Reporting Language > Creating a Graph > Selecting a Graph Type

Selecting a Graph Type

Topics:

When creating a graph, it is important to select the appropriate graph type with which to display your data. You may select from a number of basic graph types, as well as refinements on these types. Basic graph types include line graphs (connected point plots), bar graphs, pie graphs, and scatter graphs. Use the brief descriptions (see Graph Types) to select a graph type that suits the data set you are displaying and the change you want to highlight. Keep in mind that the data are the sets of numbers that you are displaying, and the scales are the numbers or variable measures displayed along the axes of the graph.

Note: When using a stacked chart of any type at least 2 series are required.

Graph Types

Following are descriptions of the types of graphs you can create:

  • Line graphs. Line graphs are useful for emphasizing the movement or trend of numerical data over time, since they allow a viewer to trace the evolution of a particular point by working backwards or interpolating. Highs and lows, rapid or slow movement, or a tendency towards stability are all types of trends that are well suited to a line graph.

    Line graphs can also be plotted with two or more scales to suggest a comparison of the same value, or set of values, in different time periods. The number of scales your graph has depends on the type of graph you select. For details, see Determining Graph Styles Using LOOKGRAPH.

  • Bar graphs. A bar graph plots numerical data by displaying rectangular blocks against a scale. The length of a bar corresponds to a value or amount. Viewers can develop a clear mental image of comparisons among data series by distinguishing the relative heights of the bars. Use a bar graph to display numerical data when you want to present distributions of data. You can create horizontal as well as vertical bar graphs.
  • Pie graphs. A pie graph emphasizes where your data fits in relation to a larger whole. Keep in mind that pie graphs work best when your data consists of several large sets. Too many variables divide the pie into small segments that are difficult to see. Use color or texture on individual segments to create visual contrast.
  • Scatter graphs. Scatter graphs share many of the characteristics of basic line graphs. Data can be plotted using variable scales on both axes. When you use a scatter graph, your data is plotted using a basic line pattern. Use a scatter graph to visualize the density of individual data values around particular points or to demonstrate patterns in your data. A numeric X-axis, or sort field, will always yield a scatter graph by default.

    It is important to note that scatter graphs and line graphs are distinguishable from one another only by virtue of their X-axis format. Line graphs can appear without connecting lines (making them look like scatter graphs) and scatter graphs can appear with connecting lines (making them look like line graphs).

  • Area graphs. Area graphs are similar to line graphs except that the area between the data line and the zero line (or axis) is usually colored or textured. Area graphs allow you to stack data on top of each other. Stacking allows you to highlight the relationship between data series, showing how some data series approach or shadow a second series.
  • 3D graphs. 3D graphs add dimension to your graphing presentation. Dimensionality allows your viewers to recognize trends based on two or more data sets easily. 3D graphs also add impact to your presentation.
  • Bipolar graphs. A bipolar graph is split along a horizontal line. This type of graph is useful for comparative trend analysis of widely disparate data values over time or other sort values.
  • Radar graphs. This is a type of circular graph used when categories are cyclical. Radar graphs are essentially analogous to a line chart, except that the scale wraps around. Radar graphs work well with any data that are cyclical, such as the months of a year.

Selecting Scales

How to:

After you have chosen a graph type, you should select an appropriate scale. A scale is a classification scheme or series of measures that you select for application to the axes of your graph. The scale provides the framework against which your data are plotted. When you choose an appropriate scale for your data, meaningful patterns can emerge, and when you modify a scale, the overall shape of your graph changes.

Steps or measures in the scale are represented along the axes of your graph by marks. The type of scale you choose determines the number of divisions along the scale. There are two general types of scales you can apply to the y-axis of your graph:

  • Linear scales
  • Logarithmic scales

A linear scale is a scale in which the values increase arithmetically. Each measure along the scale is one unit higher than the one that precedes it. Linear scales are useful when the data you are plotting are relatively small in range.

A logarithmic scale is a scale in which the values increase logarithmically. Each measure along the scale represents an exponential increase in the data value. Logarithmic scales are useful when you need to accommodate a large range of numbers.

Syntax: How to Select Scales

To use logarithmic scales in your graph, add the following to your GRAPH request: 

ON GRAPH SET STYLE *
*GRAPH_SCRIPT
setY1LogScale(value);
*END
ENDSTYLE
END

where:

value

Is one of the following:

true turns on logarithmic scaling.

false turns off logarithmic scaling. Linear scaling is used instead.

