I have written a function that estimates the inverse of e and loops through values of n until the approximated value is within a given tolerance of the actual value.
Currently I use this code:
function [approx, n] = calc_e(tolerance)
for n = 1:inf
approx = ((1-1/n)^n);
diff = (1/exp(1)) - approx;
if diff < tolerance, break; end
end
end
This works fine however I have been told that it could be more efficient by using a while loop but I can't work out how to do it in that way.
Can anybody shed some light on this?
Simply do:
function [approx, n] = calc_e(tolerance)
n = 1;
while (1/exp(1)) - ((1-1/n)^n) >= tolerance
n = n + 1;
end
end
Related
I recently stumbled upon a performance problem while implementing a simulation algorithm. I managed to find the bottleneck function (signally, it's the internal call to arrayfun that slows everything down):
function sim = simulate_frequency(the_f,k,n)
r = rand(1,n); %
x = arrayfun(#(x) find(x <= the_f,1,'first'),r);
sim = (histcounts(x,[1:k Inf]) ./ n).';
end
It is being used in other parts of code as follows:
h0 = zeros(1,sims);
for i = 1:sims
p = simulate_frequency(the_f,k,n);
h0(i) = max(abs(p - the_p));
end
Here are some possible values:
% Test Case 1
sims = 10000;
the_f = [0.3010; 0.4771; 0.6021; 0.6990; 0.7782; 0.8451; 0.9031; 0.9542; 1.0000];
k = 9;
n = 95;
% Test Case 2
sims = 10000;
the_f = [0.0413; 0.0791; 0.1139; 0.1461; 0.1760; 0.2041; 0.2304; 0.2552; 0.2787; 0.3010; 0.3222; 0.3424; 0.3617; 0.3802; 0.3979; 0.4149; 0.4313; 0.4471; 0.4623; 0.4771; 0.4913; 0.5051; 0.5185; 0.5314; 0.5440; 0.5563; 0.5682; 0.5797; 0.5910; 0.6020; 0.6127; 0.6232; 0.6334; 0.6434; 0.6532; 0.6627; 0.6720; 0.6812; 0.6901; 0.6989; 0.7075; 0.7160; 0.7242; 0.7323; 0.7403; 0.7481; 0.7558; 0.7634; 0.7708; 0.7781; 0.7853; 0.7923; 0.7993; 0.8061; 0.8129; 0.8195; 0.8260; 0.8325; 0.8388; 0.8450; 0.8512; 0.8573; 0.8633; 0.8692; 0.8750; 0.8808; 0.8864; 0.8920; 0.8976; 0.9030; 0.9084; 0.9138; 0.9190; 0.9242; 0.9294; 0.9344; 0.9395; 0.9444; 0.9493; 0.9542; 0.9590; 0.9637; 0.9684; 0.9731; 0.9777; 0.9822; 0.9867; 0.9912; 0.9956; 1.000];
k = 90;
n = 95;
The scalar sims must be in the range 1000 1000000. The vector of cumulated frequencies the_f never contains more than 100 elements. The scalar k represents the number of elements in the_f. Finally, the scalar n represents the number of elements in the empirical sample vector, and can even be very large (up to 10000 elements, as far as I can tell).
Any clue about how to improve the computation time of this process?
This seems to be slightly faster for me in the second test case, not the first. The time differences might be larger for longer the_f and larger values of n.
function sim = simulate_frequency(the_f,k,n)
r = rand(1,n); %
[row,col] = find(r <= the_f); % Implicit singleton expansion going on here!
[~,ind] = unique(col,'first');
x = row(ind);
sim = (histcounts(x,[1:k Inf]) ./ n).';
end
I'm using implicit singleton expansion in r <= the_f, use bsxfun if you have an older version of MATLAB (but you know the drill).
Find then returns row and column to all the locations where r is larger than the_f. unique finds the indices into the result for the first element of each column.
Credit: Andrei Bobrov over on MATLAB Answers
Another option (derived from this other answer) is a bit shorter but also a bit more obscure IMO:
mask = r <= the_f;
[x,~] = find(mask & (cumsum(mask,1)==1));
If I want performance, I would avoid arrayfun. Even this for loop is faster:
function sim = simulate_frequency(the_f,k,n)
r = rand(1,n); %
for i = 1:numel(r)
x(i) = find(r(i)<the_f,1,'first');
end
sim = (histcounts(x,[1:k Inf]) ./ n).';
end
Running 10000 sims with the first set of the sample data gives the following timing.
Your arrayfun function:
>Elapsed time is 2.848206 seconds.
The for loop function:
>Elapsed time is 0.938479 seconds.
Inspired by Cris Luengo's answer, I suggest below:
function sim = simulate_frequency(the_f,k,n)
r = rand(1,n); %
x = sum(r > the_f)+1;
sim = (histcounts(x,[1:k Inf]) ./ n)';
end
Time:
>Elapsed time is 0.264146 seconds.
