This work presents the experience of building and
conducting the research and development project "Scientific Visualization
and Visual Analytics" at RTU-MIREA in 2024. In 2024, I was offered to
conduct research and development on the topic "Scientific Visualization
and Visual Analytics" for 3rd-year students of the Department of Higher
Mathematics of the Institute of Artificial Intelligence at RTU-MIREA. In the
past (more than 20 years ago), I prepared and delivered a similar author's
course at the Department of Automation of Scientific Research in Computational
Mathematics and Cybernetics of Moscow State University. Naturally, everything
was interesting for me here - both the current level of students and the
current level of visualization systems, so I certainly agreed. The research and
development course was built and conducted. The main goals and objectives of
the research and development were to provide students with the basics of
processing and visual representation of calculated data and to develop practical
skills in using visualization systems. The main leitmotif of the research and
development was the following: "You will be taught to solve numerical
problems of mathematical modeling. You will solve specific problems. And how
will you show others and yourself that you did anything at all? Such attitudes
were understood by the students. When creating the R&D course, there was a
temptation to use the materials of the previous course from 20 years ago, but
this had to be abandoned, since all the programs for reading and processing
data of that time were implemented in
Fortran,
which
modern students do not study and do not know at all. Therefore, the course was
structured as follows: first, 3 introductory lectures were read on the basics
of scientific visualization and visual analytics, then the students were given
individual tasks on visualization of calculated data, which the students had to
complete while mastering visualization systems. As calculated data, we used
real open calculated data obtained by me and my colleagues at the IPM named
after M.V.
Keldysh Institute of the Russian
Academy of Sciences.
When creating research, the question of choice naturally
arose
visualization programs
for students to master. The most powerful and popular software packages, as 20
years ago, are the software packages -
TecPlot
[1],
Paraview
[2],
VisIt
[3],
Avizo
[5]. Previously, there were also a large number
of open programs with limited functionality (mainly French), which, however,
allow you to successfully visualize 1D
graphs,
2D
distributions
of quantities and
3D
distributions of quantities in volumes. I and
the staff of the Keldysh Institute of Applied Mathematics of the Russian
Academy of Sciences conducted a preliminary analysis of the current situation
with software in this area. The following was revealed. The current situation
has changed radically. Firstly, there are no programs with partial
functionality on the Internet. That is, the thesis "The Internet remembers
everything" in this case is incorrect. Secondly, the "large"
programs listed above have also changed. The Paraview [2] and VisIt [3] programs
have changed significantly in the direction of complicating the processing of
initial data, and it clearly made no sense to give them to students for
training. My colleagues and I spent several evenings solving fairly simple
examples of data for visualization. As a result, the TecPlot [1] complex (which
has a
trial
version) and
gnuplot
[4], which was widely used in the
Keldysh Institute of Applied Mathematics of the Russian Academy of Sciences,
and which there was preliminary information that most students had experience
working with, were chosen as software packages for mastering.
The research work was conducted for 4
groups of students, each group having an average of 25 people. The total number
of students was about 100 people. 30 tasks were prepared for these groups. 26
tasks were focused on using the TecPlot system and 4 tasks were focused on
using the gnuplot system. The initial data set for each task included
previously calculated and published results of solving aerogasdynamic problems
obtained in the Department of Computer Graphics and Machine Vision of the
Keldysh Institute of Applied Mathematics of the Russian Academy of Sciences.
Each student was asked to install a visualization program, study this program,
visualize the results from the initial data and write a short report on the
main functional properties of the program used and the results obtained.
Examples of tasks and the results obtained can be seen in Section 5.
It should be said that the vast majority of
students successfully completed the task, but at the same time, completing the
task required a certain amount of persistence and enthusiasm from them. While
completing the task, they constantly received consultations from me online.
On average, completing the task took 3
iterations. Most of the errors were related to insufficiently thorough study of
the program
manuals.
For example, the rather elementary
construction of cross-sections caused difficulties. If you use the section
construction tool in
TecPlot,
then when changing the coordinate direction of the section, the section itself
is not saved. An example of a consultation tip: "...How to make
perpendicular sections in one drawing? First, build one section, for example
along X, then according to the manual, turn on Data > Extract > Current
Slices, then in the section builder change the direction to perpendicular. Save
the drawing. You can continue changing directions. …».
