Running head: PSYCHOLOGY PSYCHOLOGY 2 Autism: How is it diagnosed? Autism is defined as a situation where there are impairments associated with growth and development of the brain. It is basically associated with impairments in social interaction, as well as verbal and nonverbal communication. These signs are noticeable at the age of two years. However, these signs develop gradually but some children with autism tend to reach their development stage normally and then change. Autism is caused by a combination of environmental factors as well as genetic. Autism that is caused by genetic is said to be complex. For example, if it result out of hereditary and the conditions happen to be high, then it is considered as primarily genetic. On the other hand, autism is also associated with infections that sometimes happen during pregnancy. A good example, in this case, is the one caused by the use of alcohol or other drugs like cocaine. (Speaks, A. 2011). To answer the question how autism is diagnosed, we need to focus our attention on the development changes that occur in a child. There is no medical test that can be used to diagnose autism. Therefore in this particular case, trained psychologists are used to administer autism. This is done through a specific behavioral evaluation test. However, it is very imperative to note that parents are always the first to notice any unusual behaviors in a child. Therefore some of the sign associated with autism include failure to maintain eye contact, and other weird behaviors that are not common in other children. In this regard, the first way of diagnosing kids is called the M-CHAT. This is conducted like an interview scenario where a specialist asks informative questions about a child. Therefore based on the type of answers gathered, then this gives a way ahead of whether the subject can be evaluated by a specialist. In this case, it could be a neurologist or other advanced specialists. In this particular case, parents are encouraged to find a doctor who will address this matter well and if further examination is necessary he/she can refer the matter to the appropriate specialists. Therefore efforts have been made by autism organizations to sensitize about autism signs to both physicians and parents. This basically means that every child below the age of two years should be thoroughly screened. This is to ensure that the developmental progress of a child is normal. The evaluations should involve hearing as well as using an autism-specific test like the one mentioned above. A typical diagnostic is another method used to diagnose autism victims. This approach involves using a team of specialized doctors from different fields. I.e. psychologist, therapist etc. This is another way of diagnosing autism. On the other hand, genetic testing is be very adamant combined with screening for other related medical issues. This comprehensive analysis of a child is important to parents because it helps them understand ...

1. Running head: PSYCHOLOGY
PSYCHOLOGY
2
Autism: How is it diagnosed?
Autism is defined as a situation where there are impairments
associated with growth and development of the brain. It is
basically associated with impairments in social interaction, as
well as verbal and nonverbal communication. These signs are
noticeable at the age of two years. However, these signs
develop gradually but some children with autism tend to reach
their development stage normally and then change. Autism is
caused by a combination of environmental factors as well as
genetic. Autism that is caused by genetic is said to be complex.
For example, if it result out of hereditary and the conditions
happen to be high, then it is considered as primarily genetic. On
the other hand, autism is also associated with infections that
sometimes happen during pregnancy. A good example, in this
case, is the one caused by the use of alcohol or other drugs like
cocaine. (Speaks, A. 2011).
To answer the question how autism is diagnosed, we need to
focus our attention on the development changes that occur in a
child. There is no medical test that can be used to diagnose
autism. Therefore in this particular case, trained psychologists
are used to administer autism. This is done through a specific
behavioral evaluation test. However, it is very imperative to
note that parents are always the first to notice any unusual
behaviors in a child. Therefore some of the sign associated with
autism include failure to maintain eye contact, and other weird
behaviors that are not common in other children.

2. In this regard, the first way of diagnosing kids is called the M-
CHAT. This is conducted like an interview scenario where a
specialist asks informative questions about a child. Therefore
based on the type of answers gathered, then this gives a way
ahead of whether the subject can be evaluated by a specialist. In
this case, it could be a neurologist or other advanced specialists.
In this particular case, parents are encouraged to find a doctor
who will address this matter well and if further examination is
necessary he/she can refer the matter to the appropriate
specialists. Therefore efforts have been made by autism
organizations to sensitize about autism signs to both physicians
and parents. This basically means that every child below the age
of two years should be thoroughly screened. This is to ensure
that the developmental progress of a child is normal. The
evaluations should involve hearing as well as using an autism-
specific test like the one mentioned above.
A typical diagnostic is another method used to diagnose autism
victims. This approach involves using a team of specialized
doctors from different fields. I.e. psychologist, therapist etc.
This is another way of diagnosing autism. On the other hand,
genetic testing is be very adamant combined with screening for
other related medical issues. This comprehensive analysis of a
child is important to parents because it helps them understand
their children’s strengthens and needs. However, autism
disorder can be diagnosed later in life especially in relation to
difficulties encountered in life. For example socialization.
Autism for an adult is simply diagnosed by observation or by
involving interviews by a trained specialist.
Due to the fact that they are no medical test for diagnosing
autism, it is therefore conducted based on observation of an
individual life. In most cases, however, this involves paying
attention to the victim’s social interaction as well as their
activities and interests. This is because some individuals

3. suffering from autism tend to develop some interests that are
weird compared to those of normal people. Because autism
disorders are related to other disorders, it is, therefore, possible
to conduct a medical test to identify other causes. In most cases
people with autism tend to suffer from various symptoms
especially associated with physical, behavior and mental
conditions, therefore, medical professional who are well
familiarized with autism can be able to conduct medical
diagnoses. Therefore parents should be keen when looking for
appropriate professionals. (Cicchetti, D. V., Volkmar, F., Klin,
A., & Showalter, D. 1995)
References
Cicchetti, D. V., Volkmar, F., Klin, A., & Showalter, D. (1995).
Diagnosing autism using ICD-10 criteria: A comparison of
neural networks and standard multivariate procedures. Child
Neuropsychology, 1(1), 26-37.
Speaks, A. (2011). What is autism. Retrieved on November, 17,
2011.
Assignment2/A2-12-Feedback Sample.doc
There is no need to print this document.
These are sample feedback from assignment 2 from 2012.
Student identification has been removed and there may be been
some rewording (to make it more general) of the actual specific
feedback that these students receive.
The application area was different in 2012, but the data
structure requirements were the same. Tree and maps were
mandatory then and are mandatory now.
You will find it instructive to go through this feedback when
working on your assignment 2 as a lot of the feedback relates to
the data structures that you are using now.

4. Also, going through these will give you an opportunity to learn
from others as well.
ICT209 Assignment 2, 2012
Student Name:
Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate

5. document called “evaluation.txt”
Yes
Components
Comments
UML diagram
UML is fine. All good.
Written rationale for the design + algorithm
Rationale for the maps is fine. It is a shame you didn't use your
tree, because there is quite a heavy mark reduction for not using
the required structures.
The templated BST is in the textbook. Should have attempted
the lab exercise on trees first. This lab exercise had to be done
anyway as it was assessed. If the lab exercise had been done,
you would have worked out the operation of the BST much
earlier and would have been able to incorporate the BST into
the assignment.
There are plenty of options for combining the 2 structures; store
dates in the tree, map dates to readings; store readings in the
tree, map readings to dates; map dates with time/max reading. It
didn't have to be cleanest answer, even if the rationale behind
why you map/tree something seemed illogical, it still would
have been better than getting the -30 marks for not doing it.
Algorithm is concise and clear; all good.
Program including coding style and comments
I was nearly going to put a no next to the doxygen minimum
requirement, but then I found it pretty well hidden in your VS
project files. I would have moved that to a more obvious
location before submission. That said, your doxygen comments
are good.
I get an assertion error while attempting to load the input data.

