Can someone please explain why the following symmetric function cannot pass a certain limit of negative values?
D = 0.1; l = 4;
c = #(x,v) (v/D).*exp(-v*x/D)./(1-exp(-v*l/D));
v_vec = -25:0.01:25;
figure(2)
hold on
plot(v_vec,c(l,v_vec),'b')
plot(v_vec,c(0,v_vec),'r')
Notice at the figure where the blue line chops, this is where I get inf/nan values.
It seems that Matlab is trying to compute a result that is too large, outputs +inf, and then operates on that, which yields +/- inf and NaNs.
For instance, at v=-25, part of the function computes exp(-(-25)*4/0.1), which is exp(1000), and that outputs +inf. (larger than the largest representable double precision float).
You can potentially solve that problem by rewriting your function to avoid operating of such very large (or very small) numbers, say by reorganising the fraction containing exp() functions.
I did encounter the same hurdle using exp() with arguments triggering overflow. Sometimes it is difficult to trace back numeric imprecision or convergence errors. In principle the function definition using exp() only create intermediate issues as your purpose as a transition function. The intention I guess was to provide a continuous function.
My solution to this problem is to divide the argument into regions and provide in each region an approximation function. In your case zero for negative x and proportional to x for positive x. In between you can use the orginal function. Care should be taken to match the approximation at the borders of the regions and the number of continuous differentiations which is important for convergence in loops.
Related
I have recently been trying to calculate the double integral of the functionfun = #(v,x)(10^4)*0.648*(1+v*0.001).*( exp(-2.83./( 10^-8+(sqrt(1+2*v*0.001)).*(x.^2)) ) -1).*(exp(-(v.^2)*0.33)), in the range (-1000,1000) for v and (0,a) for x, where a is either a very large number or infinity. What I have found is that while in the case a = inf the value seems to be decently accurate (it reduces to a single integral which is less cumbersome to evaluate numerically), but if I integrate from 0 to a 10^9 and from 10^9 to infinity the integrals don't sum up to the correct value, with the latter one being zero. What I am really interested in is in the integral from 0 to 10^9, but these results make me wonder if I can trust it at all.
In what I have done, I also had to use a large prefactor (10^200) in front of the function to "compensate" for the small numbers; otherwise the results were all nonsense. I have tried to use vpa, but with no success. What am I doing wrong?
Rob
Looks like your problem has to do with the different methods Matlab uses for different cases and the big numbers you are handling.
We can see your function with ezsurf just to have an idea on how does it behave.
So hint 1 is that the value is going to be a negative value, lets integrate over small limits to see an approximation on how much will it be.
integral2(fun,-100,100,0,100)
%ans =
% -5.9050e+04
And assuming that the function tends to zero, we know the final value should be on the neighborhood.
Now hint 2:
integral2(fun,-1000,1000,0,100)
%ans =
% -2.5613e-29
This doesn't make much sense, by increasing the range of the limits the integral basically became zero. After checking the documentation of integral2
'Method' — Integration method
'auto' (default) | 'tiled' | 'iterated'
Integration method, specified as the comma-separated pair consisting of 'Method' and one of the methods described below.
Integration Method Description
'auto' For most cases, integral2 uses the 'tiled' method. It uses the 'iterated' method when any of the integration limits are infinite. This is the default method.
'tiled' integral2 transforms the region of integration to a rectangular shape and subdivides it into smaller rectangular regions as needed. The integration limits must be finite.
'iterated' integral2 calls integral to perform an iterated integral. The outer integral is evaluated over xmin ≤ x ≤ xmax. The inner integral is evaluated over ymin(x) ≤ y ≤ ymax(x). The integration limits can be infinite.
Ok, so if we don't define a method it will use "tiled" if the limits are finite, and "interpolated" if they are infinte.