Determining Graph Styles With Display Commands and Sort Phrases

How to:

Each GRAPH request must include a sort phrase and at least one display field (up to five are allowed).

The fields, which are the subjects of the graph, may be real or virtual fields, with or without direct operation prefixes (AVE., MIN., MAX., etc.). They may also be calculated values.

Note: Display fields used only for calculations need not appear in the graph. You can use the NOPRINT or SUP-PRINT phrases to suppress the display of such fields. For details, see Sorting Tabular Reports.

By default, a particular combination of display commands and sort phrases determines the graph format. The combinations are:

Graph Type

Display Command and Sort Phrase

Line graph

 
PRINT A {ACROSS|BY} B

(where B is alphanumeric)

Vertical bar graph

 
SUM A ACROSS B

(where B is alphanumeric)

Horizontal bar graph

 
SUM A BY B

Pie graph

 
SET LOOKGRAPH=PIE
SUM A {ACROSS|BY} B

or

SET PIE=ON
SUM A {ACROSS|BY} B

Scatter graph (without connecting lines)

 
PRINT A ACROSS B

(where B is numeric)

Scatter graph (with connecting lines)

 
SUM A ACROSS B

(where B is numeric)

You can override the default graph format using the LOOKGRAPH parameter. For details, see Determining Graph Styles Using LOOKGRAPH.

Syntax: How to Create a Line Graph

To create a line graph, issue a GRAPH request with the following display command and sort field combination 

PRINT fieldname1 [AND] fieldname2...
{ACROSS|BY} sortfield

where:

fieldname1...

Is the name of the field to be displayed on the Y-axis of the graph. There can be a maximum of 5 display fields in a GRAPH request.

AND

Is an optional phrase used to enhance readability. It can be used between any two field names and does not affect the graph.

sortfield

Is the name of the field to be displayed on the X-axis of the graph. This must be an alphanumeric field in order to generate a line graph. If the field specified is numeric, you can still create a line graph by using the LOOKGRAPH=LINE parameter. For details, see Determining Graph Styles Using LOOKGRAPH.

Example: Creating a Line Graph

The following illustrates how to create a line graph using the LOOKGRAPH command: 

SET LOOKGRAPH = LINE, GRID=ON
SET HAXIS=600, VAXIS=315
GRAPH FILE GGORDER
HEADING CENTER
"Sample Line Graph"
SUM QUANTITY 
ACROSS PRODUCT_DESC
WHERE PRODUCT_DESC EQ 'French Roast' OR 'Hazelnut' OR 'Kona'
END

The output is:

Syntax: How to Create a Horizontal Bar Graph

To create a horizontal bar graph, issue a GRAPH request with the following display command and sort field combination 

SUM fieldname1 [AND] fieldname2...
BY sortfield

where:

fieldname1...

Is the name of a field to be displayed on the Y-axis of the graph. There can be a maximum of 5 display fields in a GRAPH request.

AND

Is an optional phrase used to enhance readability. It can be used between any two field names and does not affect the graph.

sortfield

Is the name of a field to be displayed on the X-axis of the graph. A separate group of bars is created for each value of the BY field, and each group contains one bar for each display command (SUM) object.

Syntax: How to Create a Vertical Bar Graph

To create a vertical bar graph, issue a GRAPH request with the following display command and sort field combination 

SUM fieldname1 [AND] fieldname2...
ACROSS sortfield

where:

fieldname1...

Is the name of the field to be displayed on the Y-axis of the graph. There can be a maximum of 5 display fields in a GRAPH request.

AND

Is an optional phrase used to enhance readability. It can be used between any two field names and does not affect the graph.

sortfield

Is the name of an alphanumeric field to be displayed on the X-axis of the graph.

Example: Creating a Horizontal Bar Graph

The following illustrates how to create a horizontal bar graph: 

GRAPH FILE GGORDER
HEADING CENTER
"Sample Horizontal Bar Graph"
SUM QUANTITY 
BY PRODUCT_DESC
WHERE PRODUCT_DESC EQ 'French Roast' OR 'Hazelnut' OR 'Kona'
END

The output is:

Example: Creating a Vertical Bar Graph

The following illustrates how to create a vertical bar graph: 

GRAPH FILE GGORDER
HEADING CENTER
"SAMPLE VERTICAL BAR GRAPH"
SUM QUANTITY 
ACROSS PRODUCT_DESC
WHERE PRODUCT_DESC EQ 'French Roast' OR 'Hazelnut' OR 'Kona'
END

The output is:

Syntax: How to Create a Pie Graph

To create a pie graph, issue a GRAPH request with the following SET command and display and sort field combination 

SET LOOKGRAPH=PIE
SUM fieldname1 [AND] fieldname2...
{ACROSS|BY} sortfield

where:

fieldname1...