6. But, this does correlate with your test data. Unfortunately
though, it makes it hard to test to see if your application is
functioning correctly or not.
Your coding style is okay. I only found one instance of an
attribute without an obviously clear name, which was “m” in
your main application. That could be anything at a glance. It
was a monthly reading though, so something like
monthlyReading is much clearer.
Test plan and testing
To be honest, it is good to see someone who actually puts their
failed test results in as well. It looks a lot more authentic to
have tests that fail, because in the real world it happens all the
time.
You made a comment about your test programs. I really needed
to see the test outputs from the application as well to correlate
them against the test plans. Since you have a column that says
output in your test plans I am going to presume that was copied
for application outputs.
Other advice (if any):
If you were having trouble creating the tree structure and
working out where/how/why to use it, email me. I may not reply
within 24 hours for various reasons, but we could have sorted
that out within a few days.
Details of any penalties applied:
-30: BST was not used.
Number of days late: Late penalty (10 marks a day):

7. Final Grade:
ICT209 Assignment 2, 2012
Student Name:
Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram

8. Concise and clear. Where are the details for BST and BSTNode
in the UML?
Written rationale for the design + algorithm
Perfectly fine rationales behind the use of the data structures.
From your rationales I can already see inconsistencies with
method names though; start with capitals or lower case?
Program including coding style and comments
Where's your Code::Blocks project file? I had to create the
project myself to see if it builds correctly. You had it in
assignment 1's submission. I shouldn't have to add to a students
submission to see if minimum requirements are meet (builds on
VS or C::B). As it is the first time you have done this, I'll just
give you a warning about it. Always submit the project files as
well.
Your BST and BSTNode classes stand out from the rest of the
project. Lack of UML data, lack of doxygen comments, and
mostly written in a different coding style (different naming
conventions for parameters, method, and attributes; different
location for the { in the code blocks).
The inconsistencies of the BST and BSTNode classes can invite
questions as to the authorship of these classes. If the code was
sourced elsewhere and not referenced, it becomes plagiarism.
Although I don't wish to imply that the BST code was sourced
from elsewhere, it is inconsistent with the rest of the project,
making it stand out. If say, for example, you get code examples
from books, use them, and reference them. For example, BST
code from the textbook could have been used and referenced.
Check you BST’s destructor. Deleting the root node doesn’t
delete the tree. Please see the textbook which shows how this is
done.

9. It does not make sense to expose the internals of the BST with a
public getroot() method. Makes a mockery of the encapsulation
that is done via the class mechanism.
The class declaration doxygen comments are good, but they
should go directly before the declaration in your source. For
example:
#include “someheader.h”
using namespace somenamespace;
/**
* @class example
* etc ..
*/
class MyClassName {};
The rest of your doxygen comments are very brief; just in the
form of single line /// comments. There is a good example of a
doxygen comments on the unit website @
Practice/DoxyExample/unzipped/ModelFile.h.
Generally, your code in clear enough to be read by a third party,
so that's all good. It is important to focus writing readable code
first.
Program execution is fine. No issues there.
Test plan and testing

10. Your test plans are fairly brief, but structured okay. I would
have tried to have had a bit more detail here where possible.
Other advice (if any):
Reference sources, if you use them. Supply everything that
requested (in this case, the code::blocks project file).
Details of any penalties applied:
Number of days late: Late penalty (10 marks a day):
Final Grade:
ICT209 Assignment 2, 2012
Student Name:
Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate

11. directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram
Your UML is simple, but accurately reflects your classes in
your project. All good.
Written rationale for the design + algorithm
Although I feel the goals that your maps achieve could be
handled by data validation in the date class, your rationale
behind why you chose to implement your maps in a certain way
is sound. So no issues there at all.
What I am having trouble finding is a rationale for why your
tree contains the data it does. I may be missing something, but I
read the Use of Data Structures: Maps & Binary Trees section
twice, and the first two paragraphs state that the assignment
needed a BST and that you had created one.
So I had a look at CustomeBinaryTree, and I see a lot of how,
but not really why, it stores the data it does. I'm sure this only a
misinterpretation of what you are trying to express. I'm not
marking you down for it, but personally I feel it could have
been stated clearer.
Program including coding style and comments
I had fun typing long sentences in the menu and watching the
error message spam for every character in the sentence. It's
good though, because the

12. application didn't break with inaccurate inputs.
Took me awhile to workout that when your asking the user to
enter the month, they need to enter it as a word. Which is fine,
but there would have to be a note for the user about the format
for data input.
Other than that, your application works as intended. All good.
Your doxygen comments are nicely detailed, so that is all good
too.
Test plan and testing
This is really good here. Lots of detailed testing. The test plans
are fleshed out; the outputs are shown; each unit has been tested
individually.
Probably not the most exciting part of a project to excel at, but
your testing documentation is very good.
Other advice (if any):
It is good that you have a positive attitude towards coding, as
expressed in your evaluation document. It really does help when
attempting to write something.
Taking the feedback from assignment one on-board, your testing
is great, and your doxygen comments are up to a good standard.
Details of any penalties applied:
Number of days late: Late penalty (10 marks a day):
Final Grade
ICT209 Assignment 2, 2012
Student Name:

13. Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram
UML looks fine. Had to really zoom in though to make it out -_-
Written rationale for the design + algorithm
It's good that you reference self-balancing trees. Everything
comes at a cost though; self-balancing trees take much longer to
input data as the tree is rearranging itself after each insert.

14. Rock solid rationale for your tree and map.
Clear, meaningful names for your methods and attributes. All
good there.
Algorithm is good. It's quite long and detailed, but that's no
issue.
Program including coding style and comments
Yes I did have the side-by-side configuration error again. But
it's okay because you mentioned it in your evaluation document.
But you do need to get this fixed.
Doxygen comments are good. I can see you have adjusted the
little things from the first assignment, so no issues here
whatsoever.
I like that you have extended the menu options to add more
functionality. It's simple but effective.
I think your coding standard for a 2nd year unit is very good.
Meaningful names, clear program flow, well commented; all
good. But I have to pick on you for something, so I'd say more
use of the const keyword. In particular after method
declarations.
For example:
void IDontChangeAnything() const;
and
void IChangeSomethingSoNotConst(const int&
valueThatDoesntChange);

15. All things considered it is a minor consideration.
Test plan and testing
Test plan is good. The test methods correlate to the outputs. All
good.
Other advice (if any):
Details of any penalties applied:
Number of days late: Late penalty (10 marks a day):
Final Grade:
ICT209 Assignment 2, 2012
Student Name:
Minimum Requirements:
Satisfied Y/N
UML diagram
No
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.

16. Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram
This is, rather unfortunately, what I see in your documentation:
Two large blank squares. You probably have done your UML
and attempted to insert the images, but as I cannot see them at
present, I will have to mark it down as a missing requirement.
You could have provide the uml as a separate file and referred
to this file in your write up.
Written rationale for the design + algorithm
As with the first assignment, having identifiers on your private
member attributes (the m_) is very useful for people reading
your code.
Although brief, and not actually under the section Written
Rationale, I found a rationale behind why you used your map
and tree the way you did. Which is good, as it was a
requirement (the paragraph at the end of 5. Program).
Rationales are all good.
Algorithm is good too; no dramas.
Program including coding style and comments
Avoid using #define and global variables where possible. For

17. starters, number of months is never going to change, also the
months array is only used in 2 places; month_str_to_int and
month_int_to_str. You could have mapped the date strings to
integer values in your main, or possibly considered an
enumerator to represent the months.
Your BST class uses a different naming convention to the rest
of your classes. Also, add doxygen comments for your private
methods as well. Other than that, no issues with your doxygen
comments/outputs.
The program runs good. I'm not very concerned about how fast
the application runs, as long as it runs correctly, which is seems
to do. All good there.
Test plan and testing
No pass/fail column for your test plans? Other than that, your
test plans and outputs are good.
Other advice (if any):
Double check your documentation after you save it in different
formats.
If you do well in the exam this penalty may be reassessed.
Details of any penalties applied:
-50; missing UML.
Number of days late: Late penalty (10 marks a day):
Final Grade:
ICT209 Assignment 2, 2012
Student Name

18. Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram
Your UML is fine, but I can see inconsistencies in naming
conventions for methods in the BST. Methods start with upper-
case or lower-case? Either is correct, just be consistent.

