Inverse evaluation - Give x value and get back the y value
in [Matlab]
|
Prev: How to find all different turns in a graph from Adjacency List?
Next: Gucci sale-Gucci outlet store-Cheap Gucci Bags-Handbags,40% off,Free Shipping Tax Free
From: Alex Pereira on 15 Jul 2010 20:29 Hi All, I've created a function with cfit, with the code shown bellow. Now I need to find the x value for a given y value. I know how to evaluate y values with this command y=feval(func, x). is it possible to create inverse functions ? As much as I remember, this function has no inverse form. I can't even imagine how to do it. Does anyone knows ? Thank you, Alex a=[0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 ]; b=[1:size(a,2)]; b = b(:); a = a(:); fo=fitoptions('method','SmoothingSpline','SmoothingParam',0.010000000000000000555); ft = fittype('smoothingspline'); func = fit(b,a,ft,fo);
From: John D'Errico on 15 Jul 2010 20:51 "Alex Pereira" <alexlopespereira(a)gmail.com> wrote in message <i1o94h$51j$1(a)fred.mathworks.com>... > Hi All, > I've created a function with cfit, with the code shown bellow. Now I need to find the x value for a given y value. > > I know how to evaluate y values with this command y=feval(func, x). is it possible to create inverse functions ? As much as I remember, this function has no inverse form. > > I can't even imagine how to do it. Does anyone knows ? > > Thank you, > Alex > > a=[0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 ]; > b=[1:size(a,2)]; > b = b(:); > a = a(:); > fo=fitoptions('method','SmoothingSpline','SmoothingParam',0.010000000000000000555); > ft = fittype('smoothingspline'); > func = fit(b,a,ft,fo); hint: help fzero I know, your next question will be that fzero looks for a point where the function is ZERO. But you want to find where it equals a given value. But you CAN subtract the desired value from your function. Then fzero would indeed be looking for a root. Of course, there are multiple solutions. fzero will find only one of them. John
From: Alex Pereira on 15 Jul 2010 21:30 "John D'Errico" <woodchips(a)rochester.rr.com> wrote in message <i1oado$pv2$1(a)fred.mathworks.com>... > "Alex Pereira" <alexlopespereira(a)gmail.com> wrote in message <i1o94h$51j$1(a)fred.mathworks.com>... > > Hi All, > > I've created a function with cfit, with the code shown bellow. Now I need to find the x value for a given y value. > > > > I know how to evaluate y values with this command y=feval(func, x). is it possible to create inverse functions ? As much as I remember, this function has no inverse form. > > > > I can't even imagine how to do it. Does anyone knows ? > > > > Thank you, > > Alex > > > > a=[0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 ]; > > b=[1:size(a,2)]; > > b = b(:); > > a = a(:); > > fo=fitoptions('method','SmoothingSpline','SmoothingParam',0.010000000000000000555); > > ft = fittype('smoothingspline'); > > func = fit(b,a,ft,fo); > > hint: > > help fzero > > I know, your next question will be that fzero looks for > a point where the function is ZERO. But you want to > find where it equals a given value. But you CAN > subtract the desired value from your function. Then > fzero would indeed be looking for a root. > > Of course, there are multiple solutions. fzero will find > only one of them. > > John Dear John, It will work. Thank you very much. Alex
From: Alex Pereira on 16 Jul 2010 11:22 "Alex Pereira" <alexlopespereira(a)gmail.com> wrote in message <i1ocne$lhb$1(a)fred.mathworks.com>... > "John D'Errico" <woodchips(a)rochester.rr.com> wrote in message <i1oado$pv2$1(a)fred.mathworks.com>... > > "Alex Pereira" <alexlopespereira(a)gmail.com> wrote in message <i1o94h$51j$1(a)fred.mathworks.com>... > > > Hi All, > > > I've created a function with cfit, with the code shown bellow. Now I need to find the x value for a given y value. > > > > > > I know how to evaluate y values with this command y=feval(func, x). is it possible to create inverse functions ? As much as I remember, this function has no inverse form. > > > > > > I can't even imagine how to do it. Does anyone knows ? > > > > > > Thank you, > > > Alex > > > > > > a=[0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 ]; > > > b=[1:size(a,2)]; > > > b = b(:); > > > a = a(:); > > > fo=fitoptions('method','SmoothingSpline','SmoothingParam',0.010000000000000000555); > > > ft = fittype('smoothingspline'); > > > func = fit(b,a,ft,fo); > > > > hint: > > > > help fzero > > > > I know, your next question will be that fzero looks for > > a point where the function is ZERO. But you want to > > find where it equals a given value. But you CAN > > subtract the desired value from your function. Then > > fzero would indeed be looking for a root. > > > > Of course, there are multiple solutions. fzero will find > > only one of them. > > > > John > Dear John, > It will work. > > Thank you very much. > Alex Dear John, Do you know how can I subtract a scalar from the function returned by the fit function ? Thank you, Alex
From: Steven Lord on 16 Jul 2010 13:39
"Alex Pereira" <alexlopespereira(a)gmail.com> wrote in message news:i1ptet$6gk$1(a)fred.mathworks.com... > "Alex Pereira" <alexlopespereira(a)gmail.com> wrote in message > <i1ocne$lhb$1(a)fred.mathworks.com>... >> "John D'Errico" <woodchips(a)rochester.rr.com> wrote in message >> <i1oado$pv2$1(a)fred.mathworks.com>... *snip* > Dear John, > > Do you know how can I subtract a scalar from the function returned by the > fit function ? You can't subtract a scalar from a fit object -- subtraction isn't defined for fit objects. What you can do is _evaluate_ the fit object and subtract a scalar from the result. Using the example from the reference page for FIT to try to find when the US population was 100 million: http://www.mathworks.com/access/helpdesk/help/toolbox/curvefit/fit.html % Load data load census % Perform the fit s = fitoptions('Method','NonlinearLeastSquares',... 'Lower',[0,0],... 'Upper',[Inf,max(cdate)],... 'Startpoint',[1 1]); f = fittype('a*(x-b)^n','problem','n','options',s); [c2,gof2] = fit(cdate,pop,f,'problem',2); % Determine when the population was 100 desiredPopulation = 100; objective = @(date) c2(date) - desiredPopulation; desiredDate = fzero(objective, 1900) % Evaluate the fit at the calculated time populationOnThatDate = c2(desiredDate) % Plot it plot(cdate, pop, 'k-', desiredDate, c2(desiredDate), 'r+') -- 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 |