I am trying to solve an optimization problem with several constraints and it is not working. The decision variables are the matrix entries.

Below is the code:

restart;
interface(displayprecision = 4): with( plots ):
with(linalg):with( Optimization );
[ImportMPS, Interactive, LPSolve, LSSolve, Maximize, Minimize,

NLPSolve, QPSolve]
f:=proc(x1,x2,x3,x4,x5,x6)
global lambda,mu,rho,Ls;
local eq,Lsq,g,P,n,IM,ImP,ImPi,c0,cb,Sol,i,j,t1,t2,fact,t3,t4,t5,Wq,W,Lq,L,Ws;
n:=7;
g:=array(1..n,[5,0,0,0,0,0,0]);
mu:=array(1..n,[10,5,5,5,5,5,5]);
P:=matrix([[0,x1,x2,0,0,0,0],[0,0,0,x3,x4,0,0],[0,0,0,0,0,x5,x6],
[0,0,0,0,0,0,0],[0,0,0,0,0,0,0],[0,0,0,0,0,0,0],[0,0,0,0,0,0,0]]);
IM := array(identity, 1..n,1..n):
ImP:=evalm(IM-P):
ImPi:=inverse(ImP):
lambda:=evalm(g&*ImPi):
for i from 1 to n do
rho[i] := lambda[i]/mu[i]
od:
c0:=array(1..n,[1/2,0,0,0,0,0,0]);
cb:=array(1..n,[1/5,1/5,1/5,1/5,1/5,1/5,1/5]);
for i from 1 to n do
eq[i]:=(g[i]/lambda[i])*c0[i] + sum((lambda[j]/lambda[i])*P[j,i]*((P[j,i]*(rho[j]^2*cb[j] +(1-rho[j]^2)*cx[j] ))+ (1 -P[j,i])),j=1..n)
od:
Sol:=fsolve({eq[1]-cx[1]=0,eq[2]-cx[2]=0,eq[3]-cx[3]=0,eq[4]-cx[4]=0,eq[5]-cx[5]=0,eq[6]-cx[6]=0,eq[7]-cx[7]=0},{cx[1],cx[2],cx[3],cx[4],cx[5],cx[6],cx[7]}):
assign(Sol):cx:
for i from 1 to n do
t1:= -2*(1-rho[i])/(3*rho[i]):
t2:= ((1-cx[i])^2)/(cx[i]+cb[i]):
fact := exp(t1*t2):
if cx[i] >= 1 then
fact:=1:
else
fact:
fi:
t3:=rho[i]/(1-rho[i]):
t4:= (cx[i]+ cb[i])/2:
t5:=1/mu[i]:
Wq[i] := (t3*t4*t5*fact):
W[i] := Wq[i] + t5:
Lq[i] := lambda[i]*Wq[i]:
L[i] := lambda[i]*W[i]:
od:
Ls:=add(L[i],i=1..n);Lsq:=add(Lq[i],i=1..n):Ws:=Ls/add(g[i],i=1..n):
RETURN(Ls):
end proc:

# here are the constraint procedures to ensure the probability pairs sum to one

p1 := proc (x1, x2) x1+x2-1 end proc;
proc(x1, x2) ... end;
p2 := proc (x3, x4) x3+x4-1 end proc;
proc(x3, x4) ... end;
p3 := proc (x5, x6) x5+x6-1 end proc;
proc(x5, x6) ... end;

sol := Optimization:-NLPSolve(f, {p1}, {p2}, {p3}, 0 .. 1, 0 .. 1, 0 .. 1, 0 .. 1, 0 .. 1, 0 .. 1, initialpoint = [.5, .5, .5, .5, .5, .5]);

Error, (in Optimization:-NLPSolve) unexpected parameters: {p3}

It seems to say that the problem are the constraints but this seems odd.

My PhD involves generating equations that are very long, in some cases, more than 10 pages long.

For the reader its much more useful to have some of these equations described than presented. One of the descriptions I'd like to use is number of characters and intuitively the command Length feels like it should tell you the number of characters in an expr

However what is written on the help page for this function I don't understand:
https://www.maplesoft.com/support/help/maple/view.aspx?path=length

then the example that is given is

Could anyone explain to me what is meant when it is said length(x+2y)=9? if I try an apply things recursively the expression with the most chatracters i can generate is "x+y+y"which is only 5 characters.


Is there another command that would be better suited to my purposes?

Hi,

i would like to plot a graph in R^3 of a function f(n,t), where n is integer and t is real. For every t i would like to have a sequence of points. Is it possible?

thanks, anna rita 

hello everyone, 

To illustrate my question I will get Bob and Alice 

Suppose Bob and Alice work separately on the same subject and must share some information.
They decide to exchange information among themselves by using .m files (Maple internal format files).

-------------------

1) Bob performs the first serie of calculus which ends by some "result". 
This result is the information Bob must send to Alice for her doing her part of the job.
Bob assigns this result to a variable named MyVar and saves it in a .m file :

.....
MyVar := ... :
save  MyVar   "/.../MyFile.m"

Then Bob indicates to Alice the complete path to MyFile.m ... but forgets to say her the name of the variable he has saved.

-------------------

2) Alice reads the file MyFile.m (read "/.../MyFile.m") .... 
Of course, if Alice knew the name of the variable Bob used in his "save" command, it would be the simplest thing in the World for her to browse the content of the .m file (eval(MyVar) generally works well).

But I assumed it is not the case, so my question :

Can Alice recover the name of the variable and visualize it ? 


Any answer will be greatly appreciated


postscript : it is not the first time I find myself in the situation Alice faces here. Often I have to recover the content of .m files written years ago by people who have changed their core business without having properly documented (like Bob) what these files contain

How do we specify the length of the string in external calling if we don't know the length beforehand?