19. Written rationale for the design + algorithm
Although you offer a rationale behind why some private
attributes in the Power class are prefixed with m_ and some are
not, they are still all private attributes and the m_ would
identify that clearly to anyone reading your code. I
recommended staying consistent with that in the future. Same
thing with your node in BST.
Your rationales are fine, but it is bit long winded. Scientific
documentation tends to be brief and concise (we're too easily
distracted by shiny things and buttons). Leave the 3000 word
essays to the Law students. It will save yourself a lot of typing
on future assignments =)
A good example is your BST/Map rationale. If you supplied the
list and reworded it slightly to express reasoning as well as
purpose, that would be all you need.
Algorithm is good. Clear and concise.
Program including coding style and comments
#pragma are Visual Studio directives; they aren't portable to
other compilers. Also, suppressing a warning instead of fixing it
is like hiding the body under your bed and just hoping it will go
away.
While we're on portability and re-use, the very first line of your
program calls a batch file to execute a DOS command to
generate a directory listing. Obviously this greatly reduces the
portability of the code and also making the application
dependant on the BAT file for operation.
Here is a C example of reading a directory listing:
http://www.dreamincode.net/code/snippet2295.htm.

20. All those global variables again. The 101 Coding Standards #18
states: Declare variables as locally as possible. You had already
written it for the first assignment and I wouldn't expect you to
change previous work, but by having a line in your
evaluation.txt document like, “I am aware that my global
variables should be local variables, as per 101 coding
standards”, completely covers you. It shows that you are aware
of the issue, even if the issue hasn't been rectified directly.
Doxygen comments are good. All good there.
Test plan and testing
“Your evaluation documentation does not list what does and
does not work in your program. If you feel your submission is
working as intended then include something like, “All required
features implemented and working.” in future submissions.” -
quoted myself from your assignment one feedback.
Good test plan; simple and clear.
Other advice (if any):
The only thing that stopped you from getting a HD was your
evaluation.txt document. Use it to list even the smallest things
you think you could have done better and you won't get marked
down for them. The example I gave above was the heavy use of
global variables. As it is, your evaluation document only
contains test outputs and I commented on that in the first
assignment too.
Details of any penalties applied:
Number of days late: Late penalty (10 marks a day):
Final Grade:
ICT209 Assignment 2, 2012
Student Name:

21. Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram
Brief, accurate and detailed. All good.
Written rationale for the design + algorithm
I noticed your deleteNode method is not implemented. Delete
node is also in the textbook. There is a fair amount of detail
with diagrams, algorithm and code.

22. Also see
http://www.algolist.net/Data_structures/Binary_search_tree/Re
moval. It has pictures and code examples.
What are you mapping your CollectedRecord struct to in your
map in your rationale description? Record holds the date, units,
time of max output, and max output; what is in the rest of the
map?
Good clear method names in your design.
Algorithm is clear. All good there too.
Program including coding style and comments
Once again, no issues with the comments / doxygen ouput.
Menu is fine. Program functions accurately. No issues there.
Same as what I said in the first assignment; having the
application as an interface shows a good understanding of
object-orientation concepts. Having completed the games units
your code is clear and well structured from experience with
coding now.
In your BST I would have avoided C-style #defines of values.
Consider possibly using a enum for the same result (your
custom method can return enum values instead). It's not
technically wrong by any means, only a suggestion.
Test plan and testing
This is more of a general formatting thing with your
documentation than an issue with the way the information is
presented, but using page breaks here so that each test starts on
a new page would have made it look at lot clearer and easier to

23. read.
Your test information is there, but it's not really structured in a
test plan. Topic 8 has an example test plan document (the XLS
link).
Other advice (if any):
Details of any penalties applied:
Number of days late: Late penalty (10 marks a day):
Final Grade:
ICT209 Assignment 2, 2012
Student Name:
Minimum Requirements:
Satisfied Y/N
UML diagram
Yes
Documentation asked for: Design outline including
justification/rationale
Yes
Doxygen output
Yes
Program that builds and runs.
Yes
Source code with doxygen style comments. (soft copy only)
Yes
Test plan and output of test run(s)
Yes
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.

24. Yes
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Yes
Components
Comments
UML diagram
UML is good. Relationships are accurate,
public/private/protected is all accurately depicted, and
attribute/method names correlate to your classes. So all good.
Written rationale for the design + algorithm
From assignment2.pdf; “You must use the Binary search
tree data structure and the STL map (and/or multimap) data
structure. You need to identify where these
data structures can be used and provide a rationale for their use.
There is a 30 marks penalty if you do
not do this.”.
This is your rationale for the map and bst:
map<int,Record>; STL Template class; public;
Stores elements formed by the combination of a key value and a
mapped value.
BinarySearchTree<int>; Template class; public;
Stores int elements in an ordered tree stucture. Used for fast
searching.
You need to tell us why you are using these two structures in

25. the way you are. What is the integer value stored in the map and
the tree? Why is that integer mapped to a Record? What is a
Record? There just isn't enough here to get around the -30 mark
hit. See note at the bottom about penalties.
The rest of your rationales are fine.
Algorithm is good.
Program including coding style and comments
Tried putting invalid data inputs into your application. It didn't
break it, but for example, if I put the word 'dog' in when the
application is asking me to input a value to test against, then it
treats each individual char in the string as a separate input, and
loops through the process multiple times.
Data validation wasn't a requirement of the application, so it's
not a big issue; just noting it. Otherwise your application seems
to behave itself.
Once again your doxygen comments are very brief (see what I
did there O_o). It's fine for the submission, but I was really
hoping you would flesh the comments out with more details.
There is a good example header file on the unit website @
Practice/DoxyExample/unzipped/ModelFile.h.
Your coding style is very good. Good use of the const keyword
in particular. Clear, concise, easy to read; everything code
should be really.
Test plan and testing
Test plans are clear; good. Outputs correlate. No dramas here at
all.
Other advice (if any):

26. Ultimately, the only thing that let your whole submission down
was the lack of rationale behind the use of maps and trees. But,
if you do well in the exam then penalties may be re-evaluated.
So it isn't the end of the world just yet.
Details of any penalties applied:
-30: Rationales for Maps and BST.
Number of days late: Late penalty (10 marks a day):
Final Grade:
Assignment2/Assignment2-T-16.rtf
ICT209 Assignment 2, Murdoch University 2016
9
ICT209 Assignment 2, Murdoch University 2016
ICT209 Assignment 2 2016
Objectives:

27. Demonstrate that you can do Object Oriented design.
Demonstrate that you can write Object Oriented programs using
C++.
Demonstrate that you can design and write programs using user
defined data structures. Demonstrate that you can use data
structures appropriately.Demonstrate that you can write test
plans and show evidence of systematic testing.
Demonstrate that you can draw UML diagrams.Acquire further
experience with design and code maintenance.
You do not work in groups for this assignment as this is an
individual assignment.
Worth:
30% of the unit
Due:
Midnight, end of Session 12 (last teaching week). See due date
and time (and any amendments to these) in LMS. The due date
and time depends on when is the last day of the last teaching
week for the semester or trimester at the campus you are
enrolled in.
How to submit:
Same way as assignment 1. The submission must contain the
entire directory/folder in which the source was built. Other
submission items are specified in this document.
For submitting in LMS, zip up the entire folder. Make sure that
you have included all needed files. Do not include temporary
files or files not relevant to the assignment. After submitting
the assignment, log out of LMS. Then log back in again. Check
that the assignment is submitted.
Name the zip file with the unit code, Assignment number, your
name, student number.

