I am writing a program that asks what the user would like to solve for and then pass that into the CAS Solve function. The program look like this:

[code]//Amplitude

export Amplitude()

BEGIN

Local Find,Find_1,variables,Answer;

Local Â,K,τ,ω;

print();

//K=5, Â=0.0355 τ=10 ,ω=3.1415 ,E=.5 ,A=7

A:=0;

CHOOSE(A,"Which Amplitude?","Final","Initial");

if A==1 then

variables:={"Â","K","τ","ω","E","A"};

input({{Find,variables,{56,15,0}},{Â,[0],{10,20,1}},{K,[0],{10,20,2}},{τ,[0],{10,20,3}},{ω,[0],{10,20,4}},{E,[0],{10,20,5}},{A,[0],{10,20,6}}});

Find_1:=variables[Find]; Find:=Find_1;

Answer:=CAS("SOLVE(Â=K*A/(sqrt((1-(τ*ω)^2)^2+(2*τ*ω*E)^2)),Find,1)");

print(Answer);

end;

END;[code]

When I run this it returns "Error: Constant function"

I have tried expr(Find) and also EVAL(Find) but both options failed to work.

I'm just a novice, but it looks like you mixing text and variables in the string expression. Try something like:

Answer:=CAS("SOLVE(Â=K*A/(sqrt((1-(τ*ω)^2)^2+(2*τ*ω*E)^2))," + Find + ",1)");

(10-01-2020 10:07 AM)Gene222 Wrote: [ -> ]I'm just a novice, but it looks like you mixing text and variables in the string expression. Try something like:

Answer:=CAS("SOLVE(Â=K*A/(sqrt((1-(τ*ω)^2)^2+(2*τ*ω*E)^2))," + Find + ",1)");

So it works better than before in that it doesn't return an error but it doesn't evaluate the equation. I normally ran into that problem when I didn't us the program in the solver app but I checked and made sure I was in the Solver app and it still was not running properly. I ran it in the CAS environment and it returned an error.

Code:

//Amplitude

export Sys_Ctrl_Amplitude()

BEGIN

Local Find,Find_1,variables,Answer;

print();

G:=0;

CHOOSE(G,"Which Amplitude?","Final","Initial");

if G==1 then

variables:={"Â","K","τ","ω","ζ","A"};

input({{Find,variables,{56,15,0}},{Â,[0],{10,20,1}},{K,[0],{10,20,2}},{τ,[0],{10,20,3}},{ω,[0],{10,20,4}},{ζ,[0],{10,20,5}},{A,[0],{10,20,6}}});

Find_1:=variables[Find]; Find:=Find_1;

//input({{K,[0],{10,20,2}},{τ,[0],{10,20,3}},{ω,[0],{10,20,4}},{ζ,[0],{10,20,5}},{A,[0],{10,20,6}}});

Answer:=CAS("SOLVE(Â=K*A/(sqrt((1-(τ*ω)^2)^2+(2*τ*ω*ζ)^2)),"+Find+",1)");

//print("Â= "+Answer);

print(Find+"= "+Answer);

end;

END;

I don't know how to get SOLVE or fsolve to solve your equation. Some comments on your first program are:

(1) Your choose variable A is the same as your variables[6] which is also defined as A.

(2) I tried using fsolve instead of SOLVE, but it gave {} which I think means fsolve could not find an answer for the initial guess.

(3) I tried exporting the solve variables, but that did not help.

Below are the changes I made to your program in trying to get the program to work. I predefined your list variables, because I got tired of inputting the values in the INPUT screen.

Sorry I could not help you.

PHP Code:

`EXPORT Â,τ,ω,A1,Answer;`

EXPORT Find,Find_1,variables;

export Amplitude()

BEGIN

print();

Â:=0.0355;

K:=5;

τ:=10;

ω:=3.1415;

E:=0.5;

A1:=7;

CHOOSE(A,"Which Amplitude?","Final","Initial");

if A==1 then

variables:={"Â","K","τ","ω","E","A1"};

input({{Find,variables,{56,15,0}},{Â,[0],{10,20,1}},{K,[0],{10,20,2}},{τ,[0],{10,20,3}},{ω,[0],{10,20,4}},{E,[0],{10,20,5}},{A1,[0],{10,20,6}}});

Find_1:=variables[Find]; Find:=Find_1;

Answer:=fsolve("Â=K*A/(sqrt((1-(τ*ω)^2)^2+(2*τ*ω*E)^2)),"+Find+"=1)");

print(Answer);

end;

END;

yes I tried that as well period on my calculator I already had them as global variables in another program labeled load_variables (). I just put them in as local variables so that people would not get lost reading the program. I may have to resort to waiting multiple If statements/Cases or figure out how to utilize the solver app. I'm reluctant to do this

1) because I don't know how to program utilizing an app

2) because multiple If statements would clutter up the code.

You might have a problem with the names you used for your variables. See the following posts below.

HP Prime Solver Variables Issue
Programming solve function
I also have had problems with expressions that included trig functions that gave multiple answers, where I had to either use an initial guess that would narrow the search for a solution, or put a limit on the solve variable such as 0.0001..2*pi.

I also had problems where the variables would give a division by zero error or where the change in the parameter variables would not significantly change the solve variable, so that fsolve could not find a solution. In one case, I ended up replacing fsolve with a trial and error/bisection method in order to find a solution.