20
Presentation 3: Various ways to increase the information density in Planning/Management data, while making it more “understandable”. Erik Danielsson +47 97 06 85 42 [email protected] Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Erik Danielsson +47 97 06 85 42 [email protected] Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Embed Size (px)

Citation preview

Page 1: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Presentation 3:

Various ways to increase the information density in Planning/Management data, while making it more “understandable”.

Erik Danielsson+47 97 06 85 42

[email protected]

Product Responsible - ProteusQuantum Solutions AS

30 minutes - October 2009

Page 2: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusVarious ways to increase the information density in Planning/Management data, while making it more “understandable”.

Page 3: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus

1. Introduction

2. What is “Information Density” anyway?

3. Techniques to increase information density

4. Summary

Page 4: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusIntroduction – The shortest mail correspondence in history

Page 5: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus

1. Introduction

2. What is “Information Density” anyway?

3. Techniques to increase information density

4. Summary

Page 6: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusWhat is information density (ID) anyway??

The Proteus approach to pin down a definition (first some terminology): 1. The raw (non manipulated) data set – No properties have been highlighted.2. The “enhanced” data set - Property X has been highlighted.3. When talking about ID one always relates this to a particular property, X – one

particular set of glasses to look at data with.4. Let’s call the time it takes for the brain to determine Property X for a set of raw

data, T1.5. Let’s call the time it takes for the same brain to determine Property X for the

enhanced set, T26. Finally the definition of ID:

ID = T1/T2Example: If it takes 10 seconds to find even numbers in a 3x3 matrix of numbers, and 1 second if they are colored red, then ID = 10/1 = 10.

Conclusion so far: The higher the ID, the less we have to strain the brain...

Page 7: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusExamples of increased information density (ID)

Raw data: ID = 1

Even numbers: ID = 20 Prime Numbers: ID = 50 Large Prime Numbers: ID = 100 000

Page 8: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusUsing coloring to increase ID

Raw data: ID = 1

Magnitude (3 bands): ID = X Magnitude (10 bands): ID > X? Magnitude as spectrum: ID > X?

Magnitude mapped to bar width Magnitude mapped to 3 arrows

Page 9: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusIncreased ID for more than one property

Raw data: ID = 1

Magnitude mapped to fontsize Magnitude mapped to fontsize + “Primeness” to red background

Magnitude mapped to fontsize and color bands

Page 10: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusOther common ways you increase ID... (without knowing it?)

These transformations on raw data constitute the 6 primitives for improving ID:

Conditional formatting – highlight a property of interest Filtering – How long would it take to find all values between 10 and 20 in a set of 1000 numbers? Showing list of unique values – a prerequisite for “drill downs” Sorting - How long will it take you to find the 5 smallest values in 1000 numbers? Aggregating - (how long would it take you to add 1000 numbers?) Graphing – showing numbers as spatial entities.

• Grouping (same as showing list of unique and sorted values but with expandable details)

- Show how the Gantt-Grid in Proteus contains the first 5 of these.

Page 11: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus Conditional Formatting

Page 12: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus Filtering & Showing list of unique values

Page 13: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus Sorting and Aggregating (& Grouping)

Page 14: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus Using the 6 primitives in conjunction with spatial arrangements...

Lets recap the 6 primitives for improving ID:

Conditional formatting – highlight a property of interest Filtering – How long would it take to find all values between 10 and 20 in a set of 1000 numbers? Showing list of unique values – a prerequisite for “drill downs” Sorting - How long will it take you to find the 5 smallest values in 1000 numbers? Aggregating - (how long would it take you to add 1000 numbers?) Graphing – showing numbers as spatial entities.

When using these primitives in conjunction with spatial arrangements, we can really get some remarkable ID boosts!We will look at 2 important examples...

Page 15: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusPivoting

Pivoting:Each of the dimensions (X, Y) show unique, sorted values, and each intersecting cell represents a filtered and aggregated value.

Adding graphing improves ID even further!

Page 16: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusCompletion Report

The Completion Report:Each planning object is located in a grid where its horizontal location (column) says when in time it should be complete (a list sorted in time), the values within column are sorted in value, and conditional formatting of cell’s value further enhances ID. In addition, there are summary rows for each conditional format showing aggregated number of “hits” for it.

Tooltips can help giving any detail info of relevance (including hierarchical “drill down”).

Page 17: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusCompletion Report – how its set up...

The Proteus Configurator lets Administrator set up Completion Reports with any number of formatting criteria.

Page 18: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusMapping important “facts” to visual elements

“Instant experience” can in a sense be given by coding important facts to visual elements in Gantt Chart.The Gantt chart at left use the following mappings to increase ID:1. A blue bar indicates the time-

range of all child elements. It is hatched if the children are outside of parent range.

2. A red frame is shown if the object is more than 5 days behind schedule, and it indicates where it should be located in time.

3. The objects having more than 10 000 planned hours are red (as starting color gradient).

4. The transparency of starting gradient color is higher the less hours an object has – i.e object with more hours are “more visible”.

Page 19: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

Proteus

1. Introduction

2. What is “Information Density” anyway?

3. Techniques to increase information density

4. Summary

Page 20: Erik Danielsson +47 97 06 85 42 erik@quantumsolutions.no Product Responsible - Proteus Quantum Solutions AS 30 minutes - October 2009

ProteusSummary conclusion:

The various techniques of increasing Information Density is to a large degree an “untapped grail” in the professional community...

By letting professional people (planners, PM, CC) SEE what the data means instead of having to strain their brains to deduce this meaning, a tremendous amount of time can be saved, and mental energy used creatively on other things...

The Proteus challenge:If your project team sits down and brainstorms up N ways that your project could increase the ID in its data, Proteus should be able to accommodate all of these quite easily...!