28. ICT209Asg2JoBlogs12345678.zip
or alternatively,
ICT209_Asg2_JoBlogs_12345678.zip
Textual submissions should be type-written. External
documentation can only be in the following formats:
Text (.txt)
PDF (.pdf)
RTF (.rtf)
HTML (.html)
Image formats : PNG, GIF, JPG, TIFF, BMP. (BMP and TIFF
cannot be used for Web documents)
Assignment cover sheet requirements are listed in the unit
outline found in the Admin area. Also see the end of this
document.
Mandatory Readings/work needed:
Unit textbook. “C++ Programming: Program Design Including
Data Structures” by D.S. Malik.Lecture notes.Complete lab
exercises. You need to complete all exercises up to laboratory
session 10 to cover the data structures requirements of this
assignment.Complete laboratory session 11 before submission
of this assignment.Question and Answer files (QandA) in the
assignment 2 folder.C++ Coding Standards http://0-
lib.myilibrary.com.prospero.murdoch.edu.au/Open.aspx?id=291
948" http://0-
lib.myilibrary.com.prospero.murdoch.edu.au/Open.aspx?id=291
948
Assignment Question:
Design an object-oriented solution and implement the solution
in C++ to solve the problem described below. You must read the
entire specification and work out what is needed before starting
work on this assignment.

29. Like assignment 1, this assignment is not for actually buying
and selling shares. The assignment only deals with data from
past share transactions (trades).
The solution needs to be worked out on paper before you start
coding.
Extend assignment 1 to do the tasks described below. You need
to also reflect on the design and code changes needed to convert
your assignment 1 into assignment 2.
Start early. Design the menu system and test to make sure the
menu works. Use code stubs (refer to your earlier units) to test
the overall flow of execution. If you have forgotten what a stub
is, see http://en.wikipedia.org/wiki/Method_stub"
http://en.wikipedia.org/wiki/Method_stub for a quick overview.
You already have data file reading code, so make sure that you
are able to read data files correctly. In the initial stages, read
the files and print them out to check. Once this stage is
completed, you start using the data structures after unit testing
the data structures.
Create your own test data files as given in these specifications
below. You will be submitting these files.
Make sure you design a simple menu so that users do not get
confused or annoyed – consider what it is like if you were the
end-user.
Your program should be able to deal with course of sales data
for multiple days and multiple stock exchange codes (x-code).
Each day’s trading data for each stock code (x-code) is still
stored in CSV files. This means that there will be a number of
data files for each stock code. For proper testing (for the
various test conditions), you need to create your own data files.
These test data files shouldn’t be large. There should only be

30. enough data to check the test conditions. This involves various
x-code values and dates. You decide on these values and the
decision is important as the data values are used to test your
solution.
Program input:
The data format in each course of sales file is the same as
assignment 1. As there are multiple x-codes, the date files are
arranged in directories or folders. The folder names are the x-
codes. To illustrate, assume there are course of sales data for 3
companies with their own x-codes. Assume the x-codes are
IAG, NAB, CBA. The data folder will have 3 folders called
“IAG”, “NAB”, and “CBA”. This data folder will have an index
file called code_index.txt. The contents of this index file would
just be the x-code names whose folder exists in the data folder.
The folder names match the x-code names.
So, for this example, code_index.txt will contain each x-code on
a line:
IAG
NAB
CBA
The contents of code_index.txt indicate that there are these 3
sub-folders in the data folder.
Do not make any assumption that the x-code is sorted in
code_index.txt.
The course of sales files (in CSV format, as in assignment 1) for
each day and given x-code will found in the sub-folder of the
data folder. As indicated earlier, the sub-folder is named by the
x-code. Do not make any assumptions about the naming
conventions used for the course of sales files. The names of the
files in the x-code sub-folders are listed in a file called
sales_index.txt. Each x-code subfolder will have a

31. sales_index.txt file.
Assuming that the folder IAG has course of sales files day1.csv,
day2.csv, day3.csv, the sales_index.txt file in folder IAG will
contain the following lines.
day1.csv
day2.csv
day3.csv.
Do not assume that there is any sort order in sales_index.txt.
The names of the CSV files do not indicate any chronological
order. Arrange your data files to reflect this – change the order
in which the files are listed and check.
So to read course of sales data into your program:
In the data folder, read the code_index.txt file.
For each code in code_index.txt
Read code/sales_index.txt
For each csv file in code/sales_index.txt
Load the csv file data into your program.
Endfor
Endfor
The program reads all files and loads the appropriate data
structures before the menu is displayed to the end user.
Program output:
The program reads data from the data files and produces output
according to the menu option selected by the user. Assignment 2
deals with multiple stock codes and course of sales for multiple
days. Consequently, most of the menu options are variations of
what was needed for Assignment 1.
Date is shown as dd/mm/yyyy. (02/01/2014).
Time uses the 24-hour format.
The stock code is the x-code.
Read the requirements for each option, and design your menu

32. system to be user friendly. Your design may have some of the
menu options listed below as sub-menus.
Before any menu option is displayed, the program reads all data
files and loads the appropriate data structures.
Menu option 1:
The highest share price and start time(s) of the highest share
price during the day for a user specified stock code and date.
This is printed on the screen in the following format:
Stock code: <the stock code>
Date: <the transaction date>
Highest price: <the highest price>
Start time(s):
<time when highest price transaction occurred>
<time when highest price transaction occurred>
...
The highest price could be reached more than once during the
day. If so, list each time on a new row. Do not list duplicate
time values.
Menu option 2:
The lowest share price and start time(s) of the lowest share
price during the day for a user specified stock code and date.
This is printed on the screen in the following format:
Stock code: <the stock code>
Date: <the transaction date>
Lowest price: <the lowest price>
Start time(s):
<time when lowest price transaction occurred>
<time when lowest price transaction occurred>
...
The lowest price could be reached more than once during the
day. If so, list each time on a new row. Do not list duplicate
time values.