Is the name of the field to be displayed in the graph. There can be a maximum of 5 display fields in a GRAPH request.

AND

Is an optional phrase used to enhance readability. It can be used between any two field names and does not affect the graph.

sortfield

Is the name of the field to be displayed in the graph. Each value in the sort field will be represented by a section in the pie graph.

Example: Creating a Pie Graph

The following illustrates how to create a pie graph using a BY sort phrase and the LOOKGRAPH command: 

SET LOOKGRAPH=PIE
GRAPH FILE GGORDER
HEADING CENTER
"SAMPLE PIE CHART"
SUM QUANTITY 
BY PRODUCT_DESC AS COFFEES
WHERE PRODUCT_DESC EQ 'French Roast' OR 'Hazelnut' OR 'Kona'
END

The output is:

Syntax: How to Create a Scatter Graph

To create a scatter graph, issue a GRAPH request with the following display command and sort field combination 

{PRINT|SUM} fieldname1 [AND] fieldname2...
ACROSS sortfield

where:

fieldname

Is the name of the field to be displayed in the graph. There can be a maximum of 5 display fields in a GRAPH request. When you specify more than one display field, they are represented by different symbols.

AND

Is an optional phrase used to enhance readability. It can be used between any two field names and does not affect the graph.

sortfield

Is the name of the numeric field to be displayed on the X-axis of the graph.

Example: Creating a Scatter Graph

The following illustrates how to create a scatter graph: 

GRAPH FILE GGORDER
HEADING CENTER
"Sample Scatter Graph"
PRINT QUANTITY AS 'Quantity'
ACROSS PRODUCT_CODE
WHERE PRODUCT_CODE EQ 'B144'
WHERE QUANTITY LT 51
END

The output is:

Determining Graph Styles Using LOOKGRAPH

How to:

Reference:

By default, a particular combination of display commands and sort phrases determines the graph format. You can override the default graph format by using the LOOKGRAPH parameter.

The LOOKGRAPH parameter enables you to change the format of the graph without having to set individual control parameters or restructure the graph request. However, even if you use LOOKGRAPH, you can choose to set individual control parameters (for example, SET GRID=ON).

Reference: Style Options for Line Graphs

Choose one of the following LOOKGRAPH values to change the style of connected point plots:

SET LOOKGRAPH=

Description

 
LINE

A vertical connected point plot graph.

 
HLINE

A horizontal connected point plot graph.

 
HLINE2

A horizontal connected point plot graph with two axes. 

HLINE2S

A horizontal connected point plot graph with two separate axes. 

HLINSTK

A stacked horizontal connected point plot graph. 

HLINSTK2

A stacked horizontal connected point plot graph with two axes. 

HLNSTK2S

A stacked horizontal connected point plot graph with two separate axes. 

HLNSTKPC

A stacked horizontal connected point plot graph showing percentages. 

VLINE

A vertical connected point plot graph. 

VLINE2

A vertical connected point plot graph with two axes. 

VLINE2S

A vertical connected point plot graph with two separate axes. 

VLINSTK

A stacked vertical connected point plot graph. 

VLINSTK2

A stacked vertical connected point plot graph with two axes. 

VLNSTK2S

A stacked vertical connected point plot graph with two separate axes. 

VLNSTKPC

A stacked vertical connected point plot graph showing percentages.

Reference: Style Options for Bar Graphs

Choose one of the following LOOKGRAPH values to change the style of bar graphs:

SET LOOKGRAPH=

Description

 
BAR

A bar graph with the bars displayed beside each other. 

STACK

A bar graph with stacked bars.

 
VBAR

A vertical bar graph.

 
VBAR2AX

A vertical bar graph with two axes.

 
VBAR2AXS

A vertical bar graph with two separate axes.

 
VBRSTK1

A stacked vertical bar graph.

 
VBRSTK2

A stacked vertical bar graph with two axes.

 
VBRSTK2S

A stacked vertical bar graph with two separate axes. 

VBRSTKPC

A stacked vertical bar graph that shows percentages. 

HBAR

A horizontal bar graph.