Could it be that if the range is too big, the tiles created by the "tiled" method are too big to accurately calculate the integral? If that is the case then "iterated" should not have that problem, let's check
integral2(fun,-1000,1000,0,100,'Method','iterated')
%ans =
% -5.9050e+04
Interesting, looks like we are into something. Let's try the original problem
integral2(fun,-1000,1000,0,inf)
%ans =
% -5.9616e+04
integral2(fun,-1000,1000,0,10^9,'Method','tiled')
%ans =
% -2.1502e-33
integral2(fun,-1000,1000,0,10^9,'Method','iterated')
%ans =
% -5.9616e+04
integral2(fun,-1000,1000,10^9,inf)
%ans =
% 0
That looks better. So it looks like the 'tiled' method is the problem with your function because its characteristics and the size of the range of the limits. So as long as you use 'iterated' you should be ok.
I have an integration function which does not have indefinite integral expression.
Specifically, the function is f(y)=h(y)+integral(#(x) exp(-x-1/x),0,y) where h(y) is a simple function.
Matlab numerically computes f(y) well, but I want to compute the following function.
g(w)=w*integral(1-f(y).^(1/w),0,inf) where w is a real number in [0,1].
The problem for computing g(w) is handling f(y).^(1/w) numerically.
How can I calculate g(w) with MATLAB? Is it impossible?
Expressions containing e^(-1/x) are generally difficult to compute near x = 0. Actually, I am surprised that Matlab computes f(y) well in the first place. I'd suggest trying to compute g(w)=w*integral(1-f(y).^(1/w),epsilon,inf) for epsilon greater than zero, then gradually decreasing epsilon toward 0 to check if you can get numerical convergence at all. Convergence is certainly not guaranteed!
You can calculate g(w) using the functions you have, but you need to add the (ArrayValued,true) name-value pair.
The option allows you to specify a vector-valued w and allows the nested integral call to receive a vector of y values, which is how integral naturally works.
f = #(y) h(y)+integral(#(x) exp(-x-1/x),0,y,'ArrayValued',true);
g = #(w) w .* integral(1-f(y).^(1./w),0,Inf,'ArrayValued',true);
At least, that works on my R2014b installation.
Note: While h(y) may be simple, if it's integral over the positive real line does not converge, g(w) will more than likely not converge (I don't think I need to qualify that, but I'll hedge my bets).
Has anyone tried plotting a sine function for large values in MATLAB?
For e.g.:
x = 0:1000:100000;
plot(x,sin(2*pi*x))
I was just wondering why the amplitude is changing for this periodic function? As per what I expect, for any value of x, the function has a period of 2*pi. Why is it not?
Does anyone know? Is there a way to get it right? Also, is this a bug and is it already known?
That's actually not the amplitude changing. That is due to the numerical imprecisions of floating point arithmetic. Bear in mind that you are specifying an integer sequence from 0 to 100000 in steps of 1000. If you recall from trigonometry, sin(n*x*pi) = 0 when x and n are integers, and so theoretically you should be obtaining an output of all zeroes. In your case, n = 2, and x is a number from 0 to 100000 that is a multiple of 1000.
However, this is what I get when I use the above code in your post:
Take a look at the scale of that graph. It's 10^{-11}. Do you know how small that is? As further evidence, here's what the max and min values are of that sequence:
>> min(sin(2*pi*x))
ans =
-7.8397e-11
>> max(sin(2*pi*x))
ans =
2.9190e-11
The values are so small that they might as well be zero. What you are visualizing in the graph is due to numerical imprecision. As I mentioned before, sin(n*x*pi) = 0 when n and x is are integers, under the assumption that we have all of the decimal places of pi available. However, because we only have 64-bits total to represent pi numerically, you will certainly not get the result to be exactly zero. In addition, be advised that the sin function is very likely to be using numerical approximation algorithms (Taylor / MacLaurin series for example), and so that could also contribute to the fact that the result may not be exactly 0.
