KhiCAS, le moteur formel Giac/Xcas pour toutes TI-Nspire

[krystalgamer]

It is known that KhiCAS is having issues, especially on recent versions. There isn't much that can be done about it.
Your revision (O) won't allow to early versions that might have been more stable in this regard, no.

Anyway, I shall remind you that putting a CAS engine on your non-CAS device will be considered cheating on exams relying on such non-CAS devices ; as such, if you get caught, you'll be in trouble.

Thanks for the answer, so im partially screwed . I'm a very good student but my problem is that during exams i solve them really fast(90 minutes exams i finish it having 20 or so) i endup commiting mistakes like switching signs or wrong covnersion i.e 200g=2kg, which i dont realise until it's too late. I use the calculator mostly to verify if i've done everything correctly.

I have some programming knowledge that i can apply to ndless sdk to code some neat stuff.

[Adriweb]

Well sure, but the proctors won't care

[krystalgamer]

Well sure, but the proctors won't care

Already thought on everything.
Start exam without ndless and install it after it starts(modified version with a different name) + clear recent files.
Will write my own math lib + some hooks on the keyboard so can access it very easily.
If Proctor comes to check run the hook to clear recent files + close current + open random file.

[Adriweb]

Oh but I'm not saying there is no way to trick people into thinking you were not cheating (as per their non-CAS rules, anyway), and as you describe there is at least one, but I'm rather talking about if you do get caught, one way or another - they won't care about why you did use a CAS engine in the first place

(BTW, because such tool(s) would clearly break exam rules, it wouldn't be accepted into TI-Planet archives (if you had intended to upload it), nor probably into other communities' - IDK if you had been in the community for some time, but for instance this also happened with stuff that messed with the Press-To-Test mode. We are very anti-cheating so of course, we have to remove that if it gets uploaded)

Exam boards in general are just clueless (I'm being nice in my choice of words...) about recent technology and clearly don't have rules updated for the world we live in today - not only would it be better if CAS was allowed everywhere (with exams adapted for that), but at least they should just rely on the PTT protections when all devices have it - they wouldn't have to worry about people using CAS on non-CAS (which is why TI is constantly fighting Ndless, they have to be able to tell the stupid exam boards they make devices following their rules)

[fjallbacka]

I've installed it, but I don't see any difference between this and the TI engine. Which other functionalities are?

[compsystems]

It would be interesting to create a comparison table to determine which is better TICAS or xcas, or complement the following comparison

http://www.technicalc.org/benchmark/benchmark.pdf

[Adriweb]

It'd be interesting to do that on a bigger scale, though like unit tests made of hundreds of thousands / millions of items.
I don't know how xcas is [unit]tested, though.

[parisse]

Xcas has its own regression tests (in the source check subdirectory), it is also checked against geogebra and HP database. Khicas is not well tested however, and some of the regression tests would certainly fail (i.e. reboot the calc) : one should consider khicas in a beta-release state, and that will not change until I find an efficient way to debug on the nspire, the current toolchain I'm using does not support gdb-ing in the giac source (if someone can compile and debug the giac source code with the latest toolchain, I'd be interested to have some feedback).

[Adriweb]

Oh yeah, but even comparing the CAS engines themselves (not from KhiCAS) would be interesting.
I assume it'd be easy to make giac/xcas read a big file containing expressions to be executed and output the results in a file.

It is also relatively easy to make the Nspire Computer Software's CAS engine programmatically try out lots of input expression and get the output.

Then we could compare outputs, for instance.

[parisse]

This is already available for giac on the desktop, and that's how my regression test work. Just create a text file with your commands, and run it from a terminal
giac filename
You can also export a session from Xcas in text format (File menu, export as xcas test).
You can of course redirect output to a file.

BTW, I'm reinstalling the toolchain and firebird, but I can't find any documentation on how to use firebird to debug (well there is not much documentation on firebird in general, I lost a full hour before I could send a file to the emulator, and I'm not sure it's the best way to do it. That's unfortunate because it seems to be more powerful than the previous nspire-emu).

