f



Matlab LVQ Help

Hi,

I am new on neural network toolbox. I am trying to use LVQ algorithm,but it is not working at all.
Right now I am trying to train the network for a simple sine curve, but it is not trained according to the target.
Can you help me with your suggestion,
my function is 
y=sin(x);
x is my input and y is my target output.
I classified target results in four different classes. 
0.5<=value<=1.0; class 1
0.0<=value<0.5;  class 2
-0.5<=value<0.0; class 3
-1<=value<-0.5;  class 4


I would highly appriciate if you help me with some idea, i am stucked here for a long time. 


Here is the code I am using,
---------------------------------------------------
t=-5:.1:5;
x=t;
for i=1:length(x)
  y(i)=sin(x(i));
end

for i=1:1:length(y)
  if(0.5<(y(1,i)) && (y(1,i))<=1)
  tc1(1,i)=1;
  end
end
for i=1:1:length(y)
  if(0<=(y(1,i)) && (y(1,i))<0.5)
  tc1(1,i)=2;
end
end

for i=1:1:length(y)
  if(-0.5<=(y(1,i)) && (y(1,i))<0)
  tc1(1,i)=3;
end
end

for i=1:1:length(y)
  if(-1.0<=(y(1,i)) && (y(1,i))<-0.5)
  tc1(1,i)=4;
end
end

p1=x;

check(1,:)=p1;
check(2,:)=tc1;

checkall=sortrows(check,2);

p=checkall(1,:);
tc=checkall(2,:);

t=ind2vec(tc);
targets=full(t);

sc=16;

calu(1,1)=sum(targets(1,:));
calu(2,1)=sum(targets(2,:));
calu(3,1)=sum(targets(3,:));
calu(4,1)=sum(targets(4,:));

c1=calu(1,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
c2=calu(2,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
c3=calu(3,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
c4=calu(4,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));

pc=[c1 c2 c3 c4];

net=newlvq(minmax(p),sc,pc);

net.inputWeights{1,1}.learnFcn='learnlv2';


%----------training------------
net.trainParam.epochs=1000;
net.trainParam.lr=0.05;

net=init(net);
net=train(net,p,targets);

y_output=sim(net,p);
yc=vec2ind(y_output);
%------------------------------------

Thanks

Rahil
0
1/16/2008 7:47:55 PM
comp.soft-sys.matlab 211264 articles. 26 followers. lunamoonmoon (257) is leader. Post Follow

4 Replies
1047 Views

Similar Articles

[PageSpeed] 16

On Jan 16, 11:47 am, rahil_1722 <rahil1...@yahoo.com> wrote:
> Hi,
>
> I am new on neural network toolbox. I am trying to use LVQ algorithm,but it is not working at all.
> Right now I am trying to train the network for a simple sine curve,
> but it is not trained according to the target.
> Can you help me with your suggestion,
> my function is
> y=sin(x);
> x is my input and y is my target output.
> I classified target results in four different classes.
> 0.5<=value<=1.0; class 1
> 0.0<=value<0.5;  class 2
> -0.5<=value<0.0; class 3
> -1<=value<-0.5;  class 4
>
> I would highly appriciate if you help me with some idea, i am stucked here for a long time.
>
> Here is the code I am using,
> ---------------------------------------------------

% NOTE1. You MUST clear the workspace
% NOTE2. Initialize RNGs INTEGER state (for run duplication)

close all, clear, clc
STATE = 0
rand('state',STATE)
randn('state',STATE)

> t=-5:.1:5;
> x=t;
> for i=1:length(x)
>   y(i)=sin(x(i));
> end

VECTORIZE!

y = sin(x);

> for i=1:1:length(y)
>   if(0.5<(y(1,i)) && (y(1,i))<=1)
>   tc1(1,i)=1;
>   end
> end
> for i=1:1:length(y)
>   if(0<=(y(1,i)) && (y(1,i))<0.5)
>   tc1(1,i)=2;
> end
> end
>
> for i=1:1:length(y)
>   if(-0.5<=(y(1,i)) && (y(1,i))<0)
>   tc1(1,i)=3;
> end
> end
>
> for i=1:1:length(y)
>   if(-1.0<=(y(1,i)) && (y(1,i))<-0.5)
>   tc1(1,i)=4;
> end
> end

For 1-D matrices only need one index: y(i) and tc(i)
For simpler logic
help elseif
help else

> p1=x;
>
> check(1,:)=p1;
> check(2,:)=tc1;
>
> checkall=sortrows(check,2);
>
> p=checkall(1,:);
> tc=checkall(2,:);

% NOTE3. t already defined as -5:0.1:5

> t=ind2vec(tc);
> targets=full(t);

% NOTE4. Kohonen, the creator of LVQ, recommends that
% the number of hidden nodes be determined by a clustering
% algorithm such as his SOM (AKA SOFM).
% In this simple 1-D example we see from plot(x,tc) that the
%  optimal value is
%
% sc = 10

> sc=16;
>
> calu(1,1)=sum(targets(1,:));
> calu(2,1)=sum(targets(2,:));
> calu(3,1)=sum(targets(3,:));
> calu(4,1)=sum(targets(4,:));
>
> c1=calu(1,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
> c2=calu(2,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
> c3=calu(3,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
> c4=calu(4,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));

Yuck. VECTORIZE!