There are, of course, workarounds, such as using the symbolic mathematics toolbox (see #yoh.lej's answer). In this case, you will get zero, but I won't focus on that here. Your post is questioning the accuracy of the sin function in MATLAB, that works on numeric type inputs. Theoretically with your input into sin, as it is an integer sequence, every value of x should make sin(n*x*pi) = 0.
BTW, this article is good reading. This is what every programmer needs to know about floating point arithmetic and functions: http://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html. A more simple overview can be found here: http://floating-point-gui.de/
Because what is the exact value of pi?
This apparent error is due to the limit of floating point accuracy. If you really need/want to get around that you can do symbolic computation with matlab, see the difference between:
>> sin(2*pi*10)
ans =
-2.4493e-15
and
>> sin(sym(2*pi*10))
ans =
0
I want to numerically integrate the following:
where
and a, b and β are constants which for simplicity, can all be set to 1.
Neither Matlab using dblquad, nor Mathematica using NIntegrate can deal with the singularity created by the denominator. Since it's a double integral, I can't specify where the singularity is in Mathematica.
I'm sure that it is not infinite since this integral is based in perturbation theory and without the
has been found before (just not by me so I don't know how it's done).
Any ideas?
(1) It would be helpful if you provide the explicit code you use. That way others (read: me) need not code it up separately.
(2) If the integral exists, it has to be zero. This is because you negate the n(y)-n(x) factor when you swap x and y but keep the rest the same. Yet the integration range symmetry means that amounts to just renaming your variables, hence it must stay the same.
(3) Here is some code that shows it will be zero, at least if we zero out the singular part and a small band around it.
a = 1;
b = 1;
beta = 1;
eps[x_] := 2*(a-b*Cos[x])
n[x_] := 1/(1+Exp[beta*eps[x]])
delta = .001;
pw[x_,y_] := Piecewise[{{1,Abs[Abs[x]-Abs[y]]>delta}}, 0]
We add 1 to the integrand just to avoid accuracy issues with results that are near zero.
NIntegrate[1+Cos[(x+y)/2]^2*(n[x]-n[y])/(eps[x]-eps[y])^2*pw[Cos[x],Cos[y]],
{x,-Pi,Pi}, {y,-Pi,Pi}] / (4*Pi^2)
I get the result below.
NIntegrate::slwcon:
Numerical integration converging too slowly; suspect one of the following:
singularity, value of the integration is 0, highly oscillatory integrand,
or WorkingPrecision too small.
NIntegrate::eincr:
The global error of the strategy GlobalAdaptive has increased more than
2000 times. The global error is expected to decrease monotonically after a
number of integrand evaluations. Suspect one of the following: the
working precision is insufficient for the specified precision goal; the
integrand is highly oscillatory or it is not a (piecewise) smooth
function; or the true value of the integral is 0. Increasing the value of
the GlobalAdaptive option MaxErrorIncreases might lead to a convergent
numerical integration. NIntegrate obtained 39.4791 and 0.459541
for the integral and error estimates.
Out[24]= 1.00002
This is a good indication that the unadulterated result will be zero.
(4) Substituting cx for cos(x) and cy for cos(y), and removing extraneous factors for purposes of convergence assessment, gives the expression below.
((1 + E^(2*(1 - cx)))^(-1) - (1 + E^(2*(1 - cy)))^(-1))/
(2*(1 - cx) - 2*(1 - cy))^2
A series expansion in cy, centered at cx, indicates a pole of order 1. So it does appear to be a singular integral.
Daniel Lichtblau
The integral looks like a Cauchy Principal Value type integral (i.e. it has a strong singularity). That's why you can't apply standard quadrature techniques.
Have you tried PrincipalValue->True in Mathematica's Integrate?
