HELP IF ME YOU CAN!!! Data reduction.
in [Matlab]
|
Prev: www.voguesneakers.com Cheap Jordans,Cheap Nikes,Cheap Air Max 90
Next: IDEA search - fast read and write on workspace at RUNTIME (Simulink)
From: Steven Lord on 27 May 2010 08:56 "Charles " <nkwosman(a)yahoo.com> wrote in message news:htlnfe$qkv$1(a)fred.mathworks.com... > Hi, that is not what I mean. What I need is a kind of transformation that > reduces the number of dimension from of the data from 2 columns to 1 > column. Certainly. There are many such transformations. * Take just the first column and ignore the second * Ignore the first column and take just the second * Take a linear combination of the two columns * Use HYPOT on the two columns * etc You need to be clearer about what your goal is in using this transformation -- does it need to include "information" from both columns, does it need to be linear, does it need to pass through a certain point, does it need to be monotonic, etc. As stated right now, there is insufficient information to provide you with anything more than very broad suggestions. -- Steve Lord slord(a)mathworks.com comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ To contact Technical Support use the Contact Us link on http://www.mathworks.com
From: Charles on 27 May 2010 09:16 Hi, thanks for your information. Unfortunately, I cannot take one column and ignore the other. Here is a sample of my data 0.066842 0.17721 0.57632 0.30842 1.0858 0.44874 1.5953 0.58121 2.1047 0.69558 2.6142 0.78242 3.1237 0.84037 3.6332 0.88074 4.1426 0.91695 4.6521 0.95016 5.1616 0.97679 5.6711 0.991 6.1805 0.997 6.69 1 Any kind of transformation that will reduce the data to one column will be fine. Thanks. Charles "Steven Lord" <slord(a)mathworks.com> wrote in message <htlq52$pps$1(a)fred.mathworks.com>... > > "Charles " <nkwosman(a)yahoo.com> wrote in message > news:htlnfe$qkv$1(a)fred.mathworks.com... > > Hi, that is not what I mean. What I need is a kind of transformation that > > reduces the number of dimension from of the data from 2 columns to 1 > > column. > > Certainly. There are many such transformations. > > * Take just the first column and ignore the second > * Ignore the first column and take just the second > * Take a linear combination of the two columns > * Use HYPOT on the two columns > * etc > > You need to be clearer about what your goal is in using this > transformation -- does it need to include "information" from both columns, > does it need to be linear, does it need to pass through a certain point, > does it need to be monotonic, etc. As stated right now, there is > insufficient information to provide you with anything more than very broad > suggestions. > > -- > Steve Lord > slord(a)mathworks.com > comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ > To contact Technical Support use the Contact Us link on > http://www.mathworks.com >
From: us on 27 May 2010 09:17 "Charles " > I have data like this: > > A=[1 2 3 4 5 6 7 8;1 2 3 4 5 6 7 8]; > what I want to achieve is a kind of mapping that will achieve > A=[ 2 4 3 5 3 3 5 2]. > > So that I don't have to use the command > plot(A(:,1),A(:,2)) > but instead > plot(A) > and I see the same data. > Thanks for your time. > Charles ??? can you explain how you get from A1 -> A2... us
From: Charles on 27 May 2010 09:26 That is exactly what I want to know. Please look at the previous example of my data that I provided Charles. "us " <us(a)neurol.unizh.ch> wrote in message <htlrcg$j6d$1(a)fred.mathworks.com>... > "Charles " > > I have data like this: > > > > A=[1 2 3 4 5 6 7 8;1 2 3 4 5 6 7 8]; > > what I want to achieve is a kind of mapping that will achieve > > A=[ 2 4 3 5 3 3 5 2]. > > > > So that I don't have to use the command > > plot(A(:,1),A(:,2)) > > but instead > > plot(A) > > and I see the same data. > > Thanks for your time. > > Charles > > ??? can you explain how you get from A1 -> A2... > > us
From: Charles on 27 May 2010 09:28
I noticed the data in the previous post doesn't quite look right -2.99 0 -2.4805 0.002 -1.9711 0.009 -1.4616 0.020211 -0.95211 0.037895 -0.44263 0.082421 0.066842 0.17721 0.57632 0.30842 1.0858 0.44874 1.5953 0.58121 2.1047 0.69558 2.6142 0.78242 3.1237 0.84037 3.6332 0.88074 4.1426 0.91695 4.6521 0.95016 5.1616 0.97679 5.6711 0.991 6.1805 0.997 6.69 1 "Charles " <nkwosman(a)yahoo.com> wrote in message <htlram$fb8$1(a)fred.mathworks.com>... > Hi, > > thanks for your information. Unfortunately, I cannot take one column and ignore the other. > Here is a sample of my data > > 0.066842 0.17721 > 0.57632 0.30842 > 1.0858 0.44874 > 1.5953 0.58121 > 2.1047 0.69558 > 2.6142 0.78242 > 3.1237 0.84037 > 3.6332 0.88074 > 4.1426 0.91695 > 4.6521 0.95016 > 5.1616 0.97679 > 5.6711 0.991 > 6.1805 0.997 > 6.69 1 > > Any kind of transformation that will reduce the data to one column will be fine. Thanks. > > Charles > > "Steven Lord" <slord(a)mathworks.com> wrote in message <htlq52$pps$1(a)fred.mathworks.com>... > > > > "Charles " <nkwosman(a)yahoo.com> wrote in message > > news:htlnfe$qkv$1(a)fred.mathworks.com... > > > Hi, that is not what I mean. What I need is a kind of transformation that > > > reduces the number of dimension from of the data from 2 columns to 1 > > > column. > > > > Certainly. There are many such transformations. > > > > * Take just the first column and ignore the second > > * Ignore the first column and take just the second > > * Take a linear combination of the two columns > > * Use HYPOT on the two columns > > * etc > > > > You need to be clearer about what your goal is in using this > > transformation -- does it need to include "information" from both columns, > > does it need to be linear, does it need to pass through a certain point, > > does it need to be monotonic, etc. As stated right now, there is > > insufficient information to provide you with anything more than very broad > > suggestions. > > > > -- > > Steve Lord > > slord(a)mathworks.com > > comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ > > To contact Technical Support use the Contact Us link on > > http://www.mathworks.com > > |