I'm trying to program the z-transform in wxMaxima which doesn't have it programmed but not by definition but by using the Scilab approach. Scilab to calculate the z-transform first converts the transfer function to the state space, after that the system must be discretized and after that converted to z transfer function, I need this because of some algebraic calculations that I need to do to analyze stability of a system in function of the sample period.
Right now I'm stranded with the function balanc() which finds a similarity transform such that
Ab = X^(-1) . A . X
as approximately equal row and column norms.
Most of my code in wxMaxima to reach in the near future has been done by translating the Scilab code into wxMaxima, currently I'm writing the tf2ss() function an inside that function the balanc() function is called, the problem is that I couldn't find the code for that function in Scilab installation directory, I've searched info in books and papers but every example starts with the Ab matrix given as an input to the problem, Scilab instead has the option to have as an input only the A matrix and it calculates the Ab and X matrices, so, I need help to make this function exactly as Scilab has it programmed to been able to compare all the steps that I'm doing.
Finally, wxMaxima has a function to calculate similarity transforms but it don't have the same output as Scilab what it means to me that they uses different criteria to calculate the similarity transform.
Note: I've tried to make the calculations in wxMaxima to have Ab and X matrices as elements with variables but the system of equations remains with too many variables and couldn't be solved.
Thanks in advance for the help in doing this.
In Scilab balanc() is hard-coded and based on LAPACK's dgebal (see the Fortran source at Netlib). In the algorithm the operations are quite simple (computing inf and 2-norms, swaping columns or rows of a matrix), maybe this could easily translated ?
A more readable version of the algorithm can be found on page 3 (Algorithm 2) of the following document: https://arxiv.org/abs/1401.5766.
Here is a Scilab implementation of Algorithm 3:
function [A,X]=bal(Ain)
A = Ain;
n = size(A,1);
X = ones(n,1);
β = 2; // multiply or divide by radix preserves precision
p = 2; // eventually change to 1-norm
converged = 0;
while converged == 0
converged = 1;
for i=1:n
c = norm(A(:,i),p);
r = norm(A(i,:),p);
s = c^p+r^p;
f = 1;
while c < r/β
c = c*β;
r = r/β;
f = f*β;
end
while c >= r*β
c = c/β;
r = r*β;
f = f/β;
end
if (c^p+r^p) < 0.95*s
converged = 0;
X(i) = f*X(i);
A(:,i) = f*A(:,i);
A(i,:) = A(i,:)/f;
end
end
end
X = diag(X);
endfunction
On this example the above implementation gives the same balanced matrix:
--> A=rand(5,5,"normal"); A(:,1)=A(:,1)*1024; A(2,:)=A(2,:)/1024
A =
897.30729 -1.6907865 -1.0217046 -0.9181476 -0.1464695
-0.5430253 -0.0011318 -0.0000356 -0.001277 -0.00038
-774.96457 3.1685332 0.1467254 -0.410953 -0.6165827
155.22118 0.1680727 -0.2262445 -0.3402948 1.6098294
1423.0797 -0.3302511 0.5909125 -1.2169245 -0.7546739
--> [Ab,X]=balanc(A)
Ab =
897.30729 -0.8453932 -32.694547 -14.690362 -9.3740507
-1.0860507 -0.0011318 -0.0022789 -0.0408643 -0.0486351
-24.217643 0.0495083 0.1467254 -0.2054765 -1.2331655
9.7013239 0.0052523 -0.452489 -0.3402948 6.4393174
22.23562 -0.0025801 0.2954562 -0.3042311 -0.7546739
X =
0.03125 0. 0. 0. 0.
0. 0.015625 0. 0. 0.
0. 0. 1. 0. 0.
0. 0. 0. 0.5 0.
0. 0. 0. 0. 2.
--> [Ab,X]=bal(A)
Ab =
897.30729 -0.8453932 -32.694547 -14.690362 -9.3740507
-1.0860507 -0.0011318 -0.0022789 -0.0408643 -0.0486351
-24.217643 0.0495083 0.1467254 -0.2054765 -1.2331655
9.7013239 0.0052523 -0.452489 -0.3402948 6.4393174
22.23562 -0.0025801 0.2954562 -0.3042311 -0.7546739
X =
1. 0. 0. 0. 0.
0. 0.5 0. 0. 0.
0. 0. 32. 0. 0.
0. 0. 0. 16. 0.