> pc=[c1 c2 c3 c4];
>
> net=newlvq(minmax(p),sc,pc);

% NOTE5. Quote from help NEWLVQ:
% "LEARNLV2 should only be used to finish training
%  of networks already trained with LEARNLV1."
% LEARNV1 is default learning algorithm


> net.inputWeights{1,1}.learnFcn='learnlv2';
% net.inputWeights{1,1}.learnFcn='learnlv1';

> %----------training------------

% NOTE6. epochmax = 250  is more than sufficient
% NOTE7. default lr = 0.01

> net.trainParam.epochs=1000;
> net.trainParam.lr=0.05;
>
> net=init(net);

%NOTE8: Weights are automatically initialized to midpoint
% values of p in the newlvq command. You can check if init
% does anything different or better by using IW = net.IW
% before and after net = init(net)

% NOTE9. Kohonen, the creator of LVQ, recommends that
% the number of hidden nodes AND the initial weights be
% determined by a clustering algorithm such as his SOM
% (AKA SOFM).
% In this simple 1-D example you can see from plot(x,tc)
% what good values for IW0 should be. Then initialize
% via
%
% net.IW = IW0;

> net=train(net,p,targets);

%NOTE10. Get warning from LEARNV2

% Warning: Divide by zero.
% (Type "warning off MATLAB:divideByZero" to suppress this warning.)
% > In C:\MATLAB6p5p1\toolbox\nnet\nnet\learnlv2.m at line 164
%   In C:\DOCUME~1\Gregory\LOCALS~1% \Temp\matlab_nnet\tp215164.m at
line 78
%   In C:\MATLAB6p5p1\toolbox\nnet\nnet\trainr.m at line 164
%   In C:\MATLAB6p5p1\toolbox\nnet\nnet\@network\train.m at line 278
%   In C:\MATLAB6p5p1\work\lvqexample0.m at line 109

> y_output=sim(net,p);
> yc=vec2ind(y_output);

%NOTE11. Added diagnostics
%NOTE12. MSEc = 2*MSE only if errors are between
% classes with indices differing by unity

err = targets-y_output;
MSE = mse(err)                              % 0.1287
Nerr = sum(sum(abs(err))/2)          % 26
Ntrn = length(p)                               % 101
pcterr = 100*Nerr/Ntrn                    % 25.7426

ec = tc-yc;
MSEc = mse(ec)                            % 0.2574
summary = [(1:Ntrn)' tc' yc' ec']

NOTE13. See my May 23 2007 post by searching with

greg-heath lvq glossary

Hope this helps.