You can use histcounts with r as its input:
r = rand(1,n);
sim = (histcounts(r,[-inf ;the_f]) ./ n).';
If histc is used instead of histcounts the whole simulation can be vectorized:
r = rand(n,sims);
p = histc(r, [-inf; the_f],1);
p = [p(1:end-2,:) ;sum(p(end-1:end,:))]./n;
h0 = max(abs(p-the_p(:))); %h0 = max(abs(bsxfun(#minus,p,the_p(:))));
I am trying to implement the Gauss-Seidel method in MATLAB. But there are two major mistakes in my code, and I could not fix them:
My code converges very well on small matrices, but it never converges on large matrices.
The code makes redundant iterations. How can I prevent from redundant iterations?
Gauss-Seidel Method on wikipedia.
N=5;
A=rand(N,N);
b=rand(N,1);
x = zeros(N,1);
sum = 0;
xold = x;
tic
for n_iter=1:1000
for i = 1:N
for j = 1:N
if (j ~= i)
sum = sum + (A(i,j)/A(i,i)) * xold(j);
else
continue;
end
end
x(i) = -sum + b(i)/A(i,i);
sum = 0;
end
if(abs(x(i)-xold(j))<0.001)
break;
end
xold = x;
end
gs_time=toc;
prompt1='Gauss-Seidel Method Time';
prompt2='x Matrix';
disp(prompt2);
disp(x);
disp(prompt1);
disp(gs_time);
First off, a generality. The Gauß-Seidel and Jacobi methods only apply to diagonally dominant matrices, not generic random ones. So to get correct test examples, you need to actually constructively ensure that condition, for instance via
A = rand(N,N)+N*eye(N)
or similar.
Else the method will diverge towards infinity in some or all components.
Now to some other strangeness in your implementation. What does
if(abs(x(i)-xold(j))<0.001)
mean? Note that this instruction is outside the loops where i and j are the iteration variables, so potentially, the index values are undefined. By inertia they will accidentally both have the value N, so this criterion makes at least a little sense.
What you want to test is some norm of the difference of the vectors as a whole, thus using sum(abs(x-xold))/N or max(abs(x-xold)). On the right side you might want to multiply with the same norm construction applied to x so that the test is for the relative error, taking the scale of the problem into account.
By the instructions in the given code, you are implementing the Jacobi iteration, computing all the updates first and then advancing the iteration vector. For the Gauß-Seidel variant you would need to replace the single components in-place, so that newly computed values are immediately used.
Also, you could shorten/simplify the inner loop
xold = x;
for i = 1:N
sum = b(i);
for j = 1:N
if (j ~= i)
sum = sum - A(i,j) * x(j);
end
end
x(i) = sum/A(i,i);
end
err = norm(x-xold)
or even shorter using the language features of matlab
xold = x
for i = 1:N
J = [1:(i-1) (i+1):N];
x(i) = ( b(i) - A(i,J)*x(J) )/A(i,i);
end
err = norm(x-xold)
%Gauss-seidal method for three equations
clc;
x1=0;
x2=0;
x3=0;
m=input('Enter number of iteration');
for i=1:1:m
x1(i+1)=(-0.01-0.52*x2(i)-x3(i))/0.3
x2(i+1)=0.67-1.9*x3(i)-0.5*x1(i+1)
x3(i+1)=(0.44-0.1*x1(i+1)-0.3*x2(i+1))/0.5
er1=abs((x1(i+1)-x1(i))/x1(i+1))*100
er2=abs((x2(i+1)-x2(i))/x2(i+1))*100
er3=abs((x3(i+1)-x3(i))/x3(i+1))*100
if er1<=0.01
er2<=0.01
er3<=0.01
break;
end
end
Hey i have an issuse with plotting my own function in scilab.
I want to plot the following function
function f = test(n)
if n < 0 then
f(n) = 0;
elseif n <= 1 & n >= 0 then
f(n) = sin((%pi * n)/2);
else
f(n) = 1;
end
endfunction
followed by the the console command
x = [-2:0.1:2];
plot(x, test(x));
i loaded the function and get the following error
!--error 21
Invalid Index.
at line 7 of function lala called by :
plot(x, test(x))
Can you please tell me how i can fix this
So i now did it with a for loop. I don't think it is the best solution but i can't get the other ones running atm...
function f = test(n)
f = zeros(size(n));
t = length(n);
for i = 1:t
if n(i) < 0 then
f(i) = 0;
elseif n(i) <= 1 & n(i) >= 0
f(i) = sin((%pi * n(i)/2));
elseif n(i) > 1 then
f(i) = 1;
end
end
endfunction
I guess i need to find a source about this issue and get used with the features and perks matlab/scilab have to over :)
Thanks for the help tho
The original sin is
function f = test(n)
(...)
f(n) = (...)
(...)
endfunction
f is supposed to be the result of the function. Therefore, f(n) is not "the value that the function test takes on argument n", but "the n-th element of f". Scilab then handles this however it can; on your test case, it tries to access a non-integer index, which results in an error. Your loop solution solves the problem.
Replacing all three f(n) by f in your first formulation makes it into something that works... as long as the argument is a scalar (not an array).
