CS 1112 Exercise 9

Not string but chars

We are using the type char in Matlab, not string. What we call a “string” is really an array of chars. Type each of the following statements in the Command Window and note the result.

a= pi;   % A numeric scalar
b= 'pi'  % A char array. Use SINGLE quotes to enclose a char or multiple chars

c= length(b)            % __________  b is an array, so one can use function length on it

d= ['apple '  b  'es']  % Vector concatenation. d should be the string 'apple pies'

e= [d; 'muffin']        % __________________________________________

e= [d; 'mmmuffins ']    % Note the two extra 'm's and one trailing space

[nr,nc]= size(e)        % _____________ e is a matrix, so one can use function size on it

f= e(1, 7:9)            % _________________________________________ Accessing a subarray

d(1, 7:10)= 'core'      % _________________________________________

h= 'P' - 'M';           % __________ How many letters is 'P' after 'M'?

ii= char('d' + 4);      % __________ What comes 4 letters after 'd'?

jj= ii>'A' && ii<'Z'    % __________ True or false: character stored in ii is upper case

k=  ii>'a' && ii<'z'    % __________ True or false: character stored in ii is lower case

L= strcmp('abcd', 'ab') % ________________________________ strcmp compares the arguments

m= 'abcd'=='ab'         % ERROR: attempted vectorized code on vectors of different lengths

n= 'abcd'=='abCd'       % __________________________ Vectorized code--result is a vector

o= sum('abcd'=='abCd')  % __________________________ The number of matches

n= sum('abcd'~='abCd')  % __________________________ The number of mismatches

Reverse complement

In the DNA double helix, two strands twist together and “face” each other. The two strands are reverse-complementary, i.e., reading one strand in reverse order and exchanging each base with its complement gives the other strand. ‘A‛ and ‘T‛ are complementary; ‘C‛ and ‘G‛ are complementary.

For example, given the DNA sequence AGTAGCAT
the reverse sequence is TACGATGA
so the reverse complement is ATGCTACT

(a) Write a function rComplement(dna) to return the reverse complement of a DNA strand. Use a for-loop to reverse the strand—do not use vectorized code. dna is a vector of characters. Assume that dna contains only the letters ’A’, ’T’, ’C’, and ’G’. If dna is the empty vector, then return the empty vector. Do not use any built-in functions other than length() and size().

(b) Write a function rCompBulk(mat) to return the reverse complements of a set of DNA strands. mat is a matrix of characters; each row of the matrix represents one strand of DNA (so mat contains only the letters ‘A’, ‘T’, ‘C’, and ‘G’). Return a matrix the same size as mat such that the rth row of the returned matrix is the reverse complement of the rth strand of DNA (the rth row of mat). Again use for-loops—do not use vectorized code. In order to get practice with 2D char array syntax, write code to work on 2D array mat directly—do not call function rComplement() from part (a) above.

Counting a DNA pattern

Write a function countPattern(dna,p) that returns how many times a pattern p occurs in dna. Assume both parameters to be char vectors that contain the letters ‘A’, ‘T’, ‘C’, and ‘G’ only. Note that if p is longer than dna, then p appears in dna zero times. Use a for-loop to solve this problem.

(a) Version 1: Use the built-in function strcmp() to compare two char vectors.

(b) Version 2: Do not use strcmp(); instead use vectorized code and sum() as demonstrated in Part 1 above to compare two char vectors.

Cell array vs. vector

You already know that a vector is a collection of simple data. For example, you can have a vector of numbers (each component stores a single number) or a vector of characters (each component stores a single character). In a cell array, each cell can store an item that may be more complex than just a number or a character.

Type the following code in the command window and observe the output and the display in the Workspace pane. Also read the comments given below.

v= rand(1,4)  % a VECTOR of length four, each cell stores ONE number
v(3)          % Notice that you use PARENTHESES to access a cell in a VECTOR

c= cell(1,4)  % c is a CELL ARRAY.  c's "class" in the Workspace pane is "cell."
              % Right now each cell has an empty vector.

c{2}= v       % Put a VECTOR inside the 2nd cell of the CELL ARRAY.  Notice that we use
              %   CURLY BRACKETS to access contents of a cell in a CELL ARRAY.

c(3)= 1       % Error: cannot change the type cell (left side) to type double (right side).
              %   c(3) gets the cell, not the contents INSIDE the cell.  Must use curly
              %   brackets to access the contents inside a cell of a CELL ARRAY.

c{2}          % Display what is inside cell 2 of CELL ARRAY c:  a vector!

c(2)          % Display the cell (the "wrapper"), not what is inside the cell

% So how do you display, say, the fourth value in the VECTOR in the 2nd cell of CELL ARRAY c?
c{2}(4)       % Once again, use curly brackets for the index of the CELL ARRAY; use
              %   parentheses for the index of the of VECTOR.

c{1}= 'cat'   % OK for individual cells of a cell array to have different types
c{3}= 10
c{4}= ones(2,1)

% An alternate way to create a cell array is to specify all the contents inside CURLY
%   BRACKETS using spaces, commas, or semi-colons as the separator:
d= {'cat'; 10; v; ones(2,1)}  % A cell array of four cells
e= length(d)                  % The length function works for cell arrays as well.

Deck of cards

Download the functions CardDeck() and DispCards() from the course website. Read the code and try calling the functions to make sure that you understand them. Ask if you have questions. Implement the following function as specified:

function [c, sd]= MyShuffle(d)
% d is a one-dimensional cell array representing a deck of cards.
% sd is the cell array after shuffling d.
% c is the "cut point," explained below.
% The shuffle comprises two steps:
%  - Cut the deck into two parts:  generate a random integer value such that it is equally 
%    likely to be any value in [15..38].  This value is c, the first return parameter.  
%    The Top subdeck gets the cards at positions 1 to c of the original deck; 
%    the Bot subdeck gets the cards at positions c+1 to n of the original deck.
%  - Interleave the cards from Top and Bot subdecks in this order:
%    Top{1}, Bot{1}, Top{2}, Bot{2}, ..., Top{m}, Bot{m}, where m is the smaller value 
%    between c and 52-c.  (I.e., interleave the cards from the two subdecks until the 
%    smaller subdeck has been completely incorporated.)  Then put any remaining cards
%    (from the larger subdeck) at the end of the shuffled deck.
% Use built-in functions rand and floor (or ceil) for generating the cut point. 
% Write NON-vectorized code.

Hint on generating a random cut point: Only one of these two expressions is correct: floor(rand*(38-15+1))+15 or floor(rand*(38-15))+15. Do you know which one is correct? Why?

In order to practice indexing and cell array syntax, do not try to vectorize your code. I.e., work with one cell at a time. During code development, you can call the given DispCards() inside the function MyShuffle() to show the cell array being built in order to confirm that the intermediate steps are correct. Just remove the calls to DispCards() after you’ve completed the code.