Three suggested Projects for a Visualization Course
Project One: Data Sets "IRAS"
With NASA sponsored mission IRAS (Infrared Astronomical Satellite) the entire sky was mapped in 1983 at four far-infrared wavelengths. Following are four data arrays (sky-flux plate 75, HCON 3) representing band 1 (12 micrometers), band 2 (25 micrometers), band 3 (60 micrometers) and band 4 (100 micrometers). The images show the Milky Way through the center and Lambda Orionis at the lower right. Preprocessing of images is a courtesy of the Infrared Processing and Analysis Center. Destriping and flattening by Gitta Domik..
Each of the four data arrays has the size of 512 * 512 pixels * 1 byte.
| p077h3b1.byt (band 1)
| p077h3b2.byt (band 2)
|
| p077h3b3.byt (band 3)
| p077h3b4.byt (band 4)
|
Download all four data-files in IRAS.zip:
Assumed task: Astronomers want to find "similar" objects in the images. Can you show "similarity" in measured data?
Possible solution: Merge three images in an RGB or HLS transformation (e.g. map band 4 to red, band 3 to green, band 1 to blue). Similar objects obtain the same color. Examples.
Project Two: Data Set "Leaves"
The following table shows (ficticious) measurement results of the growth of leaves from three different types of trees (maple, aspen and pear) at different growing periods:
| Type of Tree | Age of Leaf | Length of Leaf | Width of Leaf |
| Maple | 3 weeks | 2.2 cm | 1.8 cm |
| Maple | 2 months | 4.6 cm | 5.5 cm |
| Maple | 4 months | 8.8 cm | 10.0 cm |
| Aspen | 3 weeks | 1.2 cm | 1.2 cm |
| Aspen | 2 months | 3.6 cm | 3.6 cm |
| Aspen | 4 months | 7.5 cm | 7.5 cm |
| Pear-Tree | 3 weeks | 3.2 cm | 1.2 cm |
| Pear-Tree | 2 months | 7.0 cm | 2.5 cm |
| Pear-Tree | 4 months | 11.0 cm | 4.0 cm |
Task: Exploration of data - let the students decide on specific tasks.
Hint: A rigorous data model (data characteristics, e.g. what variables are quantitative, ordinal, nominal) helps to avoid errors.
No examples provided.
Project Three: Data Set "Flow"
The data set is a snap shot of water flowing through a channel. Winds acting upon the (open) surface of the water create turbulences inside the water. Movements of water particles (caused by the winds) were calculated in a supercomputing class by Lloyd Fosdick, University of Colorado, in 1992. File "field2.irreg" contains data describing the particle movement in a 2d slice perpendicular to the length of the channel. Data is given for a regular 82 x 82 grid in the following format: starting position (x,y,z) and relative movement (u,v,w).
Detailed format of field2.irreg:
| Integer value | Number of spatial dimensions ("3") |
| Integer value | Number of spatial dimensions ("3") |
| 3 Integer values | Integer values describing the extension in each dimension: (Dim_x,Dim_y,Dim_z)=("82, 82, 1") |
| Integer value | Number of spatial dimensions ("3") |
| Real array of dimension (6,82,82,1) | First value "6" describes the number of data entries for each vector, followed by "x,y,z,u,v,w" as described above. |
Suggested tasks:
(a) Give a visual overview of the data!
(b) Show Symmetry in the flow!
(c) Where are the quickest / slowest movements (whithout losing sight for the whole flow)?
Discuss advantages/disadvantages/errors of the provided solutions.
Please send back any suggestions and/or complaints to domik[at]uni-paderborn.de