Greg
0
heath (3983)
1/18/2008 6:57:48 PM
On Jan 18, 1:57=A0pm, Greg Heath <he...@alumni.brown.edu> wrote:
> On Jan 16, 11:47 am, rahil_1722 <rahil1...@yahoo.com> wrote:
> > Hi,
>
> > I am new on neural network toolbox. I am trying to useLVQalgorithm,but i=
t is not working at all.
> > Right now I am trying to train the network for a simple sine curve,
> > but it is not trained according to the target.
> > Can you help me with your suggestion,
> > my function is
> > y=3Dsin(x);
> > x is my input and y is my target output.
> > I classified target results in four different classes.
> > 0.5<=3Dvalue<=3D1.0; class 1
> > 0.0<=3Dvalue<0.5; =A0class 2
> > -0.5<=3Dvalue<0.0; class 3
> > -1<=3Dvalue<-0.5; =A0class 4
>
> > I would highly appriciate if you help me with some idea, i am stucked he=
re for a long time.
>
> > Here is the code I am using,
> > ---------------------------------------------------
>
> % NOTE1. You MUST clear the workspace
> % NOTE2. Initialize RNGs INTEGER state (for run duplication)
>
> close all, clear, clc
> STATE =3D 0
> rand('state',STATE)
> randn('state',STATE)
>
> > t=3D-5:.1:5;
> > x=3Dt;
> > for i=3D1:length(x)
> > =A0 y(i)=3Dsin(x(i));
> > end
>
> VECTORIZE!
>
> y =3D sin(x);
>
> > for i=3D1:1:length(y)
> > =A0 if(0.5<(y(1,i)) && (y(1,i))<=3D1)
> > =A0 tc1(1,i)=3D1;
> > =A0 end
> > end
> > for i=3D1:1:length(y)
> > =A0 if(0<=3D(y(1,i)) && (y(1,i))<0.5)
> > =A0 tc1(1,i)=3D2;
> > end
> > end
>
> > for i=3D1:1:length(y)
> > =A0 if(-0.5<=3D(y(1,i)) && (y(1,i))<0)
> > =A0 tc1(1,i)=3D3;
> > end
> > end
>
> > for i=3D1:1:length(y)
> > =A0 if(-1.0<=3D(y(1,i)) && (y(1,i))<-0.5)
> > =A0 tc1(1,i)=3D4;
> > end
> > end
>
> For 1-D matrices only need one index: y(i) and tc(i)
> For simpler logic
> help elseif
> help else
>
> > p1=3Dx;
>
> > check(1,:)=3Dp1;
> > check(2,:)=3Dtc1;
>
> > checkall=3Dsortrows(check,2);
>
> > p=3Dcheckall(1,:);
> > tc=3Dcheckall(2,:);
>
> % NOTE3. t already defined as -5:0.1:5
>
> > t=3Dind2vec(tc);
> > targets=3Dfull(t);
>
> % NOTE4. Kohonen, the creator ofLVQ, recommends that
> % the number of hidden nodes be determined by a clustering
> % algorithm such as his SOM (AKA SOFM).
> % In this simple 1-D example we see from plot(x,tc) that the
> % =A0optimal value is
> %
> % sc =3D 10
>
> > sc=3D16;
>
> > calu(1,1)=3Dsum(targets(1,:));
> > calu(2,1)=3Dsum(targets(2,:));
> > calu(3,1)=3Dsum(targets(3,:));
> > calu(4,1)=3Dsum(targets(4,:));
>
> > c1=3Dcalu(1,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
> > c2=3Dcalu(2,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
> > c3=3Dcalu(3,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
> > c4=3Dcalu(4,1)/(sum(calu(1,1)+calu(2,1)+calu(3,1)+calu(4,1)));
>
> Yuck. VECTORIZE!
>
> > pc=3D[c1 c2 c3 c4];
>
> > net=3Dnewlvq(minmax(p),sc,pc);
>
> % NOTE5. Quote from help NEWLVQ:
> % "LEARNLV2 should only be used to finish training
> % =A0of networks already trained with LEARNLV1."
> % LEARNV1 is default learning algorithm
>
> > net.inputWeights{1,1}.learnFcn=3D'learnlv2';
>
> % net.inputWeights{1,1}.learnFcn=3D'learnlv1';

Since LEARNLV1  is the default, just eliminate the
command.

> > %----------training------------
>
> % NOTE6. epochmax =3D 250 =A0is more than sufficient
> % NOTE7. default lr =3D 0.01
>
> > net.trainParam.epochs=3D1000;
> > net.trainParam.lr=3D0.05;
>
> > net=3Dinit(net);
>
> %NOTE8: Weights are automatically initialized to midpoint
> % values of p in the newlvq command. You can check if init
> % does anything different or better by using IW =3D net.IW

IW =3D cell2mat(net.IW)
LW =3D cell2mat(net.LW)

> % before and after net =3D init(net)
>
> % NOTE9. Kohonen, the creator ofLVQ, recommends that
> % the number of hidden nodes AND the initial weights be
> % determined by a clustering algorithm such as his SOM
> % (AKA SOFM).
> % In this simple 1-D example you can see from plot(x,tc)
> % what good values for IW0 should be. Then initialize
> % via
> %
> % net.IW =3D IW0;

This syntax doesn't work. I'll post a new thread to find out
how to do it.

> > net=3Dtrain(net,p,targets);
>
> %NOTE10. Get warning from LEARNV2
>
> % Warning: Divide by zero.
> % (Type "warning off MATLAB:divideByZero" to suppress this warning.)
> % > In C:\MATLAB6p5p1\toolbox\nnet\nnet\learnlv2.m at line 164
> % =A0 In C:\DOCUME~1\Gregory\LOCALS~1% \Temp\matlab_nnet\tp215164.m at
> line 78
> % =A0 In C:\MATLAB6p5p1\toolbox\nnet\nnet\trainr.m at line 164
> % =A0 In C:\MATLAB6p5p1\toolbox\nnet\nnet\@network\train.m at line 278
> % =A0 In C:\MATLAB6p5p1\work\lvqexample0.m at line 109
>
> > y_output=3Dsim(net,p);
> > yc=3Dvec2ind(y_output);
>
> %NOTE11. Added diagnostics
> %NOTE12. MSEc =3D 2*MSE only if errors are between
> % classes with indices differing by unity
>
> err =3D targets-y_output;
> MSE =3D mse(err) =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =
=A0% 0.1287
> Nerr =3D sum(sum(abs(err))/2) =A0 =A0 =A0 =A0 =A0% 26
> Ntrn =3D length(p) =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
 =A0 % 101
> pcterr =3D 100*Nerr/Ntrn =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0% 25.7426
>
> ec =3D tc-yc;
> MSEc =3D mse(ec) =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0% =
0.2574
> summary =3D [(1:Ntrn)' tc' yc' ec']
>
> NOTE13. See my May 23 2007 post by searching with
>
> greg-heathlvqglossary

Need spaces

greg-heath lvq glossary

Hope this helps.

