What is wrong with this command line?

function y = cos_maclaurin(x,n)
%%This function will iterate through the values in the array x and for each
%%x-value it will perform a series summation for the cosine function using
%%n-terms in the series.  The function will then output corresponding
%%y-values for each x-value.
y = 0;
for h = 0:n
a = (-1)^(h)*((x^(2*h))/(factorial(2*h)));
y = a + y;
end

line 3 = ???Undefined function or variable h

[Greg Heath]

Factorials take a lot of time for large n.

What is your error message?

[Star Strider]

Well, for a start you didn't define n or x. When I defined n = 15 and x various values from (0 2*pi), it produced acceptable values for cos(x).

I assume that's what you want it to do.

[Image Analyst]

Well, you don't need the semicolon at the end of the "for" line of code. And you have some unnecessary parentheses, and you have inconsistent use of spaces around mathematical operators. But none of those are really wrong - just a bit sloppy. If you were trying to do a Maclaurin series for cosine, see my demo below.

Answer 1

[Matt Fig]

Assuming x and n are defined before hand, there is nothing wrong with that command line. What makes you think something is wrong?

.

.

EDIT

How are you calling that function? Because when I call it I get no error.

>> cos_maclaurin(1.5,6)

ans =

          0.0707372049851851

Also, if you really want to return the results for an array input, you should put your code in a loop (at the least - other improvements could be made):

function Y = cos_maclaurin(x,n)
% help goes here....

% n = n(1);  % And other input checking goes here.  Suggestions:
%  isnumeric(x), isempty(x), same for n, etc.

Y = zeros(size(x));
for ii = 1:numel(x)
    y = 0;
    for h = 0:n
        a = (-1)^(h)*((x(ii)^(2*h))/(factorial(2*h)));
        y = a + y;
    end
    Y(ii) = y;
end

With the above code you can do:

>> cos_maclaurin([1.5 3],60)

ans =

          0.0707372016677029        -0.989992496600446

Further enhancements might include the ability to enter a single n or an array n the same size as x, as well as efficiency improvements. Good luck!

[Walter Roberson]

Assuming they are defined as scalars. There is a problem if x is a vector or matrix.

[Matt Fig]

... or a cell array or a struct or..... True enough.

Answer 2

[Image Analyst]

Brian, I think this is what you were trying to achieve:

clc;
clearvars;
format compact
workspace;

y = 0;
numberOfTerms = 10;
index = 1;
for x = 0 : 0.1 : (3 * pi)
  theSum = 0;
  for h = 0 : numberOfTerms
    thisTerm = (-1) ^ h * (x ^ (2 * h)) / factorial(2 * h)
    theSum = theSum + thisTerm
  end
  y(index) = theSum;
  index = index + 1;
end
x = 0 : 0.1 : (3 * pi);
plot(x, y, 'bo-', 'LineWidth', 2);
hold on;
plot(x, cos(x), 'r-', 'LineWidth', 2);
grid on;
title('Maclaurin Series for Cosine', 'FontSize', 20);
xlabel('x', 'FontSize', 20);
ylabel('y', 'FontSize', 20);
legend('Maclaurin Estimate', 'Cosine');
set(gcf, 'units','normalized','outerposition',[0 0 1 1]);