ti-86 program... grr!

[drow_elf]

been writing a program to do definite integral approx. but i cant get the Simpson approxiamtion to work right..... both of the others seem to work perfectly... i suspect a variable is getting changed... but i dont see anything that is changing it... and MY logic must be perfect.... anyone see the problem? (i hate programming on these things... such a cramped screen and debuggin is a real pita)

explanation of variables: x-start interval to be integrated; L=end of interval;N=number of iterations;O=offset % for rieman sum.... ooh and several variables get used for different things through the prog...

(comments are denoted by #s)

 

0->A
Prompt x,L,N,O            #get those numbers! note: we're analysing the eqn y1
x->xx                           #store a dummy; we'll use it later
(L-x)/iPart(N)->N          #make sure that #%^& user entered a whole number!
x+N*O/100->x             #offset the start sample pt for rieman sum
for(x,x,L-(n*0/100),N)  # stmt L-(n*o/100) is untested (was using an if before)
y1+A->A
end
Disp "riemann:",N*A
(evalf(y1,x,xx)+evalf(y1,x,L))/2->A   #dividing 1st and last sample pt of trap integ by 2
for(x,xx+N,L-N,N)                               
y1+A->A
end
Disp "Trap:",N*A
(L-xx)/N->N                            #get back our orginal number N and make it even
if fPart (n/2)==.5:n+1->N
(L-xx)/N->N
evalf(y1,x,xx)+evalf(y1,x,L)->A   #puting the first and last sample in
For(x,xx+N,L-2N,2N)                    #all the odd samples
4*evalf(y1,x,x)+A->A
end
For(x,xx+2N,L-N,2N)                   #even ones
2*evalf(y1,x,x)+A->A
end
Disp "Simpson:",N*A/3            #output

[Allan]

It takes like forever to program ANYTHING on the TI's. You should get a PC cable and write the programs on your computer instead.

[drow_elf]

fixed it it needed l-2n on the very last for loop.... damn i hate programming on these things... that tiny prog (the longest non-assembly one on my calc right now!) takes 5 screens!