In addition to Daniel's observation about integrating an odd integrand over a symmetric range (so that symmetry indicates the result should be zero), you can also do this to understand its convergence better (I'll use latex, writing this out with pen and paper should make it easier to read; it took a lot longer to write than to do, it's not that complicated):
First, epsilon(x)-\epsilon(y)\propto\cos(y)-\cos(x)=2\sin(\xi_+)\sin(\xi_-) where I have defined \xi_\pm=(x\pm y)/2 (so I've rotated the axes by pi/4). The region of integration then is \xi_+ between \pi/\sqrt{2} and -\pi/\sqrt{2} and \xi_- between \pm(\pi/\sqrt{2}-\xi_-). Then the integrand takes the form \frac{1}{\sin^2(\xi_-)\sin^2(\xi_+)} times terms with no divergences. So, evidently, there are second-order poles, and this isn't convergent as presented.
Perhaps you could email the persons who obtained an answer with the cos term and ask what precisely it is they did. Perhaps there's a physical regularisation procedure being employed. Or you could have given more information on the physical origin of this (some sort of second order perturbation theory for some sort of bosonic system?), had that not been off-topic here...
May be I am missing something here, but the integrand
f[x,y]=Cos^2[(x+y)/2]*(n[x]-n[y])/(eps[x]-eps[y]) with n[x]=1/(1+Exp[Beta*eps[x]]) and eps[x]=2(a-b*Cos[x]) is indeed a symmetric function in x and y: f[x,-y]= f[-x,y]=f[x,y].
Therefore its integral over any domain [-u,u]x[-v,v] is zero. No numerical integration seems to be needed here. The result is just zero.
I'm trying to do some parameter estimation and want to choose parameter estimates that minimize the square error in a predicted equation over about 30 variables. If the equation were linear, I would just compute the 30 partial derivatives, set them all to zero, and use a linear-equation solver. But unfortunately the equation is nonlinear and so are its derivatives.
If the equation were over a single variable, I would just use Newton's method (also known as Newton-Raphson). The Web is rich in examples and code to implement Newton's method for functions of a single variable.
Given that I have about 30 variables, how can I program a numeric solution to this problem using Newton's method? I have the equation in closed form and can compute the first and second derivatives, but I don't know quite how to proceed from there. I have found a large number of treatments on the web, but they quickly get into heavy matrix notation. I've found something moderately helpful on Wikipedia, but I'm having trouble translating it into code.
Where I'm worried about breaking down is in the matrix algebra and matrix inversions. I can invert a matrix with a linear-equation solver but I'm worried about getting the right rows and columns, avoiding transposition errors, and so on.
To be quite concrete:
I want to work with tables mapping variables to their values. I can write a function of such a table that returns the square error given such a table as argument. I can also create functions that return a partial derivative with respect to any given variable.
I have a reasonable starting estimate for the values in the table, so I'm not worried about convergence.
I'm not sure how to write the loop that uses an estimate (table of value for each variable), the function, and a table of partial-derivative functions to produce a new estimate.
That last is what I'd like help with. Any direct help or pointers to good sources will be warmly appreciated.
Edit: Since I have the first and second derivatives in closed form, I would like to take advantage of them and avoid more slowly converging methods like simplex searches.