 
HBAR2AX

A horizontal bar graph with two axes.

 
HBAR2AXS

A horizontal bar graph with two separate axes. 

HBRSTK1

A stacked horizontal bar graph.

 
HBRSTK2

A stacked horizontal bar graph with two axes. 

HBRSTK2S

A stacked horizontal bar graph with two separate axes. 

HBRSTKPC

A stacked horizontal bar graph that shows percentages.

Reference: Style Options for Pie Graphs

Choose one of the following LOOKGRAPH values to change the style of pie graphs:

SET LOOKGRAPH=

Description

 
PIE

A pie graph.

 
PIESINGL

A single pie graph. 

PIEMULTI

Multiple pie graphs.

 
PIERING

A ring-shaped pie graph.

 
PIEMULPR

Multiple, ring-shaped pie graphs of proportional size. 

PIEMULTP

Multiple pie graphs of proportional size.

 
PIEMULTR

Multiple, ring-shaped pie graphs.

Reference: Style Options for Scatter Graphs

Choose one of the following LOOKGRAPH values to change the style of scatter graphs:

SET LOOKGRAPH=

Description

 
SCATTER

Produces a scatter graph.

 
SCATTERD

A dual scatter graph. Values from an additional data set are displayed on a second value (Y) axis. 

SCATTRLS

A scatter graph that labels each data point with its exact numeric value. 

SCATTRLD

A dual scatter graph that labels each data point with its exact numeric value.

Reference: Style Options for Three-Dimensional Graphs

Choose one of the following LOOKGRAPH values to change the style of three-dimensional graphs:

SET LOOKGRAPH=

Description

 
3DAREAG

A three-dimensional connected group area chart. 

3DAREAS

A three-dimensional connected series area chart. 

3DBAR

A two-dimensional bar graph with three-dimensional bars.

 
3D_BAR

A three-dimensional chart with bars.

 
3DCUBE

A three-dimensional bar graph in which all data points are blocks of identical size, hovering at the position that shows their data value. 

3DGROUP

A three-dimensional chart with bars.

 
3DOCTAGN

A three-dimensional bar graph with octagon-shaped bars that have no roots. 

3DPYRAMD

A three-dimensional pyramid chart.

 
3DRIBBNG

A three-dimensional connected group ribbon chart. 

3DRIBBNS

A three-dimensional connected series ribbon chart. 

3DSPHERE

A three-dimensional bar graph in which all data points are spheres of identical size, hovering at the position that shows their data value. 

3DSTACK

A two-dimensional stack chart with three-dimensional type bars. 

3DSURFCE

A three-dimensional surface chart that graphs all data points as a three-dimensional surface, like a rolling wave. 

3DSURFHC

A three-dimensional honeycomb surface chart that graphs all data points as a three-dimensional surface using a honeycomb effect. 

3DSURFSD

A three-dimensional surface chart with sides that graphs all data points as a three-dimensional surface with solid sides.

Reference: Style Options for Area Graphs

Choose one of the following LOOKGRAPH values to change the style of area graphs:

SET LOOKGRAPH=

Description

 
VAREA

A vertical area graph.

 
VAREASTK

A stacked vertical area graph.

 
VAREAR2

A vertical area graph with two axes.

 
VARESTK2

A stacked vertical area graph with two axes.

 
VARESTKP

A stacked vertical area graph that shows percentages. 

HAREA

A horizontal area graph.

 
HAREAR2

A horizontal area graph with two axes.

 
HAREASTK

A stacked horizontal area graph.

 
HARESTK2

A stacked horizontal area graph with two axes. 

HARESTKP

A stacked horizontal area graph that shows percentages.

Reference: Style Options for Stock Charts

Choose one of the following LOOKGRAPH values to change the style of stock charts:

SET LOOKGRAPH=

Description

 
STOCK

A stock chart.

 
STOCKH

A high-low stock chart. Bars representing higher numeric values are placed behind bars representing lower numeric values. Only the top of the higher bar is visible, clearly illustrating the difference in value.

The most popular application of this type of graph is to represent stock prices. Each bar represents the highest and lowest prices for a given stock on a given day. 

STOCKHB

A bipolar high-low stock chart. Values from different data sets are displayed on separate poles. 

STOCKHD

A dual high-low stock chart. Values from an additional data set are displayed on a second value (Y) axis. 

STOCKHCL

A high-low-close stock chart. The most popular application of this type of graph is to represent stock prices. Each bar represents the highest, lowest, and closing prices for a given stock on a given day. 