For example, my regression test for limits

Code: Select all

1;
limit( exp(x)*(exp(1/x-exp(-x))-exp(1/x)), x,inf);
2;
limit( exp(x)*(exp(1/x+exp(-x)+exp(-x**2))-exp(1/x-exp(-exp(x)))), x,inf);
3;
limit( exp(exp(x-exp(-x))/(1-1/x))-exp(exp(x)), x,inf);
4;
limit( exp(exp(exp(x)/(1-1/x)))-exp(exp(exp(x)/(1-1/x-ln(x)**(-ln(x))))), x,inf);
5;
limit( exp(exp(exp(x+exp(-x))))/exp(exp(exp(x))), x,inf);
6;
limit(exp(exp(exp(x)))/exp(exp(exp(x-exp(-x)))),x,inf);
7;
limit(exp(exp(exp(x)))/exp(exp(exp(x-exp(-exp(exp(x)))))),x,inf);
8;
limit(exp(exp(x))/exp(exp(x-exp(-exp(exp(x))))),x,inf);
9;
limit(x*ln(x)*ln(x*exp(x)-x**2)**2/ln(ln(x**2+2*exp(exp(3*x**3*ln(x))))),x,inf);
10;
limit((exp(x*exp(-x)/(exp(-x)+exp(-2*x**2/(x+1))))-exp(x))/x,x,inf);
11;
limit((3**x+5**x)**(1/x),x,inf);
12;
limit(x/ln(x**(ln(x)**(ln(2)/ln(x)))),x,inf);
13;
limit(exp(exp(2*ln(x**5+x)*ln(ln(x))))/exp(exp(10*ln(x)*ln(ln(x)))),x,inf);
14;
limit((exp(4*x*exp(-x)/(1/exp(x)+1/exp(2*x**2/(x+1))))-exp(x))/exp(x)**4,x,inf);
15;
limit(exp(x*exp(-x)/(exp(-x)+exp(-2*x**2/(x+1))))/exp(x),x,inf);
16;
limit( exp(exp(-x/(1+exp(-x))))*exp(-x/(1+exp(-x/(1+exp(-x)))))*exp(exp(-x+exp(-x/(1+exp(-x)))))/exp(-x/(1+exp(-x)))**2-exp(x)+x, x,inf);
17;
limit( (log(log(x)+log(log(x)))-log(log(x)))*log(x)/log(log(x)+log(log(log(x)))), x,inf);
18;
limit(exp(log(log(x+exp(log(x)*log(log(x)))))/log(log(log(exp(x)+x+log(x))))),x,inf);
19;
limit(exp(x)*(sin(1/x+exp(-x))-sin(1/x+exp(-x**2))),x,inf);
20;
limit(exp(exp(x))*(exp(sin(1/x+exp(-exp(x))))-exp(sin(1/x))),x,inf);
21;
limit( (exp(x*exp(-x)/(exp(-x)+exp(-2*x**2/(x+1))))-exp(x))/x,x,inf);
22;
limit( log(log(x*exp(x*exp(x))+1))-exp(exp(log(log(x))+1/x)), x,inf);
23;
limit(sqrt(x+sqrt(x))-sqrt(x),x,inf);
24;
limit(sqrt(x+sqrt(x+sqrt(x)))-sqrt(x+sqrt(x)),x,inf);
25;
limit(4/9*exp(exp(5/2*x**(-5/7)+21/8*x**(6/11)+2*x**(-8)+54/17*x**(49/45)))**8/log(log(-log(4/3*x**(-5/14))))**(7/6),x,inf);
26;
limit(log(x)**2*exp(sqrt(log(x))*log(log(x))**2*exp(sqrt(log(log(x)))*log(log(log(x)))**3))/sqrt(x),x,inf);


for integration

Code: Select all

cas_setup(0,0,0,1,0,[1e-10,1e-17],12,[1,50,0,25],0,0,0),xcas_mode(0);
integrate(1/(x**4-1)**10,x);
integrate((x**4+4*x**2+6*x+4)/(x+1)**2,x);
integrate(x/((x-1)*(x+1)**2),x);
integrate(x/((x+1)*(x**4-1)),x);
integrate(1/(3*x*(x**2+x+1)*(x-1)**3),x);
integrate(1/(x**4-1)**2,x);
integrate(1/(x**4+1)**2,x);
integrate(1/(x**4+1)**4,x);
integrate(x**7/((x**4-1)*(x**2+3)),x);
integrate(((1+x)/(1-x))**(1/3),x);
integrate((sin(2*x)+1)/(cos(2*x)),x);
integrate((2*sin(x)+1)/(2*sin(x)-1),x);
integrate(exp(x)/(3+2*exp(x)),x);
integrate(sin(x)**2*cos(x)**4,x);
integrate(sin(x)/(sin(x)**3+cos(x)**3),x);
integrate(1/sqrt(2*x*t)*exp(-t**2/2),x);
integrate(sin(3*x)**4/exp((3*x+1)/cos(t)),x);
integrate(2*x/sqrt(x**2-1),x);
integrate((sin(x)+cos(x))/(sin(x)-cos(x)),x);
integrate(sin(pi/2-2*x),x);
integrate(tan(x)+tan(x)**3,x);
integrate((exp(x)-exp(-x))/(exp(x)+exp(-x)),x);
integrate(1/cos(x)**25,x);
integrate(1/(sin(x)-1)**3,x);
integrate(1/(sin(x)+1)**3,x);
integrate(ln(x+sqrt(1+x**2)),x);
integrate(atan(2*x/(1+x**2)),x);
integrate(x*sqrt(1+x**2),x);
integrate(sin(x)/cos(x)**2,x);
integrate(exp(x)/(1+exp(2*x)),x);
integrate(cos(x/2)**2/(x+sin(x)),x);
integrate(x/sqrt(x+1),x);
integrate(exp(x)/((3+exp(x))*sqrt(exp(x)-1)),x);
integrate(sqrt(x)/sqrt(a**3-x**3),x);
integrate(sqrt(a-x)/sqrt(x),x);
integrate(sqrt(x**2+a**2),x);
integrate(sin(2*x)*cos(x),x);
integrate(x*atan(x),x);
integrate(sinh(x)*cos(x),x);
integrate(atan(x)/x,x);
integrate(1/(sin(x)-2)**3,x);
integrate((2*x**2+1)*exp(x**2),x);
integrate(sin(3*x)/sin(x),x);
integrate(1/(t*ln(t)**2),t,2,x);
integrate(ln(1+2/(n*(n+3))),n,1,+infinity);
integrate((pi*t-t**2)*sin(n*t),t,0,pi);
integrate(exp(t)*cos(n*t),t,-pi,pi);
integrate(cos(x)/exp(x),x,0,+infinity);
integrate((t**4+t+1)/(t**6+t**3+2),t,1,+infinity);
assume(t>2),integrate(x*exp(1/2*abs(ln(x**2))),x,2,t),purge(t);
integrate(1/sqrt(2*x*t)*exp(-t**2/2),x,a,b);
integrate((x**2*(1-x))**(1/3),x,0,1);
normal(integrate(atan(sqrt(1-x**2)),x,0,1));
assume(x>0),integrate(ln(x**2+t**2)/(1+t**2),t,0,+infinity),purge(x);