I tried to make a 3dof(3 degrees of freedom mechanism) in matlab but i get this error and i don't know why.
this is for a school project and i need to simulate a human finger.
the code is running normal but after i enter the values for the angles it says that A,B,C are
undefined and i don't know why
a1 = input('valuarea lui q1(grade):');
a2 = input('valuarea lui q2(grade):');
a3 = input('valuarea lui q3(grade):');
L1=35;
L2=45;
L3=30;
z = [-10 10];
plot(z,10);
grid ON;
O=[0;0;0;1];
m= linspace(pi/2,pi/2+a1*pi/180,100);
n = linspace(-pi/2,a2*pi/180,100);
k=linspace(-pi/2,a3*pi/180,100);
for a=1:100
[A1,B1,C1] = Transform(m(a),n(a),k(a),L1,L2,L3);
x = [O(1) A(1) B(1) C(1)];
y = [O(2) A(2) B(2) C(2)];
Cx(i)= C1(1);
Cy(i) = C1(2);
i=i+1;
Plot = plot(x,y,'r',...1
'LineWidth',1);
title('Sumularea unui deget');
plot(Cx,Cy,'--g',...
'LineWidth',1);
pause(0.075);
delete(Plot);
end
plot(x,y,'r',...
'LineWidth',3);
function [A,B,C ] = Transform( m,n,p,l1,l2,l3 )
P = [0;0;0;1];
T1 = [cos(m) -sin(m) 0 0;sin(m) cos(m) 0 0;0 0 1 0; 0 0 0 1];
T2 = [cos(n) -sin(n) 0 11;sin(n) cos(n) 0 0;0 0 1 0; 0 0 0 1];
T3 = [cos(p) -sin(p) 0 12;sin(p) cos(p) 0 0;0 0 1 0; 0 0 0 1];
T4 = [1 0 0 13;0 1 0 0; 0 0 1 0; 0 0 0 1];
A = T1*T2*P;
B = T1*T2*T3*P;
C = T1*T2*T3*T3*P;
end
In your main function, this is the first use of the variables:
x = [O(1) A(1) B(1) C(1)];
They are never written previously. Instead A1 is written, which is a different variable. I guess you mixed the two up.
Related
I am writing a function in Matlab to model the length of stay in hospital of stroke patients. I am having difficulty in storing my output values.
Here is my function:
function [] = losdf(age, strokeType, dest)
% function to mdetermine length of stay in hospitaal of stroke patients
% t = time since admission (days);
% age = age of patient;
% strokeType = 1. Haemorhagic, 2. Cerebral Infarction, 3. TIA;
% dest = 5.Death 6.Nursing Home 7. Usual Residence;
alpha1 = 6.63570;
beta1 = -0.03652;
alpha2 = -3.06931;
beta2 = 0.07153;
theta0 = -8.66118;
theta1 = 0.08801;
mu1 = 22.10156;
mu2 = 2.48820;
mu3 = 1.56162;
mu4 = 0;
nu1 = 0;
nu2 = 0;
nu3 = 1.27849;
nu4 = 0;
rho1 = 0;
rho2 = 11.76860;
rho3 = 3.41989;
rho4 = 63.92514;
for t = 1:1:365
p = (exp(-exp(theta0 + (theta1.*age))));
if strokeType == 1
initialstatevec = [1 0 0 0 0 0 0];
elseif strokeType == 2
initialstatevec = [0 1 0 0 0 0 0];
else
initialstatevec = [0 0 (1-p) p 0 0 0];
end
lambda1 = exp(alpha1 + (beta1.*age));
lambda2 = exp(alpha2 + (beta2.*age));
Q = [ -(lambda1+mu1+nu1+rho1) lambda1 0 0 mu1 nu1 rho1;
0 -(lambda2+mu2+nu2+rho2) lambda2 0 mu2 nu2 rho2;
0 0 -(mu3+nu3+rho3) 0 mu3 nu3 rho3;
0 0 0 -(mu4+nu4+rho4) mu4 nu4 rho4;
0 0 0 0 0 0 0;
0 0 0 0 0 0 0;
0 0 0 0 0 0 0];
Pt = expm(t./365.*Q);
Pt = Pt(strokeType, dest);
Ft = sum(initialstatevec.*Pt);
Ft
end
end
Then to run my function I use:
losdf(75,3,7)
I want to plot my values of Ft in a graph from from 0 to 365 days. What is the best way to do this?
