Stretch tasks Decomposition and Abstraction

Round 1: Quiz Blitz

1. Breaking a large problem into smaller sub problems Each sub problem can then be solved and tested on its own.
2. Removing or hiding unnecessary detail What is left is only what is needed to solve the problem.
3. Decomposition Splitting into parts is decomposition.
4. abstraction Building colours never change which route is shortest.
5. False Compression keeps the information in fewer bits. Abstraction throws unneeded detail away.
6. Several programmers can work on different parts at once; Each part can be tested in isolation Decomposition also makes parts reusable and maintenance cheaper.
7. The rules of chess do not depend on it Abstraction keeps only the piece type, its square and whose turn it is.
8. Abstraction The sub program hides how, and shows only what.
9. False It is less accurate on purpose, in exchange for being simpler and faster.
10. Whether each pupil is present The register only has to record attendance.
11. The fault is narrowed down to the calculation part Testing each part separately points straight at the faulty one.
12. model A model is built using abstraction.

Round 2: Quick Fire

1. • Each sub problem can be tested in isolation
   • So a fault can be narrowed down to one part
   Model answer

   Each sub problem is small enough to be tested in isolation. If the finished program is wrong, testing each part separately narrows the fault down to the one part that is failing.

2. • Abstraction keeps only the detail needed to solve the problem
   • Kept: junctions, road lengths, cycle lanes, accident history
   • Removed: shop names, building colours
   Model answer

   Abstraction means keeping only the detail needed to solve the problem. The app would keep the junctions, road lengths, cycle lanes and accident history of each road, because they affect which route is safest. It would leave out shop names and building colours, because they never change the answer.

3. • Decomposition splits a problem into smaller parts
   • Abstraction removes unnecessary detail
   Model answer

   Decomposition breaks a problem into smaller sub problems that can each be solved separately. Abstraction removes the detail that is not needed, so only what matters for solving the problem is left.

4. • Full physics would need too much processing, so the game would be too slow
   • Players only need a believable feel, so the accuracy is unnecessary
   Model answer

   Simulating real tyre chemistry would need far more processing and memory, so the game would be too slow to run smoothly. Players only need a believable feel, so the extra accuracy is unnecessary detail.

Round 3: Big Think

1. • Break the system into smaller sub problems
   • A sub problem such as the listings and showing times
   • A sub problem such as a seat map of available seats
   • A sub problem such as payment and confirmation
   • Different programmers can write parts at the same time
   • Each part can be tested separately, so faults are easier to find
   • Parts can be reused
   • Maintenance is easier, because a change touches only one part
   Model answer

   Decomposition would break the booking system into smaller sub problems: showing the listings and showing times, displaying a seat map of available seats, taking payment by card, and sending a confirmation with the tickets. Each sub problem is small enough to solve on its own, and different programmers could write them at the same time. Each part can be tested separately, so a fault in the total is quickly traced to the payment part. Parts can be reused, such as the payment module in the chain's snack shop app, and maintenance is easier, because changing how seats are shown means editing only that part.

2. • Keep what affects flow: cars arriving, lanes, speeds, queues
   • Remove details such as car colour and drivers' names
   • Because the removed detail does not affect the result
   • It is simpler to design and program
   • It needs less memory and processing, so many timings can be tried quickly
   • But it is less accurate than the real world
   • Removing a detail that matters, such as pedestrians, gives wrong results
   • A conclusion, such as checking the model against real data
   Model answer

   Abstraction would keep only the details that affect traffic flow: the number of cars arriving per minute in each lane, their speeds, the queues and the light sequence. It would remove details such as the colour and make of each car and the drivers' names, because they do not affect how long the queues get. This makes the model simpler and quicker to design and program, and it needs less memory and processing, so it runs faster and many timings can be tried quickly. However, the model is less accurate than the real world, and if a detail that matters is removed, such as pedestrians crossing or buses stopping, its results could differ from reality. On balance, abstraction is worth it, provided the results are checked against real data from the junction.

This or That

1. Decomposition Splitting into parts is decomposition. Abstraction leaves detail out.
2. Abstraction A passenger only needs the stations and the connections between them.
3. No: easier to test, not shorter Decomposition makes code easier to test, reuse and maintain. It rarely makes it shorter.
4. Speed limits Speed limits change which route is fastest. Hedges do not.
5. Reuse A decomposed part can be reused in every system that needs it.
6. Unnecessary for this problem Abstraction is a judgement about which details matter for this particular problem.
7. No: detail is left out on purpose A model trades accuracy for simplicity and speed.
8. Naming the details kept and removed Application marks need nouns from the scenario, not a definition that fits anything.

Memory Match

• A 3 Decomposition: Splitting a big problem into smaller parts
• B 4 Abstraction: Removing unnecessary detail
• C 6 Sub problem: One small part of a larger problem
• D 1 Model: A simplified version of a real system
• E 2 Subprogram: Named code that hides how a task is done
• F 5 Algorithm: Precise steps that solve a problem

Sort It

1. Decomposition Giving the search code and the fines code to different programmers Decomposition lets work be shared out.
2. Abstraction A chess program ignoring the wood grain of the pieces The rules of chess do not depend on it.
3. Abstraction A racing game skipping real tyre chemistry A believable feel matters more than accuracy.
4. Abstraction Calling calculate_fine without knowing how it works The sub program hides its detail: abstraction.
5. Abstraction A register recording only present or absent Only the needed detail is kept.
6. Decomposition Breaking a game into menus, levels and scoring Splitting into parts is decomposition.
7. Decomposition Writing one sub program for each task Each task in its own block is decomposition.
8. Abstraction A sat nav ignoring the colour of buildings Leaving out detail is abstraction.
9. Decomposition Splitting a library system into issuing, returns and fines Breaking the system into parts is decomposition.
10. Abstraction A Tube map that ignores real distances Passengers only need the connections.
11. Decomposition Testing the login part on its own Testing a part in isolation relies on decomposition.