0. 0. 0. 0. 64.
Related
I would like to solve a differential equation in R (with deSolve?) for which I do not have the initial condition, but only the final condition of the state variable. How can this be done?
The typical code is: ode(times, y, parameters, function ...) where y is the initial condition and function defines the differential equation.
Are your equations time reversible, that is, can you change your differential equations so they run backward in time? Most typically this will just mean reversing the sign of the gradient. For example, for a simple exponential growth model with rate r (gradient of x = r*x) then flipping the sign makes the gradient -r*x and generates exponential decay rather than exponential growth.
If so, all you have to do is use your final condition(s) as your initial condition(s), change the signs of the gradients, and you're done.
As suggested by #LutzLehmann, there's an even easier answer: ode can handle negative time steps, so just enter your time vector as (t_end, 0). Here's an example, using f'(x) = r*x (i.e. exponential growth). If f(1) = 3, r=1, and we want the value at t=0, analytically we would say:
x(T) = x(0) * exp(r*T)
x(0) = x(T) * exp(-r*T)
= 3 * exp(-1*1)
= 1.103638
Now let's try it in R:
library(deSolve)
g <- function(t, y, parms) { list(parms*y) }
res <- ode(3, times = c(1, 0), func = g, parms = 1)
print(res)
## time 1
## 1 1 3.000000
## 2 0 1.103639
I initially misread your question as stating that you knew both the initial and final conditions. This type of problem is called a boundary value problem and requires a separate class of numerical algorithms from standard (more elementary) initial-value problems.
library(sos)
findFn("{boundary value problem}")
tells us that there are several R packages on CRAN (bvpSolve looks the most promising) for solving these kinds of problems.
Given a differential equation
y'(t) = F(t,y(t))
over the interval [t0,tf] where y(tf)=yf is given as initial condition, one can transform this into the standard form by considering
x(s) = y(tf - s)
==> x'(s) = - y'(tf-s) = - F( tf-s, y(tf-s) )
x'(s) = - F( tf-s, x(s) )
now with
x(0) = x0 = yf.
This should be easy to code using wrapper functions and in the end some list reversal to get from x to y.
Some ODE solvers also allow negative step sizes, so that one can simply give the times for the construction of y in the descending order tf to t0 without using some intermediary x.
I want to construct a bijective function f(k, n, seed) from [1,n] to [1,n] where 1<=k<=n and 1<=f(k, n, seed)<=n for each given seed and n. The function actually should return a value from a random permutation of 1,2,...,n. The randomness is decided by the seed. Different seed may corresponds to different permutation. I want the function f(k, n, seed)'s time complexity to be O(1) for each 1<=k<=n and any given seed.
Anyone knows how can I construct such a function? The randomness is allowed to be pseudo-randomness. n can be very large (e.g. >= 1e8).
No matter how you do it, you will always have to store a list of numbers still available or numbers already used ... A simple possibility would be the following
const avail = [1,2,3, ..., n];
let random = new Random(seed)
function f(k,n) {
let index = random.next(n - k);
let result = avail[index]
avail[index] = avail[n-k];
}
The assumptions for this are the following
the array avail is 0-indexed
random.next(x) creates an random integer i with 0 <= i < x
the first k to call the function f with is 0
f is called for contiguous k 0, 1, 2, 3, ..., n
The principle works as follows:
avail holds all numbers still available for the permution. When you take a random index, the element at that index is the next element of the permutation. Then instead of slicing out that element from the array, which is quite expensive, you just replace the currently selected element with the last element in the avail array. In the next iteration you (virtually) decrease the size of the avail array by 1 by decreasing the upper limit for the random by one.
I'm not sure, how secure this random permutation is in terms of distribution of the values, ie for instance it may happen that a certain range of numbers is more likely to be in the beginning of the permuation or in the end of the permutation.
A simple, but not very 'random', approach would be to use the fact that, if a is relatively prime to n (ie they have no common factors), then
x-> (a*x + b)%n
is a permutation of {0,..n-1} to {0,..n-1}. To find the inverse of this, you can use the extended euclidean algorithm to find k and l so that
1 = gcd(a,n) = k*a+l*n
for then the inverse of the map above is
y -> (k*x + c) mod n
where c = -k*b mod n
So you could choose a to be a 'random' number in {0,..n-1} that is relatively prime to n, and b to be any number in {0,..n-1}
Note that you'll need to do this in 64 bit arithmetic to avoid overflow in computing a*x.