The Numerical Recipes link was most helpful. I wound up symbolically differentiating my error estimate to produce 30 partial derivatives, then used Newton's method to set them all to zero. Here are the highlights of the code:
__doc.findzero = [[function(functions, partials, point, [epsilon, steps]) returns table, boolean
Where
point is a table mapping variable names to real numbers
(a point in N-dimensional space)
functions is a list of functions, each of which takes a table like
point as an argument
partials is a list of tables; partials[i].x is the partial derivative
of functions[i] with respect to 'x'
epilson is a number that says how close to zero we're trying to get
steps is max number of steps to take (defaults to infinity)
result is a table like 'point', boolean that says 'converged'
]]
-- See Numerical Recipes in C, Section 9.6 [http://www.nrbook.com/a/bookcpdf.php]
function findzero(functions, partials, point, epsilon, steps)
epsilon = epsilon or 1.0e-6
steps = steps or 1/0
assert(#functions > 0)
assert(table.numpairs(partials[1]) == #functions,
'number of functions not equal to number of variables')
local equations = { }
repeat
if Linf(functions, point) <= epsilon then
return point, true
end
for i = 1, #functions do
local F = functions[i](point)
local zero = F
for x, partial in pairs(partials[i]) do
zero = zero + lineq.var(x) * partial(point)
end
equations[i] = lineq.eqn(zero, 0)
end
local delta = table.map(lineq.tonumber, lineq.solve(equations, {}).answers)
point = table.map(function(v, x) return v + delta[x] end, point)
steps = steps - 1
until steps <= 0
return point, false
end
function Linf(functions, point)
-- distance using L-infinity norm
assert(#functions > 0)
local max = 0
for i = 1, #functions do
local z = functions[i](point)
max = math.max(max, math.abs(z))
end
return max
end
You might be able to find what you need at the Numerical Recipes in C web page. There is a free version available online. Here (PDF) is the chapter containing the Newton-Raphson method implemented in C. You may also want to look at what is available at Netlib (LINPack, et. al.).
As an alternative to using Newton's method the Simplex Method of Nelder-Mead is ideally suited to this problem and referenced in Numerical Recpies in C.
Rob
You are asking for a function minimization algorithm. There are two main classes: local and global. Your problem is least squares so both local and global minimization algorithms should converge to the same unique solution. Local minimization is far more efficient than global so select that.
There are many local minimization algorithms but one particularly well suited to least squares problems is Levenberg-Marquardt. If you don't have such a solver to hand (e.g. from MINPACK) then you can probably get away with Newton's method:
x <- x - (hessian x)^-1 * grad x
where you compute the inverse matrix multiplied by a vector using a linear solver.
Since you already have the partial derivatives, how about a general gradient-descent approach?
Maybe you think you have a good-enough solution, but for me, the easiest way to think about this is to understand it in the 1-variable case first, and then extend it to the matrix case.
In the 1-variable case, if you divide the first derivative by the second derivative, you get the (negative) step size to your next trial point, e.g. -V/A.
In the N-variable case, the first derivative is a vector and the second derivative is a matrix (the Hessian). You multiply the derivative vector by the inverse of the second derivative, and the result is the negative step-vector to your next trial point, e.g. -V*(1/A)
I assume you can get the 2nd-derivative Hessian matrix. You will need a routine to invert it. There are plenty of these around in various linear algebra packages, and they are quite fast.
(For readers who are not familiar with this idea, suppose the two variables are x and y, and the surface is v(x,y). Then the first derivative is the vector:
V = [ dv/dx, dv/dy ]
and the second derivative is the matrix:
A = [dV/dx]
[dV/dy]
or:
A = [ d(dv/dx)/dx, d(dv/dy)/dx]
[ d(dv/dx)/dy, d(dv/dy)/dy]
or:
A = [d^2v/dx^2, d^2v/dydx]
[d^2v/dxdy, d^2v/dy^2]
which is symmetric.)
If the surface is parabolic (constant 2nd derivative) it will get to the answer in 1 step. On the other hand, if the 2nd derivative is very not-constant, you could encounter oscillation. Cutting each step in half (or some fraction) should make it stable.
If N == 1, you'll see that it does the same thing as in the 1-variable case.
Good luck.
Added: You wanted code:
double X[N];
// Set X to initial estimate
while(!done){
double V[N]; // 1st derivative "velocity" vector
double A[N*N]; // 2nd derivative "acceleration" matrix
double A1[N*N]; // inverse of A
double S[N]; // step vector
CalculateFirstDerivative(V, X);
CalculateSecondDerivative(A, X);
// A1 = 1/A
GetMatrixInverse(A, A1);
// S = V*(1/A)
VectorTimesMatrix(V, A1, S);
// if S is small enough, stop
// X -= S
VectorMinusVector(X, S, X);
}
My opinion is to use a stochastic optimizer, e.g., a Particle Swarm method.