Greg
0
heath (3983)
1/19/2008 5:48:03 PM
On Jan 19, 12:48=A0pm, Greg Heath <he...@alumni.brown.edu> wrote:
> On Jan 18, 1:57=A0pm, Greg Heath <he...@alumni.brown.edu> wrote:
>
> > On Jan 16, 11:47 am, rahil_1722 <rahil1...@yahoo.com> wrote:
-----SNIP
> > > net=3Dinit(net);
>
> > %NOTE8: Weights are automatically initialized to midpoint
> > % values of p in the newlvq command. You can check if init
> > % does anything different or better by using IW =3D net.IW
>
> IW =3D cell2mat(net.IW)
> LW =3D cell2mat(net.LW)
>
> > % before and after net =3D init(net)

I did it and init didn't change anything. So, delete the init command.

> > % NOTE9. Kohonen, the creator ofLVQ, recommends that
> > % the number of hidden nodes AND the initial weights be
> > % determined by a clustering algorithm such as his SOM
> > % (AKA SOFM).
> > % In this simple 1-D example you can see from plot(x,tc)
> > % what good values for IW0 should be. Then initialize
> > % via
> > %
> > % net.IW =3D IW0;
>
> This syntax doesn't work. I'll post a new thread to find out
> how to do it.

 If you want to change net.IW and net.LW, use

net.IW{1,1} =3D IW0;
net.LW{2,1} =3D LW0;

To understand the indices see my recent post by searching

greg-heath how to use mat2cell


-------SNIP

I, too, am not getting good training results with newlvq. However,
with newff, the training results are close to perfect.

My guess is that the automatically generated initial weights aren't
good enough. After all, Kohonen not only suggest initializing via
SOFM, he suggests  continuing training with LEARNLV2 *after*
training with LEARNLV1.

I'll stick NEWFF  and NEWRB (with my recently posted
modifications).

Hope this helps.

Greg
0
heath (3983)
1/20/2008 5:40:11 AM
On Jan 16, 11:47 am, rahil_1722 <rahil1...@yahoo.com> wrote:
> Hi,
>
> I am new on neural network toolbox. I am trying to use LVQ
> algorithm, but it is not working at all.
> Right now I am trying to train the network for a simple sine
> curve, but it is not trained according to the target.
> Can you help me with your suggestion,
> my function is
> y=sin(x);
> x is my input and y is my target output.
> I classified target results in four different classes.
> 0.5<=value<=1.0; class 1
> 0.0<=value<0.5;  class 2
> -0.5<=value<0.0; class 3
> -1<=value<-0.5;  class 4

Ntrn = [35 16 15 35]
Ntrn0 = sum(Ntrn) = 101
pclass = Ntrn/Ntrn0          % ~ [0.3465 0.1584 0.1485 0.3465]

From the plot of tc vs x it is obvious that the minimal
configuration should have

Numhidnodesmin = [2 3 3 2]
Hmin                     = sum(Numhidnodesmin)      % = 10
clusterratio           = [0.2 0.3 0.3 0.2]

with cluster centroids near the midpoints of the subclass intervals.

From the MATLAB source code (type newlvq) it looks like the
subclass decomposition of H is determined by

Numhidnodes = diff[0 floor(cumsum(pclass)*H)]

which is not monotonic in H. Therefore, the solutions to

Numhidnodes >= Numhidnodesmin

should be tabulated.

for i = 1:30

H = i
J(i,:) =  diff([  0 floor(cumsum(pclass)*H)  ]);
K(i,:) =  diff([  0 floor(cumsum(clusterratio)*H)  ]);

end

summary = [ (1:30)' J zeros(30,1) (1:30)' K]

