Dear all, If I already know a prior that my data distribution should follow the shape of Laplacian distribution... the data obtained from measurement is of course a little off(not very symmtrical), how can I make the measured data more Laplacian distribution like(make it at least a little more symmtrical)? Can anybody give me an example or detailed explanation? I am kind of afraid of statistics... :=) Thanks a lot, -Walala
(statistics)how to make date more like Laplacian distribution?
Started by ●November 19, 2003
Reply by ●November 19, 20032003-11-19
"walala" <mizhael@yahoo.com> writes:> If I already know a prior that my data distribution should follow the shape > of Laplacian distribution... the data obtained from measurement is of course > a little off(not very symmtrical), how can I make the measured data more > Laplacian distribution like(make it at least a little more symmtrical)? > > Can anybody give me an example or detailed explanation? I am kind of afraid > of statistics... :=)Why do you think your data is Laplacian? If it's not symmetric, perhaps that's telling you that you don't understand what the data really should be? How was the data generated? Ciao, Peter "Apologies for answering a question with questions!" K. -- Peter J. Kootsookos "I will ignore all ideas for new works [..], the invention of which has reached its limits and for whose improvement I see no further hope." - Julius Frontinus, c. AD 84
Reply by ●November 19, 20032003-11-19
In article <bpesrd$jkk$1@mozo.cc.purdue.edu>, "walala" <mizhael@yahoo.com> wrote:> If I already know a prior that my data distribution should follow the shape > of Laplacian distribution... the data obtained from measurement is of course > a little off(not very symmtrical), how can I make the measured data more > Laplacian distribution like(make it at least a little more symmtrical)? > > Can anybody give me an example or detailed explanation? I am kind of afraid > of statistics... :=)Lets see if I understand. Your data does not follow your prior beliefs. Therefore the data is wrong, and you want to know how to modify said data so that it does exactly what you want. This makes sense - in a very distorted way. Why did you bother with measuring that blasted data in the first place? You have already decided the result. Measurements just get in the way. Please start by reading the book "How to lie with Statistics". Then return to your data, and learn from it. Is it just random noise that has given your data this property you did not expect? Or is this an indication of a problem in your measurement? Perhaps it indicates something wrong with the theory? Perhaps another factor distorts the data? Maybe your sample is just too small! I can summarize the strong suggestions I give to my students who deal with data in three words: Plot - think - learn. Only after that do I tell them to do any actual modeling, or use their data in any way. HTH, John D'Errico -- There are no questions "?" about my real address. The best material model of a cat is another, or preferably the same, cat. A. Rosenblueth, Philosophy of Science, 1945
Reply by ●November 19, 20032003-11-19
Hello Walala, How big is your sample size? If you grab only100 samples, you wouldn't expect it to exactly fit the generating distribution. Can you grab 1,000,000 samples? If still it doesn't fit, then suspect your generating dist. is different from a Laplacian. You can look into bootstrapping methods to get an estimate of the dist. Clay "walala" <mizhael@yahoo.com> wrote in message news:bpesrd$jkk$1@mozo.cc.purdue.edu...> Dear all, > > If I already know a prior that my data distribution should follow theshape> of Laplacian distribution... the data obtained from measurement is ofcourse> a little off(not very symmtrical), how can I make the measured data more > Laplacian distribution like(make it at least a little more symmtrical)? > > Can anybody give me an example or detailed explanation? I am kind ofafraid> of statistics... :=) > > Thanks a lot, > > -Walala > >
Reply by ●November 19, 20032003-11-19
"walala" <mizhael@yahoo.com> wrote in message news:bpesrd$jkk$1@mozo.cc.purdue.edu...> If I already know a prior that my data distribution should follow theshape> of Laplacian distribution... the data obtained from measurement is ofcourse> a little off(not very symmtrical), how can I make the measured data more > Laplacian distribution like(make it at least a little more symmtrical)?Walala, others have already responded that it's unusual to modify data to fit a distribution. However, it is common to fit a distribution to data. That distribution may be a "more symmetric" model for the process that generated the data. You'd probably also want to test your assumption that this is a good fit, either informally (maybe with a graph) or formally (with a hypothesis test). Could it be that you really want information about how to fit a Laplace distribution? Here's a bit of MATLAB code that generates some normal data and fits a Laplace distribution to it, then compares a histogram of the data to the re-scaled Laplace density, which is symmetric. % Generate normal data, just for example N = 50; x = 20 + 2*randn(N,1); % Fit the Laplace parameters mu = median(x); sig = mean(abs(x-mu)); % Compare a histogram to the fitted distribution hist(x,floor(min(x)):ceil(max(x))) xlim = get(gca,'xlim'); xx = linspace(xlim(1),xlim(2),200); yy = N * exp(-abs( (xx-mu)/sig ))/(2*sig); hold on; plot(xx,yy,'r-'); hold off -- Tom
