MDim eXplorer for multi-dimensional data
MDim eXplorer is an interactive data exploration application using Adobe/Flash for exploring, presenting and collaborating multi-dimensional scientific or statistical data for a single year or animated time series. A map is NOT required as for the other eXplorer applications. MDim eXplorer based on dynamic linked visual user interfaces facilitates methods for exploring large multivariate data that can uncover hidden structures and relations and let the analyst share and collaborate her findings through storytelling to a broader audience. MDim eXplorer can be used with any column EXCEL data. See figure below.
Load NEW MDim eXplorer at: https://mitweb.itn.liu.se/geovis/eXplorer/mDim/#story=0
Figure: The Human Development Index (HDI) for countries represents a summary measure of average achievement in key dimensions of human development: a long and healthy life, being knowledgeable and have a decent standard of living. The average "HDI Level" is an example of using also categorical data. Here represented by the HDI (Human Development Index): "Very Low, Low, High, Very High" (See column C in the EXCEL sheet below). The treemap, scatter plot and bar chart are here coloured based on the indicator "Life expectancy" .
The term “categorical data” here refers to thematic attributes assuming nominal or ordinal values, which denote possible states of an object. For example, states of a person may be ’at home’, ’at work’, etc. The categorical data indicator applied here in our context stems from statistical data country observations (e.g., health dimension of a country) but also derived from other data (education, standard-of-living etc.). The states can change over time, i.e., a country can move from one HDI category to another.
The treemap visualizes hierarchical data World-Continent-Country by using nested rectangles, each one covering an area proportional to here the indicator "Total Population" in column D.
The value used as variable so that larger Total Population result in larger and more prominent rectangles. Each rectangle represents a country and the size represents Total Population and the colour the HDI Level.
In the extra scatter plot, the circles are coloured based on the HDI value. See colour legend inside the scatter plot.
Column A represent the name of the data item, Column B and C are categorical data. Column D-J are numerical data items. More details about categorical versus numerical data is available below.
Available Visual Analytics and InfoVis Methods
MDim eXplorer embeds a collection of common visual analytics and information visualization methods (see figure below) including horizontal and vertical bar graph with fish eye technique, parallel axes, scatter plot, scatter matrix, table lens, treemap, distribution plot, data table, colour legends and scales. eXplorer connects these visual and data analysis components to each other and data providers can load data from various sources. Interactive features that support a spatial and temporal analytical reasoning process are exposed such as tooltips, brushing, highlight, data zoom, visual inquiry, conditioned statistics filter mechanisms that can discover outliers and methods supporting dynamically linked multiple views.
Car Data Demonstrator
Start MDim eXplorer https://mitweb.itn.liu.se/geovis/eXplorer/mDim/#story=0
MDim is already loaded with some demo story examples. You can, however, load your own data - analyse, explore and make up your own story.
The EXCEL sheet is structured in a simple fashion and is auto detectable by MDIM eXplorer. This means that the MDIM Data Management can auto detect any numeric or categorical indicators and then lets you choose in the Data Management what indicators to use in the data set. See picture below.
The indicator "Country" is considered to be a categorical indicator since the values in this indicator are not numeric values but instead indicates a certain grouping or descriptive attribute. Categorical indicators will be visualized different from numeric indicators and MDIM will automatically detect these indicators and separate them when loading the file. The indicator "Cylinders" consists of numeric values and can be visualized like any other numeric indicator - but could also work well as a categorical indicator since the values are grouped into only a few classes (4,5,6,8 cylinders). This could be done later in the Data Management step by dragging the indicator from the numeric list to the categorical list.
As a simple example based on this classic information visualization data set, the ranking of cars with “miles per gallon” may quite effectively be represented using our fish eye bar chart (see figure above and below) where the height of the bar represents the numerical variable consumption "miles per gallon". Using the same kind of representation for the categorical variable “Country” would, on the other hand, not be as useful here since there does not exist any meaningful relationship between a country name and the height of a bar. However, we can use colour to show the similarity of two variables using colour for variable “Country” and the numerical value "Miles per Gallon".
With this data set you can now start your analysis and if you prefer also begin creating a new Story. If your analysis results in any interesting discoveries then you can open the Story Editor and start a Story that can be saved and shared with your colleagues. The chapter about Storytelling provides more details how to create snapshots. You can save the any current configuration of all MDIMs components to a snapshot and a story text can be added to create a new snapshot link in your story.
When you click on the "second view", the following new story appear with two new InfoVis methods "Table lens" and "Distribution Plot".
Figure: The distribution plot shows the car models organized to each country. The variable "Miles Per Gallon" was selected. We discover that Japan has the car model with best performance "Miles per gallon" while the USA has the worst models.
Swedish Sales Management Data
Figure: An invented Swedish sales management data is stored in EXCEL data and can be imported direct into MDim eXplorer. The data has five categorical data columns (Sales person, Sales Manager, District, Gender and Age group; and four numerical data columns). The sales data is represented in the interactive figure below with a treemap, scatter plot and fish eye bar chart. Red colour indicates negative sales results (column H). The treemap is divided into "Districts" (Column C) and the "Sales person name" (Column A) is shown. The size of the rectangles are related to "Total Sales" Column F. The combination of the treemap and bar chart provides a deep insight into the sales management data.
Figure: The interactive image below (a "Vislet" (see Storytelling - http://ncva.itn.liu.se/storytelling) was produced in HTML5) analyses the Swedish sales management data above. The Vislet was created as an story example produced with MDIM eXplorer and then inserted into this web site.
Figure: Example of an interactive Vislet analysing Swedish sales management data
Last updated: 2017-01-08