33. A value of 0 for price should not be part of the output.
Menu option 3:
This option is for a user specified stock code and date.
The output goes to a file called price-change-date.csv where the
fields (individual items of data) are separated by commas. The
date part of the file name depends on the user specified date.
The format is yyyy-mm-dd. As an example, an output file can
be price-change-2014-02-10.csv. This is the output for the 10th
of February 2014. This file is created in the specified x-code
directory.
The output format is:
Stock code, Date, Start time, Price of share, Volume of shares
traded, Total value of shares traded
Make sure you write the header row. This is the first row of the
output file. This involves simply printing the format line above
as a string.
The data file will have a record (row) for each time the share
price changes. The output will have data arranged in increasing
time order.
The output fields have the following meanings: Stock code: The
x-code.Date: Date of trading.
Start time: The time when the share price changed to the value
indicated in the "Price of share" field. This time value is used to
indicate that the share price changed to a new price value. The
input data file may have multiple consecutive records (rows)
where the share price is the same. This means that there may be
multiple consecutive share transactions where the share price
remained the same. In the output data file, output.csv, the Start
time field records when the share price changed to a different
value. The intention here is that this field records the time of
initial change to a price in a series of consecutive same price
transactions.
Volume of shares traded: The number of shares traded since

34. "Start time" at the value indicated in the "Price of share" field.
Total value of shares traded: The total monetary (dollar) value
of the shares given in the "Volume of shares" field.
Treat a value of zero (0) for the price field in the input data file
the same way as other prices are treated.
Menu option 4:
This menu option is to be completed only when all other menu
options are working as expected.
Proper completion of this option will give you a 15 marks
bonus. If you attempt this option when anyof the other options
are not working, no bonus marks would be awarded.
It is theoretically possible to get 115 marks if all options are
done perfectly. However the maximum mark that will awarded
is 100.
Calculate a simple moving average (SMA) for the price of a
user specified stock code on a given date. The moving average
is over 5 transactions by default. The output goes to a file
named SMA-date.csv.
The date part of the file name depends on the user specified
date. The format is yyyy-mm-dd. As an example, an output file
can be SMA-2014-02-10.csv. This is the SMA for the 10th of
February 2014. This file is created in the specified x-code
directory.
The output format is:
Stock code, Date, Time, Price of share, simple moving average.
Make sure you write the header row. This is the first row of the
output file. This involves simply printing the format line above
as a string.
Stock code and Date have the same meaning as in option 3.
Time is the recorded time in the course of sales file. Price of

35. share is the price recorded in the course of sales file. The SMA
is described at
http://www.investopedia.com/terms/m/movingaverage.asp"
http://www.investopedia.com/terms/m/movingaverage.asp. Use
the approach described at this site but instead of using daily
closing prices, the actual transaction price at a given time
recorded in the course of sales file is used.
Provide a sub-menu of this menu that enables the default of 5
transactions for the moving average to be changed to some other
value.
Price values of 0 must not be averaged.
Menu option 5:
Exit the program
Data Structures requirement:
STL data structures and algorithms can be used in this
assignment. You must use the templatized Binary search tree
data structure and the templatized STL map (and/or multimap)
data structure.
The data for the output in the various menu options must come
from these data structures. You may use other data structures
together with the STL map and tree.
You need to identify where these data structures can be used
and provide a rationale for their use. There is a 30 marks
penalty if you do not do this.
You may use std::string and string stream classes in your
program instead of using C like strings. You may use iostream
and file handling classes and objects in C++. See laboratory
exercises.

36. When working on this assignment, think of the usability of your
program from the point of view of the user. This is the first
thing you have to do, even before you start designing the
program. This will form the menu system for the running
program. You should not make the program more complicated
than what is specified in the assignment. Make sure you provide
an option to exit the program.
Any advice and further clarifications to these requirements
would be found in the QandA file in the Assignment 2 folder.
Please ask your questions early and do not wait until near the
due date as you may not have time to incorporate any answers
into your assignment.
Documentation(all in soft copy):
UML diagram showing the design of your classes.Reflection on
the changes you had to make to cater for the requirements of
assignment 2. You describe what parts of your assignment 1
design and code enabled you to complete the requirements of
assignment 2 easily. What parts of your assignment 1 design
and code were a problem?Rationale for the use of the data
structures in the program.A high level algorithm for the
solution.Doxygen output which shows all information as was
done in the practice for session 2.Test plan and output of the
test run(s).
Do not print code. Code will only exist as soft copy. You need
to submit the entire directory/folder where you built your
program. All relevant data files need to be there and a way must
be provided to build with either Microsoft Visual Studio or
Code:Blocks on a Windows operating system.
Minimum requirements:
You must provide all of the following; otherwise 50 marks will
be deducted. UML diagram Reflection on the design and code

37. conversion from Assignment 1 to Assignment 2.Written
rationale for the data structures: tree, map/multimap and any
others used.Doxygen output Program that builds (using
Microsoft Visual C++ or code::blocks) and runs. Build
mechanism must be provided.Source code. Doxygen style
comments in header files. (soft copy only)Test plan and output
of test run(s) Executable program with associated data files in a
separate directory called “executable”. Make sure that the
executable runs on a machine which does not have a compiler.
Data files needed to enable it run should be included here.A
declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”. This is an unformatted text
file and it is not the test plan or output of test runs.
Marking
UML diagram 10%
Design reflection and rationale for data structures 20%
Program (includes coding style and comments) 55 % +
15% bonus for menu 4
Test plan and testing 15%
There is a 10 marks penalty if the cover sheet is not filled in
properly and/or the requirements of the coversheet are not met.
If the cover sheet is not provided, you get no marks. There is a
separate cover sheet available.
If progress on this assignment is not demonstrated to your tutor
in the weeks prior to the submission of the assignment, you may
not get any marks for this assignment. We need to know that
you have been working on the assignment. The assignment must
not “magically” appear. Please check the unit guide for the
requirements for passing the unit – section on “Determination of
the final grade”.

38. Penalty summary:
Specified data structures not used: -30 marks
Missing minimum requirements: -50 marks
Cover sheet problems: -10 marks (cover sheet is not required
now for LMS submissions, but you should ensure that all
requirements are met for the cover sheet and submission into
LMS.)
The lowest total mark you can get is 0.
School of Engineering and Information Technology
ICT209 COVER SHEET used for Non-LMS submission
Given Names
Surname
Student Numbers
Mode (D/X)
Email
Name of tutor: _____________________________ Day & Time
of tutorial: _________________
Assignment Number: __2___ Due Date: ______________
Date Submitted: ______________
If the given name by which your tutor knows you differs from
your name on university records, you should indicate both
names above. Tutor’s name must be entered. Penalty is 10% of
the total marks for the submission for not providing information
asked for.

39. Your assignment should meet the following requirements.
Please confirm this (by ticking boxes) before submitting your
assignment that you have checked the declarations you need to
make. Some items do not apply for online/electronic
submission.
Except where I have indicated, the work I am submitting is
my own work for the purpose of this assessment and has not
been submitted for assessment in another unit.
This submission complies with Murdoch University's
academic integrity commitments. I am aware that information
about plagiarism and associated penalties can be found at
http://www.murdoch.edu.au/teach/plagiarism/. If I have any
doubts or queries about this, I am further aware that I can
contact my Unit Coordinator prior to submitting the assignment.
I acknowledge and agree that the assessor of this
assignment may, for the purpose of assessing this
assignment:Reproduce this assignment and provide a copy to
another academic staff member; and/orSubmit a copy of this
assignment to a plagiarism-checking service. This web-based
service will retain a copy of this work for the sole purpose of
subsequent plagiarism checking, but has a legal agreement with
the University that it will not share or reproduce it in any form.
I have retained a copy of all submitted work.
I will retain a copy of the notification of receipt of this
assignment. LMS submission would provide the required
receipt. If you have not received a receipt within three days,
please check with your Unit Coordinator.
Assignment is presented on A4 size paper and is neatly
collated. (Internal students)
Assignment includes virus-free disk with machine-readable
programs & files relevant only to this submission. CD or DVD
only may be accepted. (Internal students)
Instructions relating to answering of questions, formats
used for electronic submission and LMS submission have been

40. followed.Writing is clearly legible or has been printed.
Pages have been firmly stapled. (Internal students)
____________________________
Signature for submission via the assignment box (internal
students)
Assignment2/FeedbackTemplateA2.doc
ICT209 Assignment 2, 2016Student Name:
Minimum Requirements:
Satisfied Y/N
UML diagram
Documentation asked for: Design outline, reflection,
justification/rationale
Doxygen output
Program that builds and runs.