Reply by ●November 19, 20032003-11-19
walala wrote:> Dear all, > > If I already know a prior that my data distribution should follow the shape > of Laplacian distribution... the data obtained from measurement is of course > a little off(not very symmtrical), how can I make the measured data more > Laplacian distribution like(make it at least a little more symmtrical)? > > Can anybody give me an example or detailed explanation? I am kind of afraid > of statistics... :=) > > Thanks a lot, > > -WalalaThe data you have are what you measured. If they don't suit you, do as corporations and government agencies do: lie. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●November 19, 20032003-11-19
"Peter J. Kootsookos" <p.kootsookos@remove.ieee.org> wrote in message news:s68znesvrts.fsf@mango.itee.uq.edu.au...> "walala" <mizhael@yahoo.com> writes: > > > If I already know a prior that my data distribution should follow theshape> > of Laplacian distribution... the data obtained from measurement is ofcourse> > a little off(not very symmtrical), how can I make the measured data more > > Laplacian distribution like(make it at least a little more symmtrical)? > > > > Can anybody give me an example or detailed explanation? I am kind ofafraid> > of statistics... :=) > > Why do you think your data is Laplacian? If it's not symmetric, > perhaps that's telling you that you don't understand what the data > really should be? > > How was the data generated? > > Ciao, > > Peter "Apologies for answering a question with questions!" K. > > -- > Peter J. Kootsookos > > "I will ignore all ideas for new works [..], the invention of which > has reached its limits and for whose improvement I see no further > hope." > > - Julius Frontinus, c. AD 84Dear Peter, Why people think I want to lie upon seeing my question? Oh, it's my problem that I did not clearly present the background... Here is the story: in deblocking of block DCT coded JPEG images, it was known that the DCTed coefficients are Laplacian distributed... But now I am looking at low bit rate JPEG images, so there are someking of artifacts... in order to reconstruct the original images... many algorithms have been devised... one possibility is to make the image coefficients more Laplcian like... So that came my question: how to make data more Laplican like...Please give me some detailed explanation as I am not veteran in statistics... Thanks a lot, -Walala (I am just a poor student, not lieing government agency, issurance company, weapon dealer, lawyers, and politicians... so please help me!)
Reply by ●November 19, 20032003-11-19
"John D'Errico" <?derrico@flare.n?e?t> wrote in message news:?derrico-BB6853.06593219112003@news.newsguy.com...> In article <bpesrd$jkk$1@mozo.cc.purdue.edu>, > "walala" <mizhael@yahoo.com> wrote: > > > If I already know a prior that my data distribution should follow theshape> > of Laplacian distribution... the data obtained from measurement is ofcourse> > a little off(not very symmtrical), how can I make the measured data more > > Laplacian distribution like(make it at least a little more symmtrical)? > > > > Can anybody give me an example or detailed explanation? I am kind ofafraid> > of statistics... :=) > > > Lets see if I understand. Your data does not follow > your prior beliefs. Therefore the data is wrong, and > you want to know how to modify said data so that it > does exactly what you want. This makes sense - in a > very distorted way. > > Why did you bother with measuring that blasted data > in the first place? You have already decided the > result. Measurements just get in the way. > > Please start by reading the book "How to lie with > Statistics". Then return to your data, and learn from > it. Is it just random noise that has given your data > this property you did not expect? Or is this an > indication of a problem in your measurement? Perhaps > it indicates something wrong with the theory? Perhaps > another factor distorts the data? Maybe your sample > is just too small! > > I can summarize the strong suggestions I give to my > students who deal with data in three words: > > Plot - think - learn. > > Only after that do I tell them to do any actual > modeling, or use their data in any way. > > HTH, > John D'Errico > > > -- > There are no questions "?" about my real address. > > The best material model of a cat is another, or > preferably the same, cat. > A. Rosenblueth, Philosophy of Science, 1945Dear John, Why people think I want to lie upon seeing my question? Oh, it's my problem that I did not clearly present the background... Here is the story: in deblocking of block DCT coded JPEG images, it was known that the DCTed coefficients are Laplacian distributed... But now I am looking at low bit rate JPEG images, so there are someking of artifacts... in order to reconstruct the original images... many algorithms have been devised... one possibility is to make the image coefficients more Laplcian like... So that came my question: how to make data more Laplican like...Please give me some detailed explanation as I am not veteran in statistics... Thanks a lot, -Walala (I am just a poor student, not lieing government agency, issurance company, weapon dealer, lawyers, and politicians... so please help me!)