summary =

     1     0     0     0     1     0     1     0     0     0     1
     2     0     1     0     1     0     2     0     1     0     1
     3     1     0     0     2     0     3     0     1     1     1
     4     1     1     0     2     0     4     0     2     1     1
     5     1     1     1     2     0     5     1     1     2     1
     6     2     1     0     3     0     6     1     2     1     2
     7     2     1     1     3     0     7     1     2     2     2
     8     2     2     1     3     0     8     1     3     2     2
     9     3     1     1     4     0     9     1     3     3     2
    10     3     2     1     4     0    10     2     3     3     2
    11     3     2     2     4     0    11     2     3     3     3
    12     4     2     1     5     0    12     2     4     3     3
    13     4     2     2     5     0    13     2     4     4     3
    14     4     3     2     5     0    14     2     5     4     3
    15     5     2     2     6     0    15     3     4     5     3
    16     5     3     2     6     0    16     3     5     4     4
    17     5     3     3     6     0    17     3     5     5     4
    18     6     3     2     7     0    18     3     6     5     4
    19     6     3     3     7     0    19     3     6     6     4
    20     6     4     3     7     0    20     4     6     6     4
    21     7     3     3     8     0    21     4     6     6     5
    22     7     4     3     8     0    22     4     7     6     5
    23     7     4     4     8     0    23     4     7     7     5
    24     8     4     3     9     0    24     4     8     7     5
    25     8     4     4     9     0    25     5     7     8     5
    26     9     4     3    10     0    26     5     8     7     6
    27     9     4     4    10     0    27     5     8     8     6
    28     9     5     4    10     0    28     5     9     8     6
    29    10     4     4    11     0    29     5     9     9     6
    30    10     5     4    11     0    30     6     9     9     6

which, obviously, doesn't allocate enough clusters for class 3
untill H = 17 when using pclass. However, because of the
nonmonotonicity, the solutions are

net = newlvq(minmax(p),H,pclass); (H = 17 or H >= 19)

and

net = newlvq(minmax(p),H,clusterratio); (H >= 10).

So far I still haven't obtained good results. I even tried
overwriting the initial cluster centers in net.IW{1,1} with
the midpoints of the subclass intervals.

Keep plugging and please let me know when you've
found the magic parameter settings.

Hope this helps.

Greg
0
heath (3983)
1/21/2008 9:14:26 PM
Reply:

Similar Artilces:

how could i use the matlab help without running matlab?
sometimes i just want to see some content of the help files and not intend to use matlab itself but as we know i must start matlab software first, then open the help product can i simply open the help document without running matlab? thanks! "fangfines " <fangfines@qq.com> wrote in message <i67kgh$qcr$1@fred.mathworks.com>... > sometimes i just want to see some content of the help files and not intend to use matlab itself > > but as we know i must start matlab software first, then open the help product > > can i simply open the help document without running matlab? > > thanks! Hi, The MATLAB help for the current release (R2010b) is available over the web. Choose a product such as the Signal Processing Toolbox. http://www.mathworks.com/products/signal/ and then under Support and Training, select the documentation. http://www.mathworks.com/help/toolbox/signal/index.html Wayne "Wayne King" <wmkingty@gmail.com> wrote in message <i67rfa$3bl$1@fred.mathworks.com>... > "fangfines " <fangfines@qq.com> wrote in message <i67kgh$qcr$1@fred.mathworks.com>... > > sometimes i just want to see some content of the help files and not intend to use matlab itself > > > > but as we know i must start matlab software first, then open the help product > > > > can i simply open the help document without running matlab? > > > > thanks! > > H...

help matlab help
how to automatical digitise profile to generate a vector model of the profile and how to use the digitised profile to sweep generate a graphical three dimensional model of the picture please give me some example ...........thanks very much..... ...

Need help for a matlab mini project
Hello everyone, I have a matlab in class exercise due in few days. I am new to matlab and the instructions provided with the assignment is minimal. Can Someone help me with the basic understanding as in how can I map the problem's solution to matlab script ? I am not able to post question here as it is a jpeg file. Please let me know and then we can discuss further. ...

urgent help needed: serial transmission from matlab to matlab
hi. really need help. i've to transmit an image serially between two computers using matlab on both transmitting and receiving ends. i've encoded the image into binary. been facing problems with the transmission when i used the command fprintf. when i executed that, i got a warning saying timeout occurred before terminator was reached. plus the receiving matlab hangs. i was told to save the binary packet into m-files then transmit the file over. how do i do that? is there a command to save and retrieve binary packets into and from m-files? i really need to get this done asap.thank you....

Trying to import multiple images into MATLAB and treat them as frames and overall help with matlab
Hello.I'm in desperate need of some help here, I want to import various images of clouds taken about 30 seconds in one another as a single video file the loop I used to originally do this consisted of: files = dir(Directory of files) for i = 1:100 str = strcat('Directory for workspace',files(i).name); clouds(i,:,:,:)=imread(str); end; But the images it produced were 4-d uint 8 and i was unable to read them or work with them at all. Please can you offer me some type of solution. The main goal of these processed cloud images is to be able to track the movement of clouds in the skys so if you have any demos that I could look at please post them On 28/12/10 8:49 PM, Linus Hamann wrote: > Hello.I'm in desperate need of some help here, I want to import various > images of clouds taken about 30 seconds in one another as a single video > file the loop I used to originally do this consisted of: > files = dir(Directory of files) > for i = 1:100 > str = strcat('Directory for workspace',files(i).name); > clouds(i,:,:,:)=imread(str); > end; > > But the images it produced were 4-d uint 8 and i was unable to read them > or work with them at all. Please can you offer me some type of solution. > The main goal of these processed cloud images is to be able to track the > movement of clouds in the skys so if you have any demos that I could > look at please post them That looks okay to me, except that it...