Do I need to store the values in an array first and if so what is the best way to do this?
Many ways to do this, one straightforward way is to save each data point to a vector while in the loop and plot that vector after you exit your loop.
...
Ft = zeros(365,1); % Preallocate Ft as a vector of 365 zeros
for t = 1:365
...
Ft(t) = sum(initialstatevec.*Pt); % At index "t", store your output
...
end
plot(1:365,Ft);
I want to make second matrix from the first matrix in MATLAB. Please see the following image:
AB = A and B
AC = A and C
BC = B and C
====>>>
code:
transactions={{'A','C'};{'A','B'};{'A','B','C'}};
items = unique([transactions{:}]); % A,B,C
for i = 1:size(transactions,1)
T(i,ismember(items,transactions{i,:})) = 1; %convert transactions to matrix
end
.
.
T1 = zeros(size(transactions,1), nchoosek(length(items),2));
for k=1:5
for i=1: length(items)
for j=i+1 : length(items)
z = bitand(T(k,i),T(k,j)))
% set z in matrix T1 %
end
end
end
How to set the new values and How to concate label in the result matrix?
Simply use Matlab's & operator (see documentation). Because Matlab is an interpreted language, it will automatically convert the 0's and 1's to a logical matrix.
For example:
X = [1 0 1; 1 0 0; 1 1 1; 1 0 0; 1 1 0];
Y(:,1) = X(:,1) & X(:,2);
Y(:,2) = X(:,1) & X(:,3);
Y(:,3) = X(:,2) & X(:,3);
Matrix Y will be of type logical and be the desired output:
>> Y
Y =
0 1 0
0 0 0
1 1 1
0 0 0
1 0 0
Alternatively, you can use a struct for readability:
>> X.A = [1 1 1 1 1];
>> X.B = [0 0 1 0 1];
>> X.C = [1 0 1 0 0];
>> Y.AB = X.A & X.B;
>> Y.AC = X.A & X.C;
>> Y.BC = X.B & X.C;
>> Y
Y =
AB: [0 0 1 0 1]
AC: [1 0 1 0 0]
BC: [0 0 1 0 0]
If you want to automate the process, you could do the following:
X.A = [1 1 1 1 1]';
X.B = [0 0 1 0 1]';
X.C = [1 0 1 0 0]';
names = fieldnames(X);
N = length(names);
combos = nchoosek(1:N,2);
for i=1:N
Y.(char([names(combos(i,1)) names(combos(i,2))])) = ...
X.(char(names(combos(i,1)))) & X.(char(names(combos(i,2))));
end
struct2table(Y)
Which gives the following:
ans =
AB AC BC
_____ _____ _____
false true false
false false false
true true true
false false false
true false false
Bottomline:
Matlab throws the errors below and it is not obvious to me what is the root cause. The problem seems to reside in the input arguments, but I cannot figure out exactly what it is. I would greatly appreciate any help in finding it.
Index exceeds matrix dimensions.
Error in eliminatevariables (line 42)
aux(model.precalc.index2) = value(model.precalc.jj2);
Error in optimizer/subsref (line 276)
[self.model,keptvariablesIndex] =
eliminatevariables(self.model,self.model.parameterIndex,thisData(:),self.model.parameterIndex);
Error in SCMv0_justrun (line 68)
[solutions,diagnostics] = controller{inputs};
Background:
I am trying to program a Model Predictive Control but I am not very familiar yet with either Yalmip or mathematical optimization algorithms. I have made sure that the defined inputs and the actual inputs have the same dimensions, hence why I am surprised that the error has to do with matrix dimensions.
