Experimental Computer Science And Computer Simulation: Understanding The Concepts

What does it mean that a function is eectively calculable?

Definitions of Experimental Computer Science

According to (Denning  2009). Ever since the Feldman Report got published in the year 1979, the term ‘experimental computer science’ has be discussed by computer scientists, with disagreement on the answer. Thus, different meanings and interpretation of the term exist up to date. Back then, the term ‘experimental’ was mainly used to refer to the role of computing in working & industrial environment  

According to (Feitelson 2016), with computer scientists disagreeing on the meaning of the term Experimental Computer Science, several meanings exist. Some conceive it as those large projects for system development. That is, software & computer engineering. Still, there are those who perceive it as all the non-theoretical activities of computer science, particularly the ones concerned with “hands-on” experience, (Feitelson 2016). Never the less, there are those who consider computer science as being a newborn with no actual operational definition available at the moment. Generally, experimental science classifies knowledge that is as a result of observations.

According to (Denning  2009), “experimental computer science is the mathematical modeling of the behavior of computer systems” 

Experimental computer science is both industrial and practical use of computational technology or the ability to deploy computational methods to reinforcement the other sciences.

Having adopted the terminology from other sciences, experimental computer science becomes associated to the Hypothetical-deductive Method, ( Papastergiou 2009). Just like the hypothetical-deductive method, experimental computer science identifies a problem, develops a hypothesis that is predictable and testable, devises a study to test that developed hypothesis before analyzing and evaluating the results. Modifications are made and the process repeated until a theory is developed.

Experiments grant us a chance to better existing systems or come up with new ones.

They come in handy in situations where theoretical research is not of much use. For instance, in real-time and distributed systems.

The scientific image of computer science is improved thus making it easier to acquire funds to further is development and progress, (Neri & Tirronen  2010). 

However Computer .as an experimental science doesn’t go without it critiques.

Computing falls under mathematical science which calls for an axiomatic view

rather than an experimental one. It is an engineering science making use of physical principles to a greater extent that experimentation.

According to (Cheryan et al 2009), Simulation can be defined in different ways. That is, both is a narrow and broad sense. The following are some of the definition:

1. “Simulation is a comprehensive method for studying systems by selecting a model, calculating the output of the algorithm, visualizing and studying the resultant data.”

1. “Simulation, unlike testing, uses a step-by-step method to explore the behaviour of a mathematical model whose (expected) behavior is known.”

• Simulation is any system perceived as having dynamical behavior similar enough to another system such that the former can be studied to learn about the latter.

Thus computer simulation is simulation carried out by a programmed digital computer. 

Heuristic Purposes, Predicting Data and Understanding Data

The mathematical modeling simulates relationships most of the times. Mathematical models provide a solution even to non analytical solution such as numerical solution created by computer. 

Generally, reality, mathematical models and computer simulations are intertwined as they revolve around presenting real life situations in a way that can be visualized.

Computer simulations can be used to visualize a real-life phenomena.

The mathematical model can be used to present a real-life(reality) outcome when provided with necessary parameters. The outcome of this mathematical model can them be visualized using computer simulation.

The reality and its phenomena often can be visualized and analyzed using computer simulation while taking into account physics and mathematical principles / models as the basis. Thus the three are evidently intertwined.

This simulation type employs already known theoretical concepts to express the prospective behaviours of the object being simulated in the same way those events would have unfolded in reality when same variables are provided (Feitelson 2016). Simulation of a free-falling object is an example of such simulation.

At times, this type of simulation is termed as “individual-based” simulation since it concentrates on representation of behaviours of individuals characterized by their own local set of rules. The termites simulation provided in Netlogo is an example of such simulation.

Makes use of models at different scales concurrently to visualize a system.

Analyzes the  risk levels of a given action, its set of attributes and or properties that may affect the actual decision making if the same thing was to be reproduced in the production environment. 

The usage / purpose of computer simulation are classified into three major categorizes namely heuristic purposes, predicting data,  understanding data at hand. 

1. Discover new information
2. Communicate knowledge to others

• Represent information to ourselves 

1. Predict data that we are lacking.
2. Depict the expected behaviour of some real world systems under definite conditions.

• Predict the future.

1. Understand a past event. 

Systemic prediction

Range prediction

Point predictions 

1. Create an understanding of data obtained already.
2. Answer questions on how these events could possibly have occurred;

• How events actually did occur 

Relation Between Computer Simulations & Experiments

A computer simulation can be viewed as an instance of an experiment (Identity thesis).