41. Source code with doxygen style comments. (soft copy only)
Test plan and output of test run(s)
Executable program with associated data files in a separate
directory called “executable”. Make sure that the executable
runs on a machine which does not have a compiler.
A declaration indicating what works and what does not work in
your program. This can form the summary of your test plan and
output of test runs and should be provided as a separate
document called “evaluation.txt”
Components
Comments
UML diagram
Design reflection and rationale for data structures + algorithm
Program including coding style and comments
Option 4 bonus (not looked at unless all other options are
working)
Test plan and testing
Other advice (if any):
Details of any penalties applied:
Number of days late: Late penalty (10 marks a day):
Final Grade:
PAGE
1
Assignment2/QandA.txt
Important advice about the assignment is next, followed by the

42. QandA:
Please read the requirements carefully and plan how you are
going to
tackle the problem. Do not code new functionality until you
have
completed the lab exercise on trees. (Topic 9 and lab 9)
You should of course clean up your assignment 1 code to make
it ready for
assignment 2. This includes changing the menu code, any other
classes
you may have had to deal with the data in assignment 1.
Substantial changes to the class code to cater for new
requirements
would normally imply that your original design was not very
good. This
is not a big issue because you are still learning but you need to
be
conscious of this and think of how you could have designed the
classes

43. better. Classes need to be cohesive; class behaviours (methods)
need to
be minimal and complete, functionality of any public method
must not be
duplicated in another public method - no duplication of
functionality in
public methods; friends are to be avoided; data structure classes
do not
do I/O, nor do they handle formatting issues - should have
discovered
this after completing lab 5.
Everything that is pointed out out above applies to all OO and
to
whatever implementation language you are using. In fact, all of
these
ideas existed in Computer Science before any of the languages
you know
about existed. We usually cover details of the Computer Science
theory
for this in another unit.
Keep in mind that just because you can write C++ code (or Java,

44. Javascript, VB, C#, F#, .. etc) does not mean that you are doing
OO
properly. OO design happens *before you write any code*. The
language
only implements your design - the virtual level of abstraction.
____________________________QandA___________________
_______________
The latest QandA is the one shown first. If you are reading
this for the first time, read from the bottom to the top.
Q:
So I'm interaction with my BST primarily through function
pointers
and they appear to be working correctly but when I try to used a
callback with one of my traversals passing the pointer crashes
the

45. program.
and I check that it's not crashing in the traversal function, it
doesn't even make it into the function.
the function is static so I simply pass it in as
ClassName::FunctionName I use other function pointer
similarly
earlier in the program and they appear to be operating correctly
(eg
I pass insert a compare function)
A:
See testTree.cpp in the assignment 2 area. It is annotated and
the
approach is described in this QandA file.
Q:
Quick Question again! So for Option 4, price value of 0 would
not

46. need to be in the output?
A:
The question specifies that 0 should not be used in calculation
of
SMA, so no output of 0.
Q:
Do we assume that course of sales file is sorted by time in
descending
order?
A:
You can assume the order that was in the assignment 1 data file.
Q:
Should the simple moving average directly work on course of
sales file
itself, or should be first sort the transaction in ascending order?

47. A:
It needs to work in increasing time order. It makes no sense for
it to
work backwards.
Q:
Also for my assignment 1 in question 3, I had changes of price
as 564,
however 559 is the right answer. Are we able to obtain the
sample for
that output for question 3.
A:
Yes, use the sub-totals function in the spreadsheet to work it out
for
you.
Q:
Quick question for menu option 4. As we are doing it in lots of
5 transactions, does that mean that the SMA for the first four
prices would be 0?

48. A:
See the worked calculation provided in the link. Use the same
approach.
There is no sma for the earlier ones. 0 is a valid value, so this
can't
be used. Leave blank.
Q:
it's it fair to assume that within 1 csv there wont be any
duplicate times?
A:
No, you can't make that assumption. See data file from
assignment 1.
Whatever was valid for the data file in assignment 1, is also
valid now.
Q:
I'm having trouble working out where the Binary Search Tree
would be

49. used. The only unique keys would be the file names but I don't
know
why I would need to do any traversals on the files. I do need to
do
traversals on the data within the files (menu option 3), however
I
can't use a BST to do it because there is no unique key for that
data.
A:
There is if you think about how the data is organised. Take a
step back
from the code, and look at the data organisation. What you want
to be
able to do is use the assignment 1 design/code at the design/core
of
assignment 2 with little or no modification to the assignment 1
design.
Then both the map and tree data structure usage becomes
somewhat obvious.

50. Q:
If our code doesn't work, or things like the Doxygen output or
StarUML
diagram isn't finished, do we still lose 50 marks for "Not
including
the bare minimum", if we submit the unfinished/nor working
parts
correctly?
A:
No, you don't lose 50 marks if you submit the unfinished parts
correctly. For example, if you have a program that runs and then
crashes, and you place the exe file along with the data files into
the
executable directory, then you are not going to have 50 marks
deducted.
In a situation like this, you would want to make sure that you
highlight the crashing in the evaluation.txt file.
Q:
I've been looking around the internet for resources on reading

51. file
directories, and a lot of people recommend using the dirent.h
header
file. From what I understand though, the header file is not
standard
for windows, so I was wondering if we would be able to use it,
and
whether we would have to include a copy of the file in our
submission.
A:
dirent is the usual way to go. You will learn more about this in
another unit. For this unit (ict209), it is not needed as the names
of
the directories (stock codes) are stored in code_index.txt files.
The
actual course of sales files are listed in the sales_index.txt file
in
each directory. See the description and high level algorithm
given in
the assignment question.

52. Q:
Can I use Visula Studio 2012?
A:
I have not heard of Visula Studio, but if you mean Visual
Studio,
the unit guide specifies only Visual Studio 2010 or codeblocks.
If your source does not build in either of these, it would mean
that minimum requirements were not met. You need to provide
a build mechanism too.
Q:
I was just wondering if it's possible to pass a function with a
parameter to a binary search tree?
Here is the function that I am passing:
portfolio.inOrder(MonthPurchased);

53. However both month and year are global variables.
I would like to pass them as arguments instead.
A:
No, that is not how it works.
You pass a function pointer (only an address) to the tree. The
function
pointer points to a function outside the tree and this function
can
accept parameters. The tree calls this function and sends the
parameter.
The textbook example demonstrates this. The data in the
parameter must
come from the tree.
If this function needs other data, this data needs to be "global"
to the
function. But if this "global" data is global to the program, then
it is

54. not good design - avoid using global variables.
So a better solution is to have this other data as part of a class
private area and the function pointer that is sent to the tree is
pointing to a method of the class. As it is a method of the class,
it
will have access to this data and because the data is declared as
private, the data is encapsulated and like any other private data
of a
class is not global to the program.
So you use the mechanics of the programming language to get
the good
design you want and C++ lets you do this.
The mechanism for doing this is already described in the QandA
- please
scroll down.

