Multi Series  Graph

Graph can be used to display complex stories. A single graph can sometimes be used to display complex stories elegantly but frequently it will not do . So we will cover two useful way to show this complex stories .

1)Combine multiple unit of measure

we use single graph to display multiple quantitative variable when they all use same unit of measure .but if the value of variable differ by large amount then problem occurs , cause low values to look relatively flat.

Fig…11.1

When you wish to display two units of measure for the purposes of the comparison, the best way to avoid confusion is usually to use two separate graph rather than a single graph with two Quantitative scales. Unless you are certain that your readers are comfortable with dual-scaled graphs, it is best to avoid them.

 Fig…11.2, .3(Focus on the intersection)

2)Combining multiple graph in a series

In 2D graph, we can squeeze the data to a certain limit but we have a solution to display multiple measures(more than 2) in 2D : multiple graphs arranged together as a series(Trellis chart -A series of graphics , showing  the same combination of variables , indexed by changes in another variable.)

Fig 11.4  (Focus on adding more than two series).

When you need to add one more variable (i.e., another set of  categorical subdivision) to a graph , but you have already used all the practical means to visually encode in it, you can do so by constructing a series of related graphs , in which each graph in the series displays a difference instance of the added variable. Avoid 3D graph and involves multiple graphs arranged in a series so that it will be easily

compared .

Notes : As the number of graph grows , the trick is to reduce their individual size enough to allow them to be seen together . You can arrange the graphs horizontally and vertically , or in both directions to produce a matrix arranged in  multiple columns and rows, redundant labels should be eliminated .

Best practices

Consistency            

Consistency is required for the comparison. Knowledge of first graph in a series guides you through rest. Consistency in visual includes:

• Aspect ratio of the axes
• Color used to encode data .
• Font etc.,

Graphs in a series of small multiples should be consistently designed with only one exception : text used for the labels , titles , or legends does not need to appear redundancy in each graph .

Pay particular attention to the scales along both axes  fig..(11.6)

Categorical scale also remains consistent with the same items in the same order and the same full list of items even when a value is zero or null. (fig 11.7) 

Arrangement 

Before arrangement we should answer this question :  which items do you want to make easiest for your user to compare ? And then arrange the graphs in a series of small multiples in the way that makes it as easy as possible to focus on and compare the values that are the most relevant to your reader`s interests.

Sequence

If the index variable has an intrinsic order , you should sequence the graphs in this order unless you wish to display a ranking relation .Otherwise , rank the graphs in order based on a quantitative measure associated  with the index variable.

Rule and grid lines

Only use rules or grid lines between graphs in a series when either of these two conditions exists:-

• The graph must be positioned so closely together that white space alone cannot be adequately delineate them .
• The graph are arranged in a matrix and are positioned so closely together that white space alone cannot adequately   direct your readers to scan either across or down in the manner you intend. 