parallel matlab
Hi all, I'm going to work on a project involving a parallel version of MATLAB - I'm aware of various efforts like MATLAB-*P and so on Right now I need to read up on some material regarding MATLAB internals eg: How are the ATLAS(or is it BLAS) routines called in MATLAB? How does MATLAB's C code store information in memory, cache'ing etc? How is one to get at those? Do you know of any suitable reference(s)/resource(s)? Thank you very much, Aravind ...

matlab help
hi i am student and I need help in one part of a homework... the problem is: Implements a function of birth to two sets of training classes accordance with the following specifications: each set consists of education vectors from two characteristics are among the statistically independent and the feature follows the normal distribution. The four average prices and price dispersion are four arguments as the entry-related ing. Moreover, the function receives the number of entry points for each class, even M1 and M2. The function returns two tables, measuring 3 × M1 and 3 × M2, wher...

Matlab plots outside Matlab
I'm making an end-user application where I would like to use Matlab functionality both for calculation and plotting. Matlabs GUI builder guide does not satisfy my needs (no support for f.ex. internal frames and other nescessities), so I'm planning to compile my Matlab code and use other API (Java, QT or something) for GUI. My problem is then, that I loose access to the neat plotting commands availible under matlab GUI... Is there a way of integrating Matlabs plot functionality into another GUI system? - has it been done? - I appreciate any response! - P�l Berge ...

Matlab Help!!!!!!!!!!!!!!!
Dear All I am working in project using matlab gui... I finished them and run prefectly. Then I used Matlab Compiler to change to stand-alone C application and all my file works perfectly ..... except one give me error and I could not figure out what the hell is the problem!! Here are the errors: I use ==> mcc -g -B sgl rx_gui.m Error rx_gui.c 102: Type mismatch in redeclaration of 'system' Error C:\BC5\INCLUDE\stdlib.h 353: Earlier declaration of 'system' Error rx_gui.c 4356: Two operands myst evaluate to the same type in function Mrx_gui_rx_gui_OpeningFcn MBUILD.BAT...

Matlab Help?
I have been given an assignment and have no idea where to start. Please Help!!! 1. Open humpty_mixed.gif (2D image) and do what all of the king’s horses and men could not do – put humpty back together again. The image is 600x412 and split evenly along the rows and columns. (Note: We may use the imread(), imshow(), and imwrite() commands here.) * The picture is a scrambled picture of humpty dumpty that I am supposed to unmix. 2. Open humpty.jpg (3D image) and section off old humpty such that the top?half is red, bottom?left is green, and bottom?right is blue. For the top?half, force al...

matlab help
Hello guys i have to control 4 wheels by 4 dc motor through matlab. i am newbee in matlab, also i have to make respective GUI, how to proceed.. aditya wagh <adipatil110@gmail.com> wrote in message <4999e3da-44a3-463b-bc16-fcc521231536@googlegroups.com>... > Hello guys i have to control 4 wheels by 4 dc motor through matlab. > i am newbee in matlab, also i have to make respective GUI, how to proceed.. Start here: <<http://www.mathworks.com/help/matlab/getting-started-with-matlab.html>> ...

Help with matlab.
I'm failry new at matlab, and I was wondering if anyone could help me figure out a problem I have. I have the following assigned values: i=1:500; j=122+i; k=j+1; l=i+1; What I want is to display my anwers in 4 columns, i j k l, and also somehow delete each line with a specific i value. For example I don't want every 5th value of i, so when I get my results 5,10,15 etc. of i j k l would not display in my answers. Thank you all for your help. In article <ef57c22.-1@webcrossing.raydaftYaTP>, Franco Alvarez <emgineering@gmail.com> wrote: >I'm failry new at matlab, and ...

Help in matlab
Hi, I am a beginner of using matlab. I am now doing a project about cancer detection in matlab. I've got 120 samples which have classified as tumor or normal cells from patients. For each sample, there is 19200 parameters which are gene information of the cell. My objective is to find the best combination of parameters to classify the sample whether it is tumor or normal. Can anyone teach me how to do this? thanks very much cyly <ynlee519@hotmail.com> wrote in message <430718512.15579.1277718307555.JavaMail.root@gallium.mathforum.org>... > Hi, I am a beginner of usin...

Help for Matlab!
Hi everyone! I hope you can help me with my probem. I want to reduce noise from a speech signal using an FIR Wiener filter. I am using the Wiener- Hopf equation to get the filter parameter, and in this equation, we need to compute autocorrelations ( or power spectrum). But first, I want to window my speech signal using 256 size windows. I only succeded in writing a matlab program that windows my signal in consecutive windows( a 256 window after a 256 window..etc). My problem is that I want the windows to partially overlap: That the second 256n window starts after 100 samples from the beginning...