55. Q:
I understand that we can pass a function as an argument in the
traversal
functions. However this function can only modify the values not
return
them. I am able to use the callback function to place pointers to
the
elements of the tree in a global variable. I was wondering if
there is a
way to use the callback function to return the items (or pointers
to the
items) without using a global variable?
A:
The callback routine (in the client) can be sent the data by the
tree
and you can do whatever you want with this particular data. But
if you
want to keep all the data that is coming back from the tree, the
easiest
way is to use a global.

56. The better way is to send a ptr to a static method of a class.
When the
tree calls the method of this class, the method stores the data
being
returned in a static data structure that is private to this class.
This
data struct can be a simple array or a vector or something else.
If you
have time, try it out in prac 9 with the unit object.
// client class used to collect data returned from tree traversals
class A{ // use meaningful names. this is just an illustration of
the approach
public:
static void f1(Unit& d){ // assume the Unit class from prac 9
// this is the call back. d comes in, you print it out or keep
track of it in a static data struct
// so store in the private b array (in this case) or vector or
.. whatever but use better names

57. }
// other methods ... doesn't have to be static
private:
static Unit b[100]; // Unit from prac 9. most simple form just
an array but could be a vector, ... etc
... other stuff that you might need
}; // the next line is needed for static data - check textbook
Unit A::b[100]; // this creates memory. the declaration above
makes it private so can't access like a global.
For everything else, follow the design of classes we have talked
about.
For example you need a public method to access data stored in
b.
========

58. How to use in the client
BST<Unit> bu; // your tree of units from prac 9 - use better
names than this
A a1; // for collecting all the objects sent by the traversal - use
better names than this
// do stuff like insert unit objs into tree bu.
bu.inOrder(A::f1); // will call the static method f1 of A if you
got the function pointers working
// alterntively bu.inOrder(a1.f1); // object.method notation
passing the function pointer
// after this call to inOrder, object a1 will contain all the unit
objects returned by the bu.inOrder. Assuming you coded the
body of f1
correctly. Don't make it un-necessarily complicated. The code
in f1
should look obvious. f1 is the setter/insert for A. It just happens

59. that you sent this setter/insert to the tree.
Q;
From the UML diagram comments:
1)You have said that:
"Date’s setdate takes in a string parameter. This creates a
dependence
on the format. If a different format is required, this method
would need
to be modified. So it is not future-proofed."
Which I can understand, but then in the Trans action class,
where I have
provided SetQuantity(quantity: unsigned) and
SetQuantity(quantity:
string) you have said:
"Transaction is a little bloated. Functionality is duplicated.
Same date

60. is passed just with different types float and string."
A:
"Same date..." meant data.
Yes, only one way is to be provided. It is quantity and
quantities are
communicated using numbers. Whether the data is read in as a
number, or
as string or as a series of binary digits is not Transaction's
problem.
Q:
So I’d like to know what the best solution for this is. Using the
SetDate as an example, should I just provide 1 SetDate method
that takes
all unsigned variables -
SetDate(unsigned: day, unsigned: month, unsigned: year)- than

61. let any
other class or Main program that calls it deal with the data
format?
So the Main program might take in a string in "dd/mm/yyyy"
format,
then split the string, convert to an unsigned variables day,
month,
year, then call the SetDate(day, month, year) method?
Wouldn’t this style make the Main program messy because we’d
need to
do extra processing to accommodate the necessary data type
changes?
A:
Format is a client problem - so client (main) has to deal with
this. It
is up to the client to come up with a clean approach. The date
class
design should not be ruined and a coupling to date format
introduced to

62. deal with issues like this.
Q:
My intention was to make the class as flexible as possible with
the data
formats, but I do remember you saying that methods are like
holes
through a firewall. The more you have, the less secure it all is.
A:
Yes about the holes. Introducing coupling to client side issues
actually
reduces flexibility.
Once you have completed the class, it should not be touched
again to
deal with issues like the client wants the format to be different.
If
the client wants the format to be different, the client can put a
wrapper around the date class.

63. An object is designed to have behaviour that is consistent with
the
object's nature.
All of this is *not* just about C++. It is about any OO design
and programming
that you might do.
Q:
From the Program including coding style and comments:
2)You have said:
"Doxygen comments shouldn’t tell how something is done." For
example
//** ///

64. /* @brief Return value of m_size///
/* @warning None///
/* ///
/* Unsigned function to return value of unsigned variable///
/* m_size as a constant///
/* ///
/* @return value of private m_size member///
/* @pre None///
/* @post m_size value returned///

65. /*////
/unsigned GetSize() const;/
//
Saying that it returns private m_size, defeats the whole purpose
of
information hiding and preventing client dependencies from
occurring."
What you should say, it returns the size of the vector.
A:
The doxygen comments must treat the class and the methods as
black boxes.
Q:
3)You have also said:

66. "In .h files, do not include other .h files which are not needed.
Date.h has got the following:
#include <iostream> // why?
#include <string> // ok but indicated design issue see earlier
#include <sstream> // why stringstream code in the .h?
#include "EXCEPTION.h" // are exceptions dealt with in the .h?
Most of the above is needed in date.cpp. So include these
there."
I think the penny has finally dropped on this one. The header
file is
purely and simply that a header file only. It should only say the

67. information the public need to know, for both methods and
variables.
#Includes and anything else needed only for programming
shouldn’t be
included in the header file.
A:
Yes, that observation is correct.
Q:
I realise that stl::map stores its data in a balancing BST using
the key
to sort, but for the assignment you asked that we use both the
stl::map
and our own BST. I'm not sure whether to have our BST like
stl::map and
encapsulate stl::map or something else

68. A:
This is a good question.
As the user of the map, you must *not* assume anything about
how the map
data is stored. The map is *known only by its public
behaviours*
otherwise you introduce a coupling between the client (your
program) and
private parts (data or algorithm) of map. Defeats the whole
purpose of
having private and information hiding.
So for the assignment, you use the BST you built and the map -
two
different data structures.
Q:
Are we allowed to use more than one tree? I thought I could use
only
one tree and map.

69. A:
It wasn't ever a case that no one was allowed to use multiple
trees. It
is usual that some people don't realise that because a class
exists,
multiple objects of that class can be created. This is after all a
good
reason for creating a class - so that it can be reused.
So a vector of trees, queues of trees, stacks of trees, trees of
vectors, trees of queues, trees of stacks, trees of trees, maps of
trees
.. etc are possible. There are so many choices. So the best way
is to
make (and unit test) the templated tree needed (Prac 9b) and
then step
back from the code and look at the requirements of the problem
to be
solved along with the specified constraints and think at the
highest
(most abstract) level what a solution can be and why.

70. If you think at the code level, then the chances of getting lost in
the
detail is high. It will be a case of "missing the forest for the
trees"
(no pun intended).
Q:
Is there an perfect way to use the tree and map in the
assignment?
A:
The intention was that there shouldn't be. If there is perfect
way,
then there is not a lot of thinking involved to meet the data
structure
constraints on the assignment. I am looking for what is going on
in your
mind when you made the choice of where you going to use the
tree and map.