It should be noted that mastering the
programs caused great enthusiasm from mastering new skills and abilities. Which
sometimes led to some funny incidents. In a burst of enthusiasm, some students
tried to build streamlines for obviously scalar quantities, such as temperature
or pressure. Here it was necessary to draw the students' attention to the
physical nature of the solved aerogasdynamic problem.
It should also be noted that fairly simple
tasks with
gnuplot
caused the
greatest number of difficulties for students. This should be attributed to the
less user-friendly interface of this program. Here, completing tasks required
up to 4 corrections in practice.
This section provides typical examples of
tasks and general results of completing tasks.
Example of a task for
TecPlot
1. TARGET SETTING:
Mastering and practical application of the universal data
visualization program
TecPlot.
2. CONTENT OF PRACTICE:
2.1 Learn: the universal data visualization
program
TecPlot,
write a brief study report including
the main data types and a list of the main functions provided to the user.
2.2 Practical implementation:
Construct
a visualization of the numerical array on the figure based on the data from
the MPrSh.
lay
file as a set of planes. The file contains a three-dimensional data
array representing the critical velocity of transition between different flow
regimes for variations of the Mach, Prandtl, Reynolds and Strouhal numbers.
Present the figure as a set of cross-sections using the program tools. Provide
the figures in the report.
2.3 Get acquainted
with the
possibilities provided by this program.
The results of the task are presented in
Figures 1, 2, 3, which show 2 cross-sections, 3 cross-sections and 4 animation
steps, where one of the cross-sections moves.
Figure 1. A set of two cross-sections (along
the
OX
and
OY axes).
Figure 2. A set of three cross-sections
(along
the OX,
OY,
O
Z axes).
Figure 3. 4 steps of animation for a section
parallel to the
OX axis.
Example of a task for
gnuplot
1.
TARGET SETTING:
Mastering and practical application
of the universal data visualization program
gnuplot.
2.1 Learn: the universal data visualization program
gnuplot.
write a brief
study report including the main data types and a list of the main functions
provided to the user.
2.2 Practical
implementation:
Plot an image of the two-dimensional
array presented in the file
Test2D.dat.
The file contains
a two-dimensional array of data representing the distribution of the
scalar value F in a rectangular area. In reality, these data correspond to the
temperature distribution on a plane. Plot an image of the data as a
three-dimensional surface.
Provide the drawings
in the report.
2.3 Get acquainted
with
the possibilities provided by this program.
3. ADDITIONAL
TASK:
4. ORGANIZATIONAL AND METHODOLOGICAL
INSTRUCTIONS:
use
gnuplot for
data visualization,
take the program version from the site
http://www.gnuplot.info/.
The results of constructing the data in the
form of a three-dimensional surface are shown in Fig. 4.
Figure 4. The model constructed using the
gnuplot
program
In general, the results of the research
work can be assessed as quite successful. The majority (more than 90%) of
students passed the research work with a positive assessment. Such independent
work related to obtaining new specific practical skills aroused interest among
students, which they reflected in many reports on the research work.
This paper describes the experience of
constructing and conducting the research project "Scientific Visualization
and Visual Analytics" at RTU-MIREA in 2024. The research project was
conducted for 3rd-year students of the Department of Higher Mathematics of the
Institute of Artificial Intelligence at RTU-MIREA. The construction and
organization of the research project are described. Examples of tasks and their
implementation are given. This work may be of interest for teaching similar
disciplines in this subject area. Such organization of work serves as a source
for students to acquire specific practical skills that are in demand in science
and industry.
The author
expresses deep gratitude to the management of the Institute of Artificial
Intelligence and the head of the Department of Higher Mathematics, Doctor of
Physical and Mathematical Sciences Albina Viktorovna Shatina for their
assistance and support in creating and conducting research, as well as junior
researchers of the Keldysh Institute of Applied Mathematics of the Russian
Academy of Sciences Artem Evgenievich Kuvshinnikov and Nicole Aleksandrovna
Bondareva.
1. Tecplot Visualization and Analysis Tools for CFD Post- pricessing https://tecplot.com/
2. Paraview https://www.paraview.org/
3. VisIt visualization system https://hpc.llnl.gov/software/visualization-software/visit
4. gnuplot homepage http://www.gnuplot.info/
5. Avizo 3D https://www.thermofisher.com/software-em-3d-vis/customerportal/download-center/amira-avizo-3d-installers/