Matlab help !
Can any1 help me in matlab for image processing using blind deconvolution algorithm ...

help matlab
Devo svolgere con matlab problema di equazioni differenziali con eulero trapezi usando function e non la funzione ODE!... Grazie anticipatamente L.M. i need to resolve with matlab differenzial equation with eulero (trapezi) method without use ode functin. i need just use simple matlab function. t u L.M. "lorenzomaccioni@hotmail.com" wrote: > > Devo svolgere con matlab problema di equazioni differenziali con > eulero trapezi usando function e non la funzione ODE!... > Grazie anticipatamente > L.M. > > i need to resolve with matlab differen...

Matlabe Help
I am trying to minimize the following loglikehood function and would like your help in matlab code thanks alot r(t) = a0 + a1 * r(t-1) + e(t), e(t) = (0, SD(e)) and n=N(0,1) e(t) = h(t)^0.5 * n(t) or n(t)= e(t)*h(t)^-0.5 h(t) = b0 + b1 * e(t-1)^2 + b2 * h(t-1)^2 s(t) = c0 + c1 * n(t-1)^3 + c2 * s(t-1) k(t) = d0 + d1 * n(t-1)^4 + d2 * k(t-1) LLF = -0.5 * log(h(t)) - 0.5*n(t)^2 + log(psi(n)^2) - ln(gamma(t)) psi = 1 + s(t)/6 * (n(t)^3 - 3*n(t)) + (k(t) - 3)/24 * (n(t)^4 - 6* n(t)^2 +3) gamma(t) = 1 + s(t)/6 + (k(t) - 3)^2 / 24 what am doing wrong? file 1. sk.m function [parameters, l...

Help in Matlab...
Hello Dear; I’m a programmer and developer of applications in Matlab with more than 5 years of experience and I offer services. Development projects and work on issues of science and engineering (mechanical, hydraulic systems, civil, industrial, environmental, chemical, electrical, electronics, physics, mathematics, statistics, ...) using programming software and graphical display as Matlab. I teach courses, seminars, trainings, workshops and virtual programming and graphic display of Matlab, the courses are mainly oriented to signal and image processing, numerical methods, statistics and general data processing, development and automation of processes and systems through implementing various types of models. Besides providing advice and assistance in developing all kinds of university-level technical and scientific environmental and physical sciences, mathematics and statistics related to Matlab (implementation of functions, encryption code, programs, script, GUI, executables ...). The service costs are negotiable, 50% at the beginning and 50% to finish the job. To contract and information: E-mail: correa.dave30@gmail.com Blog: http://fismatlab.blogspot.com/ Dave <correa.dave30@gmail.com> wrote in message <1706039780.58322.1280963817173.JavaMail.root@gallium.mathforum.org>... > Hello Dear; > > I&#8217;m a programmer and developer of applications in Matlab with more than 5 years of experience and I offer services. > > Development p...

Calling Compiled MATLAB from MATLAB
Hello everyone, How do I compile MATLAB M files into a form that can then be called from within the MATLAB interpreter? A roundabout way would seem to be to (somehow) Turn the matlab code into C (I think mcc does this), then compile this C code into MEX format and call that. Is there an easier way? I have the MATLAB compiler, so I just need to know what to do with it to get this to work. Thanks in advance, Sina sina.tootoonian wrote: > > > Hello everyone, > > How do I compile MATLAB M files into a form that can then be called > from within the MATLAB interpreter? A roundabout way would seem to > be > to (somehow) Turn the matlab code into C (I think mcc does this), > then > compile this C code into MEX format and call that. Is there an > easier > way? I have the MATLAB compiler, so I just need to know what to do > with > it to get this to work. > > Thanks in advance, > > Sina > > HELP MEX > HELP MEX Or, better, HELP MCC help pcode pseudo code is useful to obfuscate code - protect algorithms, yet still run at full speed and be distributable to other matlab users sina.tootoonian wrote: > > > Hello everyone, > > How do I compile MATLAB M files into a form that can then be called > from within the MATLAB interpreter? A roundabout way would seem to > be > to (somehow) Turn the matlab code into C (I think mcc does this), > then > compile this C code into MEX format and call that...