When a computer simulation is an instance of an experiment, then the computer simulation provides a guarantee to consider their results as being truthful as well. This is termed as Epistemological Dependence. 

Similarities Between Computer Simulations & Experiments

Both Computer Simulations & experiments can be used to depict real-life situations/outcomes.

They both can be used to carry out tests before the actual mass production of a product is deployed. 

Simulation is clearly different from experiment in that its target is separate

from the object being manipulated. Objects used in simulation are virtual while those in an experiment most of the times are physical objects that can be touched and interacted with. 

Relation between Computer Simulations and Experiments

Computer Simulations eliminate the potential dangers in real life that would have emanated when experimenting with the same. For instance, simulation of a car crash with passengers on board. 

Computer Simulations have the ability to be completed in a relatively quicker time as one can accelerate the simulation while experiments may lack such capabilities at times.

Exercise 0.5 (1 POINTS)

A graph is termed as being a strongly connected graph when there exists a directed path from any given node to any other node within that graph. Thus, for each pair q,r of vertices in the graph, there exists a path from ‘q’ to ‘r’ and ‘r’ to ‘q’.

“Letting G be a graph with adjacency matrix A with respect to the ordering v₁ , v₂ , ..., v_n of the vertices of the graph (with directed or undirected edges, that allow multiple edges and loops), the

number of different paths of length r from v_i to v_j , where r is a positive integer, equals the (i, j)^th

entry of A r.” 

The degree of a given node is equivalent to the number of edges linking to that node. It is the summation of the number of ‘in-degree’ and ‘out-degree’ 

A walk of length ‘W’, is a sequence of ‘W’  number of edges, such that the second node of each given edge happens to be the first node in the succeeding edge of that walk.

Picking any node in G and walking randomly without using the same edge more than once, each node is of even degree. When a node is entered, there will be an unused edge you exit through, except at the starting point, at which you can get stuck.                

“The Total graph T (G) of a graph G is one with the following criteria met:”

• “V (T (G)) corresponds to V (G) ∪E(G)”
• “two vertices are adjacent in T (G) if and only if their corresponding elements are either incident or adjacent in G.” 

“A total graph with at least 10 vertices,  with its set of vertices , edges and a path including at least 5 vertices (1,2,6,9,7,(8)).”

1. A setup procedure that cleans the world when needed & then creates one turtle.

1.     setup procedure

to setup        ;; the setup procedure -cleans the world when called, then creates one turtle.   clear-all     ;; clear   crt 1           ;; create-turtles 1 – this creates one turtle. end              ;; terminate the setup procedure

2.     ray procedure

to ray [l n]              ;; a procedure ‘ray’ that accepts two parameters: l and n   pd                         ;; pen down- so as to start performing visible actions   repeat n [              ;; repeat the actions below ‘n’ times.     fd l                     ;; move forward by l steps as parsed to the ray procedure.     bk l                     ;; move backwards by l steps as parsed to the ray procedure     rt (360 / n) ]       ;; right turn by an angle of 360 degree divided by the value of the n variable.   pu                        ;; pen up – actions and movements past this command will not be visible end                        ;; terminate the setup procedure

3.    blade procedure

to blade [l n d]        ;; a procedure blade that accepts three parameters: l, n and d   pd                        ;; pen down - so as to start performing visible actions   ifelse (d = 1)        ;; check whether parameter d is 1 then execute the code ‘[ ray l n ]’   [ ray l n ]             ;; execute the ray procedure when d = 1   [     repeat n [           ;; repeat the actions below ‘n’ times.       fd l                  ;; move forward l steps       blade (l / 3) (n) (d - 1)  ;; recursion       bk l                 ;; move backwards l steps       rt (360 / n)      ;; right turn by and angle defined by 360/n     ]                        ;; closes repeat   ]                          ;; closes ifelse statement for when false (d is not 1)   pu                       ;; pen up – actions and movements past this command will not be visible end                       ;; terminate the setup procedure

The blade procedure makes use of recursion.

It calls itself ‘d’ number of times as long as parameter ‘d’ remains greater than one.

The parameter ‘d’ is parsed to the blade procedure as the third parameter.

When ‘d’ is 1, the ray procedure is called as there is no need to recursively call the blade procedure.

When ‘d’ is greater than 1, the blade procedure is called recursively while reducing ‘d’ by 1;  (d-1)

I noticed that when ‘d’ is greater than ‘n’ , then the output image is made bolder.