Graph Design

Components	Practices
Points	When sets of points cannot be clearly distinguished , correctly by: Enlarging the points Selecting objects that are more visually distinct . When point overlap such that some are obscured , correct by: Enlarging the graph and/or reducing the size of the points. Removing the fill colors.
Bars	Use horizontal bars when their categorical labels bars won't fit side by side. Never use horizontal bar for the time series value . Proximity Set the width of white space separating bars that are labeled along the axis equal to the width of the bars, plus or minus 50%. Do not include white space between bars that are differentiated by a legend. Do not overlap bars . Fill Avoid the use of fill pattern . Use fill colors that are clearly distinct . Use fill colors that are fairly balanced in intensity for the data sets that are equal in importance. Use fill colors that are more intense than the others to highlight particular things . Only place borders around bars when one of the two following conditions  exists : The fill color of the bars is not distinct against its background , in which case you can use a subtle border(e.g. grey). You wish to highlight one or more bars compared to the rest. Always start bars at a baseline of zero.
Lines	Distinguish lines using different hues whenever possible. Include points on lines only when values for the same point in time on different lines must be precisely  compared .
Boxes	Follow the principles for the bar design , except when box plots are connected with a line to show change through time , which might require greater distance between the boxes.
Combination	Use boxes and lines for the distribution through time . Use bars and lines in the form of the Pareto charts for the featuring the contribution of the largest portions of the whole . Use bar and points for the uncluttered comparisons.
Trends line	In most cases , use moving averages rather than straight lines of best fit to show the overall nature of change through time . Only use linear line in a scatter plot when the shape of the data is linear rather than curved .
Reference lines	Use references  line to mark meaningful threshold and regions, especially for the measures of the norm.
Annotations	Use text to feature and comment on the values directly when doing so is important to the story .
Log scales	Use log scales to reduce the visual difference between quantitative  data sets with significantly different values so they can be clearly displayed together. Use log scales to compare differences in a value as percentages .
Tick marks	Mute tick marks in comparison to the data objects . Use tick marks with quantitative scales but not with categorical scales , except in line graphs  when slightly more precision is needed. Aim for a balance between including so many tick marks that the scale looks cluttered and using so few that your readers have difficulty determining the values of data objects that fall between them . Avoid using tick marks to denote values at odd interval.
Grid Lines	Thin , light grid lines may be used in graphs for the following purposes : Ease look-up of values. Ease comparison of values Ease perception and comparison of localized patterns.
legends	Use legends for the categorical labels when the labels are not associated with a categorical scale along the an axis  and cannot be directly associated with the data objects . Place legends as close as possible to objects they label without interfering with other data . Render legends less prominent than the data objects they label. Use borders around legends only when necessary to separate legends from other information.
Axes	Don’t manipulate the aspect ratio to distort perception of the values
Data Regions	Keep the background clean and light.

General Design for Communication

With the basic understanding of visual perception ,we can build a set of visual design principles, beginning with those that apply to table and graphs. Our primary visual design objectives will be to present content  to the readers in a manner that highlights what`s important , arranges it for clarity , and leads them through it in the sequence that tells the 

Story best.

We use visual designer to communicate .There are stories in the numbers that will be perceived and acted upon or will go unnoticed and be ignored , depending on our knowledge of visual design and our ability to apply that knowledge to the important task of communication .

Communication -oriented  design support two fundamental objectives:

• Highlight
• Organize

Highlight

We Highlight  important information to give it a voice that comes through loudly and clearly , without distraction . We organize information to lead readers through it in a manner that promotes optimal understanding and use .

Edwards Tufte: "Above all else show the data "

Data-Ink Ratio : It is an amount of ink that presents information compare to the total amount of ink .

The objective is to reduce the non-data ink to more than what `s necessary to make the data ink understandable.

We highlight data through a design process that involves activities of two types :

1. Reducing the non-data ink .
2. Enhancing the data ink 

Reducing the non-data ink

The process of reducing the non-data ink involve two steps:

Subtract unnecessary non-data ink = Resist the temptation to keep the things just because they are cute or because you worked hard to create them .You must carefully select the content that is essential to the message and trim all else away .

De-emphasize and regularize the remaining non-data ink = Once non data ink is reduced , you should push the non-data ink that remains far enough into the background to enable the data to stand out clearly in the foreground .this can be reduced by visual prominence of the non-ink data components. Non-data items consistent with their supporting role, should stand out just enough from the background to serve their purpose but not so much that they drew attention to themselves.

Enhancing the data ink

You can enhance the data ink through a process that consists of two steps:

Subtract unnecessary data ink = Not all information are equally important .Don`t remove anything that`s important , but be sure to remove all that is peripheral to the interests and purposes of your readers .

Emphasize the most important data ink= Each step in the process of highlighting data results in simplicity. In the Communication of Quantitative information , Simplicity of design is the essence of elegance . Your message might be complex , but its design - the form in which you present it - should be so simple that to your readers it is nearly invisible.