I'm trying to maximize the portfolio return subject to 5 constraints:
1.- a certain level of portfolio risk
2.- the same above but oposite sign (I need that the risk to be exactly that number)
3.- the sum of weights have to be 1
4.- all the weights must be greater or equal to cero
5.- all the weights must be at most one
I'm using the optiSolve package because I didn't find any other package that allow me to write this problem (or al least that I understood how to use it).
I have three big problems here, the first is that the resulting weights vector sum more than 1 and the second problem is that I can't declare t(w) %*% varcov_matrix %*% w == 0 in the quadratic constraint because it only allows for "<=" and finally I don't know how to put a constraint to get only positives weights
vector_de_retornos <- rnorm(5)
matriz_de_varcov <- matrix(rnorm(25), ncol = 5)
library(optiSolve)
restriccion1 <- quadcon(Q = matriz_de_varcov, dir = "<=", val = 0.04237972)
restriccion1_neg <- quadcon(Q = -matriz_de_varcov, dir = "<=",
val = -mean(limite_inf, limite_sup))
restriccion2 <- lincon(t(vector_de_retornos),
d=rep(0, nrow(t(vector_de_retornos))),
dir=rep("==",nrow(t(vector_de_retornos))),
val = rep(1, nrow(t(vector_de_retornos))),
id=1:ncol(t(vector_de_retornos)),
name = nrow(t(vector_de_retornos)))
restriccion_nonnegativa <- lbcon(rep(0,length(vector_de_retornos)))
restriccion_positiva <- ubcon(rep(1,length(vector_de_retornos)))
funcion_lineal <- linfun(vector_de_retornos, name = "lin.fun")
funcion_obj <- cop(funcion_lineal, max = T, ub = restriccion_positiva,
lc = restriccion2, lb = restriccion_nonnegativa, restriccion1,
restriccion1_neg)
porfavor_funciona <- solvecop(funcion_obj, solver = "alabama")
> porfavor_funciona$x
1 2 3 4 5
-3.243313e-09 -4.709673e-09 9.741379e-01 3.689040e-01 -1.685290e-09
> sum(porfavor_funciona$x)
[1] 1.343042
Someone knows how to solve this maximization problem with all the constraints mentioned before or tell me what I'm doing wrong? I'll really appreciate that, because the result seems like is not taking into account the constraints. Thanks!
Your restriccion2 makes the weighted sum of x is 1, if you also want to ensure the regular sum of x is 1, you can modify the constraint as follows:
restriccion2 <- lincon(rbind(t(vector_de_retornos),
# make a second row of coefficients in the A matrix
t(rep(1,length(vector_de_retornos)))),
d=rep(0,2), # the scalar value for both constraints is 0
dir=rep('==',2), # the direction for both constraints is '=='
val=rep(1,2), # the rhs value for both constraints is 1
id=1:ncol(t(vector_de_retornos)), # the number of columns is the same as before
name= 1:2)
If you only want the regular sum to be 1 and not the weighted sum you can replace your first parameter in the lincon function as you've defined it to be t(rep(1,length(vector_de_retornos))) and that will just constrain the regular sum of x to be 1.
To make an inequality constraint using only inequalities you need the same constraint twice but with opposite signs on the coefficients and right hand side values between the two (for example: 2x <= 4 and -2x <= -4 combines to make the constraint 2*x == 4). In your edit above, you provide a different value to the val parameter so these two constraints won't combine to make the equality constraint unless they match except for opposite signs as below.
restriccion1_neg <- quadcon(Q = -matriz_de_varcov, dir = "<=", val = -0.04237972)
I'm not certain because I can't find precision information in the package documentation, but those "negative" values in the x vector are probably due to rounding. They are so small and are effectively 0 so I think the non-negativity constraint is functioning properly.
restriccion_nonnegativa <- lbcon(rep(0,length(vector_de_retornos)))
A constraint of the form
x'Qx = a
is non-convex. (More general: any nonlinear equality constraint is non-convex). Non-convex problems are much more difficult to solve than convex ones and require specialized, global solvers. For convex problems, there are quite a few solvers available. This is not the case for non-convex problems. Most portfolio models are formulated as convex QP (quadratic programming i.e. risk -- the quadratic term -- is in the objective) or convex QCP/SOCP problems (quadratic terms in the constraints, but in a convex fashion). So, the constraint
x'Qx <= a
is easy (convex), as long as Q is positive-semi definite. Rewriting x'Qx=a as
x'Qx <= a
-x'Qx <= -a
unfortunately does not make the non-convexity go away, as -Q is not PSD. If we are maximizing return, we usually only use x'Qx <= a to limit the risk and forget about the >= part. Even more popular is to put both the return and the risk in the objective (that is the standard mean-variable portfolio model).