Similarities and Differences between Computer Simulations and Experiments

Thus, ‘d’ can casually be viewed as a recursion factor/ pattern repetition factor.

The ‘l’ variable sets the length / number of steps moved in a given direction by the turtle when the pen is down. The greater the value of ‘l’, the larger the output will appear and vice versa.

The parameter ‘l’ can be viewed as a scaling factor to shrink or magnify the output image. 

The ‘n’ variable sets the number of times the repeat command loops when performing the actions specified within its scope. The parameter ‘n’ is also responsible for determining the shape of the outcome as it specifies the Angel made when the turtle makes a right turn” 

Whenever the blade procedure was run with different parameters for l, n or d, different amazing shapes and patterns would be formed. However, setting the values of n and d too high often lead to an image that took a lot of time to complete.

turtles-own [   message1?                  ;; does this specific turtle currently have the message1 (boolean) ?   message2?                  ;; does this specific turtle currently have the message2 (boolean) ?   nomessage?                ;;  does this specific turtle lack both message1 and message2 (boolean) ? ] to setup                        ;; the setup procedure to initialize the world and turtles’ defaults   clear-all                      ;; clear everything from the world and   create-turtles 400 [     ;; set the number of turtles to 400     set message1? false     set message2? true     setxy random-xcor random-ycor ;; randomizer     set size 2                      ;; set turtle size.   ]   ask one-of turtles [        ;; request a turtles in close range to get message1     set message1? true      ;; parse message1 to  one of such turtles in range   ]   ask turtles [     recolor                        ;; recolor turtles that posses  message1     create-links-with n-of links-per-node other turtles   ]   reset-ticks end to go    ;; terminate the simulation when all turtles   if all? turtles [ message1? or message2] [ stop ]     ask turtles [ communicate ]   ask turtles [ recolor ]   tick end to msg2   ifelse ( (count link-neighbors with [ message2? ] ) > 0 )     [set message1? false]     [set message2? true] end to communicate     if any? link-neighbors with [ message1? ]     [ set message1? true       set message2? false   ] end ;; recolor turtles that have message1 to red, those lacking message1 are recolored to blue to recolor  ;; recolor procedure   ifelse message1?     [ set color red ]     [ set color blue ] end

This is a random network. The properties of this random network dictates that, while the number of agents is less than the number of edges between those agents, on of the agents asks another agent to create a link with her until there are no isolated agents left without a message. The message1 and message2 are passed from two agents to the other turtles until all the turtles have a message with them. In my NetLogo simulation, turtles with message1 become red in color while those with message two are recolored to blue. When there are more turtles with message1, the turtles with message1 are open to receive message2 from the fewer carrying the message2 until the agents bearing message1 equates to that of those with message2. 

References

Denning, P.J., 2009. The profession of IT Beyond computational thinking. Communications of the ACM, 52(6), pp.28-30.

Feitelson, D.G., 2016. Experimental computer science: The need for a cultural change. Internet version: https://www. cs. huji. ac. il/~ feit/papers/exp05. Pdf.

Papastergiou, M., 2009. Digital game-based learning in high school computer science education: Impact on educational effectiveness and student motivation. Computers & Education, 52(1), pp.1-12.

Neri, F. and Tirronen, V., 2010. Recent advances in differential evolution: a survey and experimental analysis. Artificial Intelligence Review, 33(1-2), pp.61-106.

Cheryan, S., Plaut, V.C., Davies, P.G. and Steele, C.M., 2009. Ambient belonging: How stereotypical cues impact gender participation in computer science. Journal of personality and social psychology, 97(6), p.1045.

Soukov, S.A., Gorobets, A.V. and Bogdanov, P.B., 2018. Portable Solution for Modeling Compressible Flows on All Existing Hybrid Supercomputers. Mathematical Models and Computer Simulations, 10(2), pp.135-144.

Wilensky, U. and Rand, W., 2015. An introduction to agent-based modeling: modeling natural, social, and engineered complex systems with NetLogo. MIT Press.

Wilensky, U. and Rand, W., 2015. An introduction to agent-based modeling: modeling natural, social, and engineered complex systems with NetLogo. MIT Press.

Thiele, J.C. and Grimm, V., 2010. NetLogo meets R: Linking agent-based models with a toolbox for their analysis. Environmental Modelling & Software, 25(8), pp.972-974.