The error originates in when my code calls the optimizer.
My code is based on: https://yalmip.github.io/example/standardmpc/
Here is my code (the first part of the code is only needed to define the optimization problem and is marked between "%%%%%"; the error occurs near the end):
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
yalmip('clear')
clear all
% Model data
A = eye(3);
B = [1 0 -1 0 0 0 0; 0 1 0 -1 0 0 0; 0 0 0 0 1 -1 -1];
nx = 3; % Number of states
nu = 7; % Number of inputs
% MPC data
Q = [10 10 20]';
R = [10 10 1 1 5 3 3]';
C = [50 0; 0 30];
N = 90;
ny = 2;
E = [0 0 0 0 0 1 0; 0 0 0 0 0 0 1];
u = sdpvar(repmat(nu,1,N),repmat(1,1,N));
x = sdpvar(repmat(nx,1,N+1),repmat(1,1,N+1));
r = sdpvar(repmat(ny,1,N+1),repmat(1,1,N+1));
d = sdpvar(ny,1);
pastu = sdpvar(nu,1);
dx = 0.05;
Gx=[-1*eye(3);eye(3)];
gx = [0 0 0 500 500 1000]';
COVd = [zeros(5,7);0 0 0 0 0 10 0; 0 0 0 0 0 0 10];
COVx = zeros(nx,nx);
auxa = eye(5);
auxb = zeros(5,2);
Gu = [-1*eye(7,7); auxa auxb;0 0 0 0 0 1 1];
gu = [zeros(7,1); 200; 200; 50; 50; 100; 500];
Ga = [0 0 0.5 0.5 -1 0 0];
constraints = [];
objective = 0;
for k = 1:N
r{k} = r{k} + d;
objective = objective + Q'*x{k} + R'*u{k} + (r{k}-E*u{k})'*C*(r{k}-E*u{k});
constraints = [constraints, x{k+1} == A*x{k}+B*u{k}];
COVx = A*COVx*A' + B*COVd*B';
COVGx = Gx*COVx*Gx';
StDevGx = sqrt(diag(COVGx));
chance = gx - norminv(1-dx/(length (gx)*N))*StDevGx;
constraints = [constraints, Ga*u{k}==0, Gu*u{k}<=gu, Gx*x{k}<=gx-chance];
end
objective = objective + Q'*x{N+1};
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
parameters_in = {x{1},[r{:}],d,pastu};
solutions_out = {[u{:}], [x{:}]};
controller = optimizer(constraints, objective,[],parameters_in,solutions_out);
x = 100*ones(nx,1);
clf;
disturbance = randn(ny,1)*10;
oldu = zeros(nu,1);
hold on
for i = 1:150
future_r = [4*sin((i:i+N)/40);3*sin((i:i+N)/20)];% match dimensions of r
inputs = {x,future_r,disturbance,oldu};
[solutions,diagnostics] = controller{inputs};
U = solutions{1};oldu = U(1);
X = solutions{2};
if diagnostics == 1
error('The problem is infeasible');
end
x = A*x+B*u;
end
It's a bug in the latest version of YALMIP.
I am trying to make a parity check matrix from non-systematic to systematic. Hence, I am attaching my code below. Somewhat it is correct, but there are some problems. It would be really great if someone could help me in this.
Subject: Information theory and coding. I am working on LDPC coding and decoding. Please check the code below
MATLAB CODE:
H = [1 0 1 1 0; 0 0 1 0 1; 1 0 0 1 0; 1 0 1 1 1]
[m,n] = size(H);
k = n-m;
for i = k+1:n
%H(:,i)
ind = find(H(:,i),1,'last');
% exchanging (ind)th row and (i-k)th row
if ind < i-k
continue;
end
if ind ~= i-k
temp = H(ind,:);
H(ind,:) = H(i-k,:);
H(i-k,:) = temp;
end
I = find(H(:,i));
% Guassian elimination
for j = 1:length(I)
if I(j) ~= i-k
H(I(j),:) = mod(H(I(j),:)+H(i-k,:),2);
end
end
end
Hsys = H
For e.g.