A possible solver for solving non-convex quadratic problems under R is Gurobi.
In some cases, a loop needs to run for a random number of iterations that ranges from min to max, inclusive. One working solution is to do something like this:
int numIterations = randomInteger(min, max);
for (int i = 0; i < numIterations; i++) {
/* ... fun and exciting things! ... */
}
A common mistake that many beginning programmers make is to do this:
for (int i = 0; i < randomInteger(min, max); i++) {
/* ... fun and exciting things! ... */
}
This recomputes the loop upper bound on each iteration.
I suspect that this does not give a uniform distribution of the number of times the loop will iterate that ranges from min to max, but I'm not sure exactly what distribution you do get when you do something like this. Does anyone know what the distribution of the number of loop iterations will be?
As a specific example: suppose that min = 0 and max = 2. Then there are the following possibilities:
When i = 0, the random value is 0. The loop runs 0 times.
When i = 0, the random value is nonzero. Then:
When i = 1, the random value is 0 or 1. Then the loop runs 1 time.
When i = 1, the random value is 2. Then the loop runs 2 times.
The probability of this first event is 1/3. The second event has probability 2/3, and within it, the first subcase has probability 2/3 and the second event has probability 1/3. Therefore, the average number of distributions is
0 × 1/3 + 1 × 2/3 × 2/3 + 2 × 2/3 × 1/3
= 0 + 4/9 + 4/9
= 8/9
Note that if the distribution were indeed uniform, we'd expect to get 1 loop iteration, but now we only get 8/9 on average. My question is whether it's possible to generalize this result to get a more exact value on the number of iterations.
Thanks!
Final edit (maybe!). I'm 95% sure that this isn't one of the standard distributions that are appropriate. I've put what the distribution is at the bottom of this post, as I think the code that gives the probabilities is more readable! A plot for the mean number of iterations against max is given below.
Interestingly, the number of iterations tails off as you increase max. Would be interesting if someone else could confirm this with their code.
If I were to start modelling this, I would start with the geometric distribution, and try to modify that. Essentially we're looking at a discrete, bounded distribution. So we have zero or more "failures" (not meeting the stopping condition), followed by one "success". The catch here, compared to the geometric or Poisson, is that the probability of success changes (also, like the Poisson, the geometric distribution is unbounded, but I think structurally the geometric is a good base). Assuming min=0, the basic mathematical form for P(X=k), 0 <= k <= max, where k is the number of iterations the loop runs, is, like the geometric distribution, the product of k failure terms and 1 success term, corresponding to k "false"s on the loop condition and 1 "true". (Note that this holds even to calculate the last probability, as the chance of stopping is then 1, which obviously makes no difference to a product).
Following on from this, an attempt to implement this in code, in R, looks like this:
fx = function(k,maximum)
{
n=maximum+1;
failure = factorial(n-1)/factorial(n-1-k) / n^k;
success = (k+1) / n;
failure * success
}
This assumes min=0, but generalizing to arbitrary mins isn't difficult (see my comment on the OP). To explain the code. First, as shown by the OP, the probabilities all have (min+1) as a denominator, so we calculate the denominator, n. Next, we calculate the product of the failure terms. Here factorial(n-1)/factorial(n-1-k) means, for example, for min=2, n=3 and k=2: 2*1. And it generalises to give you (n-1)(n-2)... for the total probability of failure. The probability of success increases as you get further into the loop, until finally, when k=maximum, it is 1.
Plotting this analytic formula gives the same results as the OP, and the same shape as the simulation plotted by John Kugelman.