In a local directory I have a number of worksheets that uses the same procedures, so I want to collect these procedures in a package, where the file with the package should be stored locally, since it is only relevant locally.

Is that possible, and how can it be done ?

Hi,

I have to find the root of an equation corresponding to the maximum absolute value. I am using root finding package to get all the roots. But after getting all the roots i am not able to apply abs function. Maple sheet is attached.

Download question.mw

I will be really thankful for the help.

Regards

Sunit

I'm running into a very simple problem with the way that Maple integrates Heaviside functions. Naively, it should act like a step function, but it is not integrating properly. See the attached document.

Note that the symbolic integration of the Heaviside function (defined to be 1 inside the unit circle and 0 outside) gives zero, whereas it should clearly give the area of the unit circle, which the numerical integration does. I even checked that the (suposedly equivalent) piecewise definition symbolically evaluates to the area, and it, too, gets the right answer.

Anyone have any clue as to why the symbolic integration of this Heaviside function is so wrong? My understanding is that if we do the integral as two nested 1D integrals, the returned function (as a function of y) is zero everywhere except at y=0, but that result cannot be right either.

Thoughts?

Download Heaviside-error.mw

I am trying to find the optimal routing probabilities in a Maple procedure where the Mean Value Analysis is used to compute the queueing values. The Maple code is below. It first tries to compute the visit ratios where the probability routing values are the decision variables. There is one specified constraint on the sum of the probability decision variables.

restart;
interface(warnlevel=0): interface(displayprecision = 4): with( plots ):
with(linalg):with( Optimization ); with(Student[NumericalAnalysis]):
[ImportMPS, Interactive, LPSolve, LSSolve, Maximize, Minimize,

NLPSolve, QPSolve]
f:=proc(x1,x2,x3)
global T,lambda,nq,u;
local i,j,pop,Sum;
n:=3;N:=2;M:=3;
#
# Gauss-Seidel iterations
#
A:=Matrix([[1,-x1,-x2],[0,1,-x3],[0,0,1]]);
b:= Vector([1,0,0]);
v := IterativeApproximate(LinearAlgebra:-Transpose(A), b, initialapprox = Vector([1, 3/4, 3/4]), tolerance = 10^(-3), maxiterations = 20, stoppingcriterion = relative(infinity), method = gaussseidel);
mu:=array(1..n,[2.0,1.0,1.0]);
nq:=array(1..M,[0,0,0]);# must initialize queue lengths
for i from 1 to N do

pop:=i;
for j from 1 to M do # mean waiting times
T[j]:=t[j]*(1 + nq[j]) od;
Sum := 0.0;
for j from 1 to M do # mean cycle time
Sum := Sum + v[j]*T[j] od;
for j from 1 to M do #compute the throughputs
lambda[j] := (v[j]*pop)/Sum od;
for j from 1 to M do #compute the queue lengths
nq[j]:= lambda[j]*T[j] od;
for j from 1 to M do #compute the utilizations
u[j]:= lambda[j]*t[j] od;
od;
RETURN(lambda[1]);
end proc;

proc(x1, x2, x3) ... end;

sol := Optimization:-NLPSolve(f, {}, {proc (x1, x2, x3) options operator, arrow; x1+x2+x3-5/3 end proc}, 0 .. 1, 0 .. 1, 0 .. 1, initialpoint = [.75, .25, .6667], assume = nonnegative); 1;

Error, (in Optimization:-NLPSolve) non-numeric result encountered

I am not sure why I get the error message

I was trying to call a C shared library which calls a Python function. I got the following error:

Error, external linking: error loading external library

/directory/of/my/library/libbe.so: libpython2.7.so.1.0: cannot open shared object

file: No such file or directory

Since I was running this using a pbs file, I made sure that I exported LD_LIBRARY_PATH (to libpython.so) inside the specific node. Does anyone have ideas on why this happens? Thank you!

Hello people in mapleprimes,

I have a question. I hope someone give an answer to me.

rho:=sqrt((x-a)^2+(y-b)^2);

limit(x-a+y-2,rho=0);

brings an error

Error, invalid input: limit expects its 2nd argument, p, to be of type Or(name = algebraic, set(name = algebraic)), but received ((x-a)^2+(y-b)^2)^(1/2) = 0

Isn't other way than the following?

limit(x-a+y-2,{x=a,y=b});

Best wishes.

taro

I am working in Document mode, 2D input.  I have doing some complex number evalutions

Lets say I have a complex number Z= 2+3i

when I ask maple for the "abs(Z)" I get the display of "abs(2+3i)" rather than the numberic answer. This occurs with all the complex operation arg, abs, polar conversion, 

polar(abs(2+3*i), argument(2+3*i))   this is from Maple. 

How do I display the numeric value for these functions in my document

Thanks Bill

Hello,

I've got a load of data that contains a lot of noise in the form of sinusoidal interference patterns. This is quite thick and it disturbes what I am trying to look at. I've uploaded a picture to represent what I am trying to show. I know I need a cut-off frequency I just don't know how to implement it. Thank you in advance for your help!

Gambia Man

Hi,
In maple18 has:

with(Student[Basics][ExpandSteps])

but how to see step-by-step factor ??

(Sorry! I'm not very good at English.)

Thank you!

Dear Community,

I've made a nonlinear curve fit with the Minimize routine (see attachment). What would be an easy and elegant way to rerun the model (Model) with the fitted values of a, b, c and plot the result together with the measured points in the same chart? I'm stuck here.

Tx in advance,

best regards

Andras

BroSzem_Data.xlsx

Nonlin_Curve_Fit.mw