have at least three values: On/yes (or a specific value), off/no (or a zero/empty value), and undefined. If the setting is undefined/unchecked at the highest level, it tries to find the setting value at a lower level, until it reaches the built-in defaults.

Settings are applied using the following hierarchy, sorted by priority:

• Colors/sizes/shapes settings which currently have their query entity stored in the Colors/Sizes/Shapes drag boxes. I.e. anything dragged into these boxes will have the highest priority to ensure that it changes the visualization no matter what other settings are applied.
• Settings applied to a specific value in the query entity. Value priorities go according to the order they were added, with the last getting the highest priority.
• Settings applied to the specific query entity currently being displayed, as selected in the Entity drop-down.
• Settings applied to another query entity which has the Apply To setting configured to affect the current query entity. For example, a User entity that affects the color of the Sales values.
• User defined defaults: The entity drop-down in the Settings dialog contains a [Default] entry. This allows applying settings to all query entities globally.
• Built-in defaults. If the setting isn't defined at any level, hard-coded lowest-denominator defaults are used that make sense for the most amount of graphs.

Many settings affect the various graph types differently, or have no effect whatsoever on some graph types. For example, a bar is only affected by the Shape setting if it is in 3D, Swap Scales only works for XY graphs, a separate Plot Area background only exists for XY and Radar graphs, Labels appear in different places for each type of graph, etc.

For a list of how the color, sSales by User by Yearhape and size settings affect graph types, see the Graph Types section above.

Axis Range & Ticks: These settings are used only as a suggestion and may be ignored if the values are not reasonable. For example if the range is 1-1000000 with ticks every 10 points, this will not work.

Ticks show values every major point, and only tick lines for minor points.

The Axis Range setting is not enabled for the [Default] entity because it does not make sense to apply a global setting for all Indicators and Dimensions.

3D Gap/Depth: This only applies with 3D XY graphs between multiple sets/series and even then, it is ignored for stacked/side/percent bars and area graphs.

The 3D Graph Options section is only relevant for Surface, Pie, Donut, Pyramid and Cone graphs, but this may change in the future.

Swap Scales isn't the same as swapping the entities in the Columns/Rows. It affects how certain graph types are drawn. E.g. bars and boxes will be drawn from side to side instead of from top to bottom, and lines will be drawn from top to bottom.

Reverse vs. sort: See the Axes section above.

Several axes settings have a value listed as 'First/last'. This only has an effect when pivots are shown. By default, the same headers or axes would be shown repetitively for each pivot section. This setting allows you to show the axis lines only for the first/last pivot section, or to show the table series headers only for the first table pivot row, etc. This saves space, but normally it requires that you switch on the “Equal Axes” setting so that multiple graphs/pivot sections can be compared visually without their axes. In other words by using the 'first/last' setting, multiple graphs or table pivot sections will have different elements drawn depending on whether it is first/last or not, causing them to be drawn with different sizes, which makes it hard to compare them visually at a glance.

The difference between 'first/last' and 'first/last only' (graphs only) is that with 'first/last' it leaves room for the element (e.g. the axis title) but doesn't actually display it so that the graph will be the same size as the rest, and with 'first/last only' it maximizes the space, completely removing elements that are not used (such as axis values), thus saving space but causing the graphs to appear with different sizes.

Plot Area Margin: Although only the XY and Radar graphs have a Plot Area background, the Plot Area Margin setting applies to all graphs. This setting is very important and allows the user to adjust the margin between the border and graph in case axes or labels exceed the boundaries and are cut off. For example with a Surface chart, often the 3D axes will be cut off. The margin values can also be negative to remove extra spacing, or to purposely cut off the axes.

The Frame Depth setting can be positive or negative numbers for creating a protruding or indented frame.

Some settings work differently depending on whether they are defined for the [Default] entity, or a specific query entity, but only if multiple graphs are shown:

• The graph background settings will affect the whole graph area if defined as the default, but will only affect the specific pivot graph if set for a query entity.
• Note that this gives you three levels of background colors in some cases: The main graph area including the pivots, the graph area of each individual graph, and the plot area inside the graph.
• Same goes for the Frame settings: Complete graph vs. individual graph.
• Same goes for the graph Title settings.
• Legends can be switched off for specific query items only, using the Legend settings in that query entity.

The Labels settings are not the same as axis labels. These show up inside the plot area for all graph types, including gauges and pie graphs.

Columns Per Page: See the Axes section above.

Zones, Marks and Color Zones use static values for now, but you can add dynamic values from the database as areas, lines or boxes to have a similar effect as marks and zones.

This BI tool is new to Digital-Clay version 9 and it necessitated significant changes to the infrastructure and querying code. 8.x and older queries will be converted on the fly when they are loaded to the best of its abilities.

Note that there are no issues with converting list/browser queries. Only with analysis queries.

The possible issues are:

• Various graphical changes and improvements depending on the graph type.
• The gauges will look very different and now fill up all of the space given to it, which may cause much bigger gauges if the size wasn't set appropriately.
• Gauge axes ranges are now set to the maximum value range by default instead of double the range.
• The “Show series horizontally” setting becomes a full pivot instead of a partial pivot as before, which means that all other columns will be included inside the pivot when converted.
• Gauges no longer have default zone colors unless explicit highlighting settings are set. Old highlighting rules will be converted to Zones.
• The Merge feature for merging multiple analysis queries has been discontinued, and conversion of these will only be successful or partially successful depending on the complexity and contents of the merged query. Most common cases will be converted, and the rest may be rebuilt manually using the Analysis Matrix feature.
• Conversion places priority on whether the default display option is a table or graph. In some cases the primary display option (e.g. the table) will be converted properly but not the other (e.g. graph).
• In tables, unless you specified a column order, it will use the new system defaults which may not match the previous defaults. So some columns may show in an unexpected order.
• A bubble graph that uses a series value for bubble sizes will not be converted properly - but in most cases bubbles use indicators which will be converted successfully.
• Custom functions added via the analysis table Columns button are no longer supported. These must be added as custom function fields.