Incidentally the R code to do this is as follows
plot_probability_mass_function = function(maximum)
{
x=0:maximum;
barplot(fx(x,max(x)), names.arg=x, main=paste("max",maximum), ylab="P(X=x)");
}
par(mfrow=c(3,1))
plot_probability_mass_function(2)
plot_probability_mass_function(10)
plot_probability_mass_function(100)
Mathematically, the distribution is, if I've got my maths right, given by:
which simplifies to
(thanks a bunch to http://www.codecogs.com/latex/eqneditor.php)
The latter is given by the R function
function(x,m) { factorial(m)*(x+1)/(factorial(m-x)*(m+1)^(x+1)) }
Plotting the mean number of iterations is done like this in R
meanf = function(minimum)
{
x = 0:minimum
probs = f(x,minimum)
x %*% probs
}
meanf = function(maximum)
{
x = 0:maximum
probs = f(x,maximum)
x %*% probs
}
par(mfrow=c(2,1))
max_range = 1:10
plot(sapply(max_range, meanf) ~ max_range, ylab="Mean number of iterations", xlab="max")
max_range = 1:100
plot(sapply(max_range, meanf) ~ max_range, ylab="Mean number of iterations", xlab="max")
Here are some concrete results I plotted with matplotlib. The X axis is the value i reached. The Y axis is the number of times that value was reached.
The distribution is clearly not uniform. I don't know what distribution it is offhand; my statistics knowledge is quite rusty.
1. min = 10, max = 20, iterations = 100,000
2. min = 100, max = 200, iterations = 100,000
I believe that it would still, given a sufficient amount of executions, conform to the distribution of the randomInteger function.
But this is probably a question better suited to be asked on MATHEMATICS.
I don’t know the math behind it, but I know how to compute it! In Haskell:
import Numeric.Probability.Distribution
iterations min max = iteration 0
where
iteration i = do
x <- uniform [min..max]
if i < x
then iteration (i + 1)
else return i
Now expected (iterations 0 2) gives you the expected value of ~0.89. Maybe someone with the requisite math knowledge can explain what I’m actually doing here. Because you start at 0, the loop will always run at least min times.
I use this algorithm to calculate the length of a quadratic bezier:
http://www.malczak.linuxpl.com/blog/quadratic-bezier-curve-length/
However, what I wish to do is calculate the length of the bezier from 0 to t where 0 < t < 1
Is there any way to modify the formula used in the link above to get the length of the first segment of a bezier curve?
Just to clarify, I'm not looking for the distance between q(0) and q(t) but the length of the arc that goes between these points.
(I don't wish to use adaptive subdivision to aproximate the length)
Since I was sure a similar form solution would exist for that variable t case - I extended the solution given in the link.
Starting from the equation in the link:
Which we can write as
Where b = B/(2A) and c = C/A.
Then transforming u = t + b we get
Where k = c - b^2
Now we can use the integral identity from the link to obtain:
So, in summary, the required steps are:
Calculate A,B,C as in the original equation.
Calculate b = B/(2A) and c = C/A
Calculate u = t + b and k = c -b^2
Plug these values into the equation above.
[Edit by Spektre] I just managed to implement this in C++ so here the code (and working correctly matching naively obtained arc lengths):
float x0,x1,x2,y0,y1,y2; // control points of Bezier curve
float get_l_analytic(float t) // get arclength from parameter t=<0,1>
{
float ax,ay,bx,by,A,B,C,b,c,u,k,L;
ax=x0-x1-x1+x2;
ay=y0-y1-y1+y2;
bx=x1+x1-x0-x0;
by=y1+y1-y0-y0;
A=4.0*((ax*ax)+(ay*ay));
B=4.0*((ax*bx)+(ay*by));
C= (bx*bx)+(by*by);
b=B/(2.0*A);
c=C/A;
u=t+b;
k=c-(b*b);
L=0.5*sqrt(A)*
(
(u*sqrt((u*u)+k))
-(b*sqrt((b*b)+k))
+(k*log(fabs((u+sqrt((u*u)+k))/(b+sqrt((b*b)+k)))))
);
return L;
}
There is still room for improvement as some therms are computed more than once ...
While there may be a closed form expression, this is what I'd do:
Use De-Casteljau's algorithm to split the bezier into the 0 to t part and use the algorithm from the link to calculate its length.
You just have to evaluate the integral not between 0 and 1 but between 0 and t. You can use the symbolic toolbox of your choice to do that if you're not into the math. For instance:
http://integrals.wolfram.com/index.jsp?expr=Sqrt\[a*x*x%2Bb*x%2Bc\]&random=false
Evaluate the result for x = t and x = 0 and subtract them.