If you want test to be able to accept vector arguments without making a loop, the problem is that n < 0 is a vector of the same size as n. My solution would use logical arrays for indexing each of the three conditions:
function f = test(n)
f = zeros(size(n));
negative = (n<0);//parentheses are optional, but I like them for readability
greater_than_1 = (n>1);
others = ~negative & ~greater_than_1;
f(isnegative)=0;
f(greater_than_1)=1;
f(others) = sin(%pi/2*n(others));
endfunction
Here is the code I have so far :
function [u]=example222(xrange,trange,uinit,u0bound,uLbound);
n = length(xrange);
m = length(trange);
u = zeros(n,m); %%%
Dx = (xrange(n)-xrange(1))/(n-1);
Dt = (trange(m)-trange(1))/(m-1);
u(:,1)=uinit';
u(1,:)=u0bound;
u(n,:)=uLbound;
gegu=0.08;
alpha=0;
koefa=(-Dt*gegu/(2*Dx));
koefb=(alpha*Dt/(Dx)^2);
u
% first time step
for i = 2:n-1
u(i,2) = u(i,1)+2*koefa*(u(i+1,1)-u(i-1,1))+(koefb/2)*(u(i-1,1)-2*u(i,1)+u(i+1,1))
end;
% subsequent time steps
for j = 2:m-1
for i = 2:n-1
u(i,j+1)=u(i,j-1)+koefa*(u(i+1,j)-u(i-1,j))+koefb*(u(i-1,j)-2*u(i,j)+u(i+1,j))
end;
end;
______________________________________
x = (0:0.1:1);
t = (0:0.8:8) ;
u0=zeros;uL=zeros;
uinit=1-(10*x-1).^2;
[u]=example222(x,t,uinit,u0,uL);
surf(x,t,u,'EdgeColor','black')
Next thing I need to do is to implement an interval for uinit=1-(10*x-1).^2
IF x-0.08*t < 0.2. then => uinit=1-(10*x-1).^2;
else uinit=0;
Can someone help me with that please. I was trying to do it with if clauses and loops and couldn't make it work. Help is greatly appreciated.
There are many ways to define a piecewise function my usual method is as follows:
uinit = zeros(size(x));
I = x<(0.08*t+0.2); %// Find the indices where x<(0.08*t+0.2)
uinit(I) = 1-(10*x(I)-1).^2;
This case is easier than what you may often have since you want all the other values of uinit to be zero. If you had another function in the other region (say x^2) you could also do:
uinit(1-I) = x(1-I).^2;
The operation 1-I gives 0 where I==1 and 1 where I==0, therefore your get the complement of I.
I'm trying to optimize the performance (e.g. speed) of my code. I 'm new to vectorization and tried myself to vectorize, but unsucessful ( also try bxsfun, parfor, some kind of vectorization, etc ). Can anyone help me optimize this code, and a short description of how to do this?
% for simplify, create dummy data
Z = rand(250,1)
z1 = rand(100,100)
z2 = rand(100,100)
%update missing param on the last updated, thanks #Bas Swinckels and #Daniel R
j = 2;
n = length(Z);
h = 0.4;
tic
[K1, K2] = size(z1);
result = zeros(K1,K2);
for l = 1 : K1
for m = 1: K2
result(l,m) = sum(K_h(h, z1(l,m), Z(j+1:n)).*K_h(h, z2(l,m), Z(1:n-j)));
end
end
result = result ./ (n-j);
toc
The K_h.m function is the boundary kernel and defined as (x is scalar and y can be vector)
function res = K_h(h, x,y)
res = 0;
if ( x >= 0 & x < h)
denominator = integral(#kernelFunc,-x./h,1);
res = 1./h.*kernelFunc((x-y)/h)/denominator;
elseif (x>=h & x <= 1-h)
res = 1./h*kernelFunc((x-y)/h);
elseif (x > 1 - h & x <= 1)
denominator = integral(#kernelFunc,-1,(1-x)./h);
res = 1./h.*kernelFunc((x-y)/h)/denominator;
else
fprintf('x is out of [0,1]');
return;
end
end
It takes a long time to obtain the results: \Elapsed time is 13.616413 seconds.
Thank you. Any comments are welcome.
P/S: Sorry for my lack of English
Some observations: it seems that Z(j+1:n)) and Z(1:n-j) are constant inside the loop, so do the indexing operation before the loop. Next, it seems that the loop is really simple, every result(l, m) depends on z1(l, m) and z2(l, m). This is an ideal case for the use of arrayfun. A solution might look something like this (untested):
tic
% do constant stuff outside of the loop
Zhigh = Z(j+1:n);
Zlow = Z(1:n-j);
result = arrayfun(#(zz1, zz2) sum(K_h(h, zz1, Zhigh).*K_h(h, zz2, Zlow)), z1, z2)
result = result ./ (n-j);
toc
I am not sure if this will be a lot faster, since I guess the running time will not be dominated by the for-loops, but by all the work done inside the K_h function.