Dear All, I am completely new to MATLAB and the digital world. Could anyone help me with my following query. What are the methods used to implement a filter from an fdatool design in MATLAB
Dear All, I am completely new to MATLAB and the digital world. Could anyone help me with my following query. What are the methods used to implement a filter from an fdatool design in MATLAB On Wed, 13 Feb 2013 21:39:17 +0000, savina wrote: > Dear All, I am completely new to MATLAB and the digital world. Could > anyone help me with my following query. What are the methods used to > implement a filter from an fdatool design in MATLAB (...as well as 'completely new' to writing a compact subject line.) Depends on what you mean by "methods". If you launch the FDA tool, a 'la: > fdatool ....and look near the bottom of the GUI, you should see quite a lot of information on available filter synthesis options. ...

matlab
any one please can help me sploving the following question 1________Given the causal system y[n] =3D 0.9y[n-1] + x[n], find H(z) and sketch the pole-zero plot. See the functions roots.m and zplane.m. 2___ Plot the magnitude and phase of H(z) on the frequency axis from 0 =E2=80=93 =EF=81=B0, or normalized frequency 0 - 1, and label the graphs. S= ee the freqz.m command. hani_al_yazidi@hotmail.com wrote: > any one please can help me sploving the following question > > > 1________Given the causal system > > y[n] = 0.9y[n-1] + x[n], > > find H(z) and sketch the pole-zero plot. See the functions roots.m and > zplane.m. > > 2___ Plot the magnitude and phase of H(z) on the frequency axis from 0 > – , or normalized frequency 0 - 1, and label the graphs. See the > freqz.m command. It looks like homework. State what you tried and what error messages you got and someone who knows Matlab will probably help. We try not to do homework here, but (like gods) we do help those who help themselves. Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯&...

Matlab
Hello , I'm a new user in Matlab. So should you advise matlab notes , documentation please ? <azatserver113@gmail.com> wrote in message news:7b4e75e2-40c6-43bb-acc6-2553123ba649@googlegroups.com... > Hello , > > I'm a new user in Matlab. So should you advise matlab notes , > documentation please ? > At the MATLAB command prompt type this command: doc This will show the documentation included as part of your installation. I recommend reading through the Getting Started section of the MATLAB documentation first, and executing the examp...

can Matlab support 32-bit matlab version scripts on 64-bit version of matlab
can Matlab support 32-bit matlab version scripts on 64-bit version of matlab in windows platform?.if yes , how is the performance better ? please do answer this question thank u "Gangamma " <gangvsng@gmail.com> wrote in message <hvmr7h$sn0$1@fred.mathworks.com>... > can Matlab support 32-bit matlab version scripts on 64-bit version of matlab in windows platform?.if yes , how is the performance better ? Yes, Matlab program written in 32/64 bits are compatible accepted MEX file. 64-bit version can handle larger array but does *NOT* improve speed. Br...

Urgent help needed in matlab!! please help!!!
i have a matrix which has 16 columns and varying number of rows.. i need to replace the elements ranging from -50 to 50 with zero and the rest should remain the same.. i have coded for a single element search and replace which is [m, n] = size(Data); for i = 1: m if ( Data(i) == 1 ) Data(i) = 0; end end but how do i do it for a range of -50 to 50 instead of a 1 in my code.. could anyone help me out.. also please suggest me if there are any matlab function which does this with out the use of the for loops... Regards, Mohan i have a matrix which has 16 columns and...

Web resources about - Matlab LVQ Help - comp.soft-sys.matlab

Undocumented Matlab
Undocumented, unsupported and hidden features in Matlab

Matlab Upazila - Wikipedia, the free encyclopedia
... 23°21′N 90°42.5′E  /  23.35°N 90.7083°E  / 23.35; 90.7083 Coordinates : 23°21′N 90°42.5′E  /  23.35°N 90.7083°E  / 23.35; 90.7083 Matlab ...

Open Directory - Science: Math: Software: MATLAB
The MathWorks - Developers of MATLAB, Simulink and Stateflow. Offers products for data analysis, visualization, application development, simulation, ...

MATLAB Mobile on the App Store
Read reviews, compare customer ratings, see screenshots, and learn more about MATLAB Mobile. Download MATLAB Mobile and enjoy it on your iPhone, ...

The Smart Emergency Response System Using MATLAB and Simulink - YouTube
In the aftermath of an earthquake, every second counts for disaster victims. Imagine a system that can locate people quickly and help first responders ...

The A-Z of Programming Languages: MATLAB
What prompted the development of MATLAB?

Matlab: Invalid .NET Framework. Either the specified framework was not found or is not currently supported ...
If you are trying to compile Matlab into .Net framework, you might encounter this message, even though you have the .Net framework installed ...

Matlab edges closer to electronic design automation world
Matlab, one of the world's most widely used products for algorithmic development, has long occupied its own niche far removed from EDA and hardware ...

Can I/How do I use DMTCP with a parallel pool of MATLAB workers? · Issue #223 · dmtcp/dmtcp · GitHub
dmtcp - DMTCP: Distributed MultiThreaded CheckPointing matlab -nodisplay -r "parpool('local',4); parfor i=1:10, pause(1); fprintf('%i ',i); end; ...

Using MATLAB and SDR to Reverse Engineer 433MHz Messages
... so he focused on studying the methodology and the tools available for analyzing 433MHz RF transmission. He describes the process of using MATLAB ...

Resources last updated: 3/20/2016 5:22:25 AM