Organize

When your readers looks at a page or screen of information . They immediately begin to organize what they see in an effort  to make sense of it .As a designer of communication it is our job to organize the information for them in a manner  that tells the story as clearly as possible . The page and screen that serves as your medium of communication will often contains more than  a single table or graph . Your message ,may require multiple tables, multiple graphs , or a combination of both , along with the additional text in the form of annotations , sentences or even whole paragraphs. When you arrange the information on the page , you must consciously do so to tell a story . What should I say first ? What should I save for the last ? What should I emphasize more than the rest ? The answer to these question take on the form of visual attributes designed to accomplish the following :

• Group (i.e., segment information into meaningful sections).The Gestalt principles of visual perception reveal a number of techniques that can be used to group information meaningfully. The simplest approach - proximity - is often the best . Table primarily use Gestalt principles of proximity and continuity to organize the different categories into columns  and rows .Graphs use many techniques , such as principle of similarity  and connection.
• Prioritize (i.e., rank information by importance).
• Sequence (i.e., Provide direction for the order in which information should be read )  

Table and its types

Table should be structured to  suit the nature of the information they mean to display .so purpose of this section is to identifying what table can be used to display , followed by how they can be structured visually.

Information that we display in table always exhibits a specific relationship between individual values . These are the possible relationships which guide us to design table effectively . 

Quantitative to Categorical Relationships

		Structure type
Relationship	Primary function	Unidirectional	Bidirectional
Between a single set of quantitative value and a single set of Categorical items.	Look up	yes	Not applicable because there is only one set of categorical items
Between a single set of quantitative value and the intersection of multiple Categories.	Look up	Yes sometimes this structure is preferred because of the convention	Yes. structure save the space
Between a single set of quantitative values and the intersection of multiple hierarchical Categories.	Look up	Yes this structure can clearly display the hierarchical relationship by placing the separate levels of the hierarchy side by side in adjacent columns.	Yes. However, this structure does not display the hierarchy as clearly if its separate levels are split between the columns and rows.

Quantitative to Quantitative Relationships

		Structure type
Relationship	Primary function	Unidirectional	Bidirectional
Among a single set of quantitative values associated with the multiple Categorical items.	Comparison	Yes	Yes. This structure works especially well because the Quantitative values are arranged closely together for easy comparison.
Among distinct set of quantitative values associated with the single  Categorical items.	Comparison	Yes	Yes. However this structure tends to get messy as you add multiple sets of Quantitative values .

Variations in table design

Unidirectional - categorical items are laid out in one direction only (i.e. either across columns and down the rows).  e.g.

Bidirectional - categorical items are laid out in both directions.  e.g.

Comparison on  same table  (diagram)

Table Vs. Graph

Tables and graphs are two primary means to structure and communicate Quantitative information. This sections help you to select tables and graphs for your particular purpose .

When the Quantitative information  which you want to convey consists only of a single number and two , written language is an effective means to communicate (form of sentence..).e.g.

As describe in previous section about Quantitative  and categorical value , it plays an important role  while taking any design decision .but for more clarification , number do not always represent the Quantitative value but use as to label thing and not having any Quantitative meaning. e.g. patient Id, number that identify the year (2010).These are the example that express categorical data . So before categorized your data ask one Question from yourself that "Would it make sense to add these number up , or to perform any mathematical operation on them".

So to select the appropriate medium of communication , we must understand the needs of our audience as well as  the purposes for which the various forms of display can be used .

Table definition :

A table is a structure for organizing and displaying information ; a table exhibits following characteristics :

• Information is arranged in column and rows.
• Information is encoded as text (including numbers and words).

Whenever you have more than one set of values, and relationship exist between values in the separate set  , you may use a table to align the related values by placing them in the same row or column .

A table works best when:

• The display will be used to look up individual values.
• It will be used to compare individual values but not entire series of values to one another.
• Precise values are required .
• The Quantitative information to be communicated involves more than one unit measure .
• Both summary and detail values are included.