This is my H matrix:
H =
1 0 1 1 0
0 0 1 0 1
1 0 0 1 0
1 0 1 1 1
I want to have an identity matrix inside the matrix. The dimension on H matrix here is (mxn) which is (4x5).
Generally we use Gaussian elimination method to make the Identity matrix.hence, we make operations between rows. This is how we make it systematic.
I should have matrix as this in the result:
Hsys =
0 1 0 0 0
0 0 1 0 0
1 0 0 1 0
0 0 0 0 1
I should have an identity matrix of dimension m.
Here is how I'd do it (using Gauss-Jordan elimination):
% Not your matrix since it does not have any ones in the second column.
H=[1 1 0 1 1 0 0 1 0 0;
0 1 1 0 1 1 1 0 0 0;
0 0 0 1 0 0 0 1 1 1;
1 1 0 0 0 1 1 0 1 0;
0 0 1 0 0 1 0 1 0 1];
rows = size(H, 1);
cols = size(H, 2);
r = 1;
for c = cols - rows + 1:cols
if H(r,c) == 0
% Swap needed
for r2 = r + 1:rows
if H(r2,c) ~= 0
tmp = H(r, :);
H(r, :) = H(r2, :);
H(r2, :) = tmp;
end
end
end
% Ups...
if H(r,c) == 0
error('H is singular');
end
% Forward substitute
for r2 = r + 1:rows
if H(r2, c) == 1
H(r2, :) = xor(H(r2, :), H(r, :));
end
end
% Back Substitution
for r2 = 1:r - 1
if H(r2, c) == 1
H(r2, :) = xor(H(r2, :), H(r, :));
end
end
% Next row
r = r + 1;
end
I want to generate a matrix that is "stairsteppy" from a vector.
Example input vector: [8 12 17]
Example output matrix:
[1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1]
Is there an easier (or built-in) way to do this than the following?:
function M = stairstep(v)
M = zeros(length(v),max(v));
v2 = [0 v];
for i = 1:length(v)
M(i,(v2(i)+1):v2(i+1)) = 1;
end
You can do this via indexing.
A = eye(3);
B = A(:,[zeros(1,8)+1, zeros(1,4)+2, zeros(1,5)+3])
Here's a solution without explicit loops:
function M = stairstep(v)
L = length(v); % M will be
V = max(v); % an L x V matrix
M = zeros(L, V);
% create indices to set to one
idx = zeros(1, V);
idx(v + 1) = 1;
idx = cumsum(idx) + 1;
idx = sub2ind(size(M), idx(1:V), 1:V);
% update the output matrix
M(idx) = 1;
EDIT: fixed bug :p
There's no built-in function I know of to do this, but here's one vectorized solution:
v = [8 12 17];
N = numel(v);
M = zeros(N,max(v));
M([0 v(1:N-1)]*N+(1:N)) = 1;
M(v(1:N-1)*N+(1:N-1)) = -1;
M = cumsum(M,2);
EDIT: I like the idea that Jonas had to use BLKDIAG. I couldn't help playing with the idea a bit until I shortened it further (using MAT2CELL instead of ARRAYFUN):
C = mat2cell(ones(1,max(v)),1,diff([0 v]));
M = blkdiag(C{:});
A very short version of a vectorized solution
function out = stairstep(v)
% create lists of ones
oneCell = arrayfun(#(x)ones(1,x),diff([0,v]),'UniformOutput',false);
% create output
out = blkdiag(oneCell{:});
You can use ones to define the places where you have 1's:
http://www.mathworks.com/help/techdoc/ref/ones.html