Reply by ●November 19, 20032003-11-19
"Jerry Avins" <jya@ieee.org> wrote in message news:bpg5pp$oa7$2@bob.news.rcn.net...> walala wrote: > > > Dear all, > > > > If I already know a prior that my data distribution should follow theshape> > of Laplacian distribution... the data obtained from measurement is ofcourse> > a little off(not very symmtrical), how can I make the measured data more > > Laplacian distribution like(make it at least a little more symmtrical)? > > > > Can anybody give me an example or detailed explanation? I am kind ofafraid> > of statistics... :=) > > > > Thanks a lot, > > > > -Walala > > The data you have are what you measured. If they don't suit you, do as > corporations and government agencies do: lie. > > Jerry > -- > Engineering is the art of making what you want from things you can get. > ����������������������������������������������������������������������� >Dear Jerry, Why people think I want to lie upon seeing my question? Oh, it's my problem that I did not clearly present the background... Here is the story: in deblocking of block DCT coded JPEG images, it was known that the DCTed coefficients are Laplacian distributed... But now I am looking at low bit rate JPEG images, so there are someking of artifacts... in order to reconstruct the original images... many algorithms have been devised... one possibility is to make the image coefficients more Laplcian like... So that came my question: how to make data more Laplican like...Please give me some detailed explanation as I am not veteran in statistics... Thanks a lot, -Walala (I am just a poor student, not lieing government agency, issurance company, weapon dealer, lawyers, and politicians... so please help me!)
Reply by ●November 19, 20032003-11-19
"Clay S. Turner" <physics@bellsouth.net> wrote in message news:oKKub.3456$gU2.511@bignews6.bellsouth.net...> Hello Walala, > > How big is your sample size? If you grab only100 samples, you wouldn't > expect it to exactly fit the generating distribution. Can you grab1,000,000> samples? If still it doesn't fit, then suspect your generating dist. is > different from a Laplacian. You can look into bootstrapping methods to get > an estimate of the dist. > > Clay > > > > > "walala" <mizhael@yahoo.com> wrote in message > news:bpesrd$jkk$1@mozo.cc.purdue.edu... > > Dear all, > > > > If I already know a prior that my data distribution should follow the > shape > > of Laplacian distribution... the data obtained from measurement is of > course > > a little off(not very symmtrical), how can I make the measured data more > > Laplacian distribution like(make it at least a little more symmtrical)? > > > > Can anybody give me an example or detailed explanation? I am kind of > afraid > > of statistics... :=) > > > > Thanks a lot, > > > > -Walala > > > > > >Dear Clay, Why people think I want to lie upon seeing my question? Oh, it's my problem that I did not clearly present the background... Here is the story: in deblocking of block DCT coded JPEG images, it was known that the DCTed coefficients are Laplacian distributed... But now I am looking at low bit rate JPEG images, so there are someking of artifacts... in order to reconstruct the original images... many algorithms have been devised... one possibility is to make the image coefficients more Laplcian like... So that came my question: how to make data more Laplican like...Please give me some detailed explanation as I am not veteran in statistics... Thanks a lot, -Walala (I am just a poor student, not lieing government agency, issurance company, weapon dealer, lawyers, and politicians... so please help me!)