71. Q:
How much testing do I need to do?
A:
There is the usual unit tests to make sure that each class you
have
created, works. All methods of the classes must be tested to
make sure
that what you expect is what you get. A test plan that is created
when
you doing the class design in UML should list the tests.
Then there is a test which check all the classes working
together. This
is usually the test for the program that solves the problem (in
this
case, the assignment). Again, a test plan would show what is to
be
tested and why. As part of the test, you need to demonstrate the
the

72. calculations work as expected. This means that you have test
data
covering the necessary conditions. You would have calculated
the
expected output manually. The program is run with this test data
as
input and you need to check that the program gives the expected
result.
You would have a number of buy transactions and a number of
sell
transactions. The last sell transaction to be considered needs to
account for earlier sell transactions. It is important to make sure
that
calculations are correct.
Simple errors can cause significant problems and really hurt a
lot of people.
One example is the goto fail problem in Apple's code. This was
incorrectly
written if statement and this wasn't tested or reviewed by
someone else.
See discussion of it here:

73. https://www.imperialviolet.org/2014/02/22/applebug.html
(problem is described here)
https://news.ycombinator.com/item?id=7281378 (some
discussion of the problem)
Source code where the error was supposed to have occurred:
http://opensource.apple.com/source/Security/Security-
55471/libsecurity_ssl/lib/sslKeyExchange.c
For some examples of the consequences of errors, see this NYT
article by
a Nobel Laureate titled "The Excel Depression" where one of
the
questions asked is ".. did an Excel coding error destroy the
economies
of the Western world?"
http://www.nytimes.com/2013/04/19/opinion/krugman-the-
excel-depression.
html. A follow-up on this same topic is
http://www.nytimes.com/2013/04/26/opinion/krugman-the-one-
percents-
solution.html which discusses the snowball effect of error.

74. Q:
What is "mimimal but complete"?
A:
Follow the KISS principle as *originally* envisaged by the
person who
created this idea. See
http://en.wikipedia.org/wiki/KISS_principle#Origin (sections on
"Origin"
as well as "Variants" and what Antoine de Saint Exupéry had to
say about
perfection has been a past repeated exam question. You need
*not* have
heard of Antoine, but what he said about perfection would make
total
sense in the design of software systems. Scroll one page up to
see
section on "Quotations"
http://en.wikipedia.org/wiki/Antoine_de_Saint_Exup%C3%A9ry

75. #See_also.
Q:
I thought from your lecture that "this->" was used to de-
reference
class member pointer objects?
Any chance you can explain the difference with *this?
A:
Within a class, a member can be referred to as either the
member name on
its own or via the "this" pointer.
Example:
class A{
public:
f1(int b);

76. private:
int data;
};
// all 3 ways to assign to private member data mean the same
thing.
// the pointer "this" exists only when the object exists and so the
"this"
// pointer is not available in static methods
A::f1(int b){
data = b; //1st way
this->data = b; //2nd way
(*this).data = b; //3rd way
}
The 2nd and 3rd way uses the "this" pointer and are equivalent.
The 3rd way

77. de-references this to get the object pointed by "this" first, then
the object.name
notation is used to access the member.
You can think of it like this:
When a class client uses the class, an object of the class is
created.
A obj1;
obj1.f1(5); // can be imagined as if the compiler arranged for
a pointer
// to obj1 to be passed in as a hidden
// parameter to method f1 so imagine that the call was
f1(&obj1, 5);
// within f1, the hidden pointer to itself, i.e. obj1,
is named "this".
// note that obj1 is an object of class A.

78. Q:
Is this assignment like using binary search tree, map and
vector(assignment 1 bring over) or we can just use binary
search tree
and map
A:
You need to use the tree and map and explain why. It is up to
you if you
want to use the vector from assignment 1 as there is no
requirement
which says that you must use the vector from assignment 1.
Q:
I can do the assignment without using the tree or map. It works
and
does what is required according to the question. Must the tree
and map
be included?

79. A:
The assignment can be done in many ways where no tree, map
or even
vectors are used. In fact, the assignment can be done without
any
Object Orientation. But the specifications require Object
Orientation
and the use of both the tree and the map. You must use both tree
and
map. And you need to also provide justification for their use.
This
justification is *not* because the specification asked for it. You
need
to think of the properties of trees and maps and the justification
needs to be related to their properties. You may have to think
really
hard about this because it is possible that trees or maps might
not
really needed for this assignment but you need to make a case
for their
use.

80. You can provide the justification in the same table as the data
dictionary.
If you do no do what is asked, there is a 30 marks penalty.
There are also
other penalties that you should be mindful of. See assignment
question.
Assignment2/sample_data.zip
data/CBA/sales_index.txt
data/code_index.txt
CBA
IAG
NAB
data/IAG/day1.csv
Course Of Sales
Time,Price ($),Volume,Value ($),Condition
10/10/2013 04:57:27 PM,5.81,5000,29050.00,LT XT
10/10/2013 04:48:05 PM,5.81,62728,364449.68,SX XT
10/10/2013 04:10:33 PM,.00,0,.00,

81. 10/10/2013 04:10:33 PM,.00,0,.00,
10/10/2013 04:10:33 PM,.00,0,.00,
10/10/2013 04:10:33 PM,5.81,451,2620.31,
10/10/2013 04:10:33 PM,5.81,5000,29050.00,
10/10/2013 04:10:33 PM,5.81,145,842.45,
10/10/2013 04:10:33 PM,5.81,9241,53690.21,
10/10/2013 04:10:33 PM,5.81,8759,50889.79,
10/10/2013 04:10:33 PM,5.81,1875,10893.75,
10/10/2013 04:10:33 PM,5.81,58,336.98,
10/10/2013 04:10:33 PM,5.81,1370,7959.70,
data/IAG/day2.csv
Course Of Sales
Time,Price ($),Volume,Value ($),Condition
15/12/2015 04:57:27 PM,5.81,5000,29050.00,LT XT
15/12/2015 04:48:05 PM,5.81,62728,364449.68,SX XT
15/12/2015 04:10:33 PM,.00,0,.00,
15/12/2015 04:10:33 PM,.00,0,.00,
15/12/2015 04:10:33 PM,.00,0,.00,

82. 15/12/2015 04:10:33 PM,5.81,451,2620.31,
15/12/2015 04:10:33 PM,5.81,5000,29050.00,
15/12/2015 04:10:33 PM,5.81,145,842.45,
data/IAG/day3.csv
Course Of Sales
Time,Price ($),Volume,Value ($),Condition
20/05/2015 11:24:55 AM,5.85,148,865.80,
20/05/2015 11:24:34 AM,5.84,124,724.16,
20/05/2015 11:24:28 AM,5.85,148,865.80,
20/05/2015 11:24:24 AM,5.84,364,2125.76,
20/05/2015 11:24:03 AM,5.85,1,5.85,CX
20/05/2015 11:24:01 AM,5.85,14,81.90,
20/05/2015 11:24:01 AM,5.85,17,99.45,
20/05/2015 11:24:01 AM,5.85,118,690.30,
20/05/2015 11:23:49 AM,5.84,336,1962.24,
20/05/2015 11:23:40 AM,5.84,93,543.12,
20/05/2015 11:23:39 AM,5.84,310,1810.40,
20/05/2015 11:23:36 AM,5.84,167,975.28,
20/05/2015 11:23:31 AM,5.84,734,4286.56,

83. 20/05/2015 11:22:58 AM,5.85,73,427.05,
20/05/2015 11:22:58 AM,5.85,39,228.15,
20/05/2015 11:22:58 AM,5.85,36,210.60,
20/05/2015 11:22:16 AM,5.85,65,380.25,
20/05/2015 11:22:16 AM,5.85,83,485.55,
20/05/2015 11:21:58 AM,5.85,844,4933.18,CX
data/IAG/sales_index.txt
day1.csv
day2.csv
day3.csv
data/NAB/sales_index.txt