Graphs definition:

Graph exhibits following characteristics:-

• Value are displayed within in an area delineated by one or more axes.
• Values are encoded as visual objects positioned in relation to the axes.
• Axes provide scales (quantitative and categorical ) that are used to label and assign values to the visual objects.

So Graph is a visual display of the quantitative information .whereas tables encode Quantitative values as text .the visual nature of graphs endows them with their unique power to reveal patterns of various  types, including changes , differences, similarities, and exception .

A Graph works best when:

• The message is contained in the shape of the values (e.g. , pattern, trends, and exceptions).
• The display will be used to reveal relationships among whole set of values

Understanding of Quantitative information

Quantitative information  forms the core of what organization must know to operate effectively .It give us the stories  from which we can get the involved number and its relationship. In other words , it simply belong to the class of information , that communicate the quantity of something .It is not boring and interesting it is up to us to give a clear and unhindered voice to that information and its story , using language that is easily understood by your audience .

Quantitative  stories include two types of data :

• Quantitative  ( Quantitative values measures things  e.g. no . Of patient , count of activities)
• Categorical    (Categories divide information into useful groups. e.g. activity name , activity categories.)

Quantitative stories always  feature relationships.

These relationships involve either:

• Simple association between Quantitative values and corresponding Categorical items  and
• More complex association among multiple sets of Quantitative values

Categorical items exhibit  four type of relationships:

1. Nominal
2. Ordinal
3. Interval
4. Hierarchical

Quantitative values exhibit  three type of relationships:

1. Ranking
2. Ratio
3. Correlation

Measure of money

• When comparison of monetary value are expressed across time , adjusting the value to account for the inflation produces the most accurate results .

• When reporting monetary values that combine multiple currencies , you must first convert them into a common currency .  

Introduction

"InSightive 1.0" product which talks about data, enable us to make informed decisions. The way InSightive helps to determine success and failure is always based primarily on the numbers. We derive great value from the stories that numbers tell , yet we rarely consider the significant of how we present them . So main purpose of this document is to learn some design practices which make our InSightive more effective . 

Since number cannot speak for themselves .Number have an important story to tell , and it is up to us to help them to tell it .Time is wasted struggling to understand the meaning  and significance of numbers - time that could be better spent doing something about them . Some best practices can solve this problem. Table and graph are usually the best means to Communicate Quantitative information() .we should think critically and creatively about the design of the table and graphs. There is no Single correct solution to the task at hand but we have one which give us an optimal way to express our data. With the advent of various tool everyone can produce reports of Quantitative information in the form of table and graphs But we have lost sight of the real purpose of Quantitative  displays : to provide the reader with important , meaningful and useful insight. To communicate Quantitative information effectively requires an understanding of the number and the ability to display their message for the accurate and efficient interpretation by the user .

 As the presenter of Quantitative Information , it is our responsibility to do more than sift through the information and pass it on ;we must help our users gain the insight contained therein . We must design the message in a way that leads the readers on  a journey of discovery , making sure that what's important is clearly seen and understand . The right numbers have important stories to tell . They rely on us to find those stories , understand them and then tell them to other in a way that is clear , accurate , and compelling.

Tables and graph of Quantitative  business information can be used for  four purposes:

• Analyzing
• Communicating
• Monitoring
• Planning

When you use tables and graphs to discover the message in data , you are performing analysis .

When you use them to pass a message on to others , your purpose is communication .

When you use them to track information about performance , such as the gain /loss or quality of treatment , you are engaged in monitoring .

When you use them to predict and prepare for the future , you are planning .

Although tables and graphs are both vehicles for the preparing information visually , the role that visual perception plays in reading and interpreting the information presented differs significantly for the tables and graphs. Graphs are perceived almost by our visual system while table interact primarily with our verbal system .with graph we see the pattern and relationship as a whole but in tables information stored sequentially , reading down columns or across rows , comparing this number to that number, one pair at a time .So neither method of Quantitative display is better than other, but each is better than the other for the particular communication task , and both play a vital role.