STOCKHCB

A bipolar high-low-close stock chart. Values from different data sets are displayed on separate poles.

The most popular application of this type of graph is to represent stock prices. Each bar represents the highest, lowest, and closing prices for a given stock on a given day. 

STOCKHCD

A dual high-low-close stock chart. Values from an additional data set are displayed on a second value (Y) axis. 

STOCKHOC

A high-low-open-close stock chart.

The most popular application of this type of graph is to represent stock prices. Each bar represents the highest, lowest, opening, and closing prices for a given stock on a given day. 

STOCKHOB

A bipolar high-low-open-close stock chart. Values from different data sets are displayed on separate poles. 

STOCKHOD

A dual high-low-open-close stock chart. Values from an additional data set are displayed on a second value (Y) axis. 

STOCKHV

A high-low-volume stock chart.

 
STOCKHOV

A high-low-open-close-volume stock chart.

 
STOCKC

A candle stock chart.

 
STOCKHC

A high-low candle stock chart.

 
STOCKCV

A volume candle stock chart.

 
STOCKHCV

A high-low-volume candle stock chart.

Reference: Style Options for Polar Charts

Choose one of the following LOOKGRAPH values to change the style of polar charts:

SET LOOKGRAPH=

Description

 
POLAR

A polar chart that displays data points on a circle. 

POLAR2

A dual polar chart. Values from an additional data set are displayed on a second value (Y) axis.

Reference: Style Options for Radar Charts

Choose one of the following LOOKGRAPH values to change the style of radar charts:

SET LOOKGRAPH=

Description

 
RADARA

A radar area chart.

 
RADARL

A radar line chart.

 
RADARL2

A dual radar line chart. Values from an additional data set are displayed on a second value (Y) axis.

Reference: Style Options for Bubble Charts

Choose one of the following LOOKGRAPH values to change the style of bubble charts:

SET LOOKGRAPH=

Description

 
BUBBLE

A bubble chart.

 
BUBBLED

A bubble chart with a dual axis.

 
BUBBLEDL

A bubble chart with a dual axis and labels.

 
BUBBLEL

A bubble chart with labels.

Reference: Style Options for Spectral Charts

Choose one of the following LOOKGRAPH values to change the style of spectral charts:

SET LOOKGRAPH=

Description

 
SPECTRAL

A spectral map chart. This is a chart with a row or column matrix of markers that is colored according to the data values.

Reference: Other Graph Types

SET LOOKGRAPH=

Description

 
GANTT

Provides a visual representation of project oriented time critical events. Gantt charts require six display fields and one sort field, in that order. Conditional styling and drill-down are not supported for GANTT charts. 

POSITION

Product position charts provide a visual representation of market share and growth versus revenue and measurement (past, present, future). Product position charts require a set of three display fields. 

VWATERFL

Vertical waterfall graph. 

HWATERFL

Horizontal waterfall graph. 

PARETO

Displays data following Pareto 80:20 rule. Pareto charts require only one display field. 

MULTI3Y
MULTI4Y
MULTI5Y

Stacks charts in order to make it easier to read, analyze and manage them.

Reference: Options for HTML5-Only Chart Types

The following LOOKGRAPH values are valid only when generating an HTML5 chart:

SET LOOKGRAPH

Description

 
BUBBLEMAP

A bubblemap is a chart in which proportionally sized bubbles are displayed on relevant areas of the map. 

CHOROPLETH

a chloropleth is a chart in which areas on a map are shaded or patterned in proportion to the value of the measure being represented, 

MEKKO

A Mekko chart is a variant of a stacked bar chart, in which the width of the bars is adjusted relative to its value in the data set. 

PARABOX

A Parabox (or parallel coordinates chart) is similar to a regular line chart, except that each group in the line chart has a unique and interactive numeric axis. Each line represents one series of data. Each vertical bar represents a numeric axis. You can click and drag along each of the axes to select (filter) the lines that pass through that part of the axis 

STREAM

A streamgraph is a simplified version of a stacked area chart. In a streamgraph, there are no axes or gridlines. The baseline is free, which makes it easier to perceive the thickness of any given layer across the data. 

TAGCLOUD

A tagcloud is a visual representation of frequency. It displays only group labels. The size of each label is proportional to its data value. 

TREEMAP

A treemap chart displays hierarchical data as a set of nested rectangles.

WebFOCUS

Feedback