Calculators Calculator giving weird answers?

Click For Summary
SUMMARY

The forum discussion centers on the unexpected outputs of the sine function for multiples of pi on various calculators, specifically the TI-83 Plus and TI-36X Pro. Users report that while sin(4π) and sin(6π) yield values close to zero, they return small floating-point errors such as -2E-13. This behavior is attributed to the limitations of floating-point arithmetic in calculators, which can introduce rounding errors during calculations. The discussion highlights the importance of understanding how calculators handle transcendental functions and floating-point precision.

PREREQUISITES
  • Understanding of floating-point arithmetic
  • Familiarity with trigonometric functions
  • Knowledge of calculator models, specifically TI-83 Plus and TI-36X Pro
  • Basic graphing techniques
NEXT STEPS
  • Research floating-point precision and its impact on mathematical calculations
  • Explore the behavior of trigonometric functions in programming languages like Python
  • Learn about numerical methods for evaluating transcendental functions
  • Investigate the differences in calculator algorithms for trigonometric calculations
USEFUL FOR

Mathematicians, educators, students in mathematics or engineering, and anyone interested in the intricacies of calculator algorithms and floating-point arithmetic.

1MileCrash
Messages
1,338
Reaction score
41
For sin(4pi), my calculator gives -2^-13.

Obviously, that is really close to the correct value of 0, but it just says 0 for ANY other multiple of pi I've tried.

Weird right?

TI-83 plus.. common calculator. Anyone have issues like this?
 
Computer science news on Phys.org
Ha! Just tried it on my TI-83+ (from 1997), and it shows the same thing. Says the same for Sin4pi. Sin6pi returns 2E-13, and sin8pi returns -4E-13.

My TI-36X Pro, which I use almost exclusively now, returns 0.
 
Yep, I get those same results as you, never tried those, so my capital ANY was not a very meaningful "ANY," heh.

When it showed this error, I actually decided to graph "sin x" on it, then find the value when x = 4pi.. still returns -2e-13 that way.

Weird.
 
Welcome to the joys of floating point arithmetic.
 
hmmmm transcendentals likely calculated by a series...


and 10-13 is about 1/(2^43)

i'm guessing their internal binary arithmetic is good to ~43 bits.
Anybody got an old HP31?

old jim
 
It must be doing the 4*pi calculation first, incorporates some rounding error, then takes the sin of that.
 
Moonbear said:
It must be doing the 4*pi calculation first, incorporates some rounding error, then takes the sin of that.

But it works fine for something like pi/2 or even 10pi.
 
Try 8 pi.
 
um, so, I was kind of bored this evening and I saw this thread. Since I love excel and graphs... and numbers, I just started punching multiples of pi into my TI- 83 Plus calculator and noting down the results from 1pi to 100pi.

I got these graphs

the first is the whole shebang, the second is a zoom in on the first portion of the graph

it's interesting I guess, but I don't know enough about calculators to have any idea of why it does it like this. I'm going to keep going, since I have nothing else to do tonight, and see what happens on out from 100pi.

also there is the excel file if anyone feels like looking at the numbers and messing with them
 

Attachments

  • calculator stuff.png
    calculator stuff.png
    23.2 KB · Views: 601
  • calculator stuff2.png
    calculator stuff2.png
    22.3 KB · Views: 577
  • calculator stuff.xls
    calculator stuff.xls
    37.5 KB · Views: 351

Similar threads

Replies
22
Views
4K
  • · Replies 2 ·
Replies
2
Views
10K
  • · Replies 10 ·
Replies
10
Views
26K
  • · Replies 1 ·
Replies
1
Views
9K
  • · Replies 15 ·
Replies
15
Views
9K
  • · Replies 4 ·
Replies
4
Views
10K
  • · Replies 4 ·
Replies
4
Views
9K
  • · Replies 5 ·
Replies
5
Views
7K
  • · Replies 3 ·
Replies
3
Views
6K
  • · Replies 1 ·
Replies
1
Views
9K