When finding a formula for truncation error, why do you only take the absolute value of the top?
Ex: \[\frac{ 128lxl ^{7} }{ 1-2x }\]
Why don't I take the absolute value of the bottom part?
can you show me the process whereby we formulate a truncation error?
What do you mean?
They gave me a function 1/(1-2x) and they asked for the error associated with P6(x)
do we just magically have formulas? or are they developed thru more basic measures, like taking limits of this or that and working it out.
So I generated the series for the function using a/(1-r), and then I created a series for every term after the 6th term
using a/(1-r) again
so we want to create a polynomial that represents 1/(1-2x) if we cut off (truncate) the poly at the 6th degree; then the error of truncation is the sum of the parts from 7 to infinity ... am i understanding this correctly?
Exactly
i havent tried to work it so ill ahve to review some stuff first to refresh my memory on this. if none else come by with an answer ill post what i can dig up.
Ok thanks. I got the answer, but I didn't understand the absolute value thing
one thing ive got in the back of my head has to do with an alternating error thrm; the size of the error of truncation is +- no more than, the first neglected term. but i cant be sure that this applies here or not. so im reviewing :)
post your work if you can ...
Ok. the function was 1/(1-2x). So I said a=1 and r=2x, so I generated the series 1+2x+4x^2+8x^3+16x^4+32x^5+64x^6+128x^7+256x^8... I cut it off after 64x^6, and I made a series a/(1-r) with 128x^7 as the first term. That series was 128x^7/(1-2x)
and the series converges of course if |2x| < 1, or |x|<1/2 its just my thought, but i think it has to do with this .... but i could be completely wrong at the moment :)
They did give me the interval (-1/2,1/2), so I dont know if that had anything to do with it
lagrange form of Rn \[R_n=f^{(n+1)}(c)\frac{(x-a)^{n+1}}{(n+1)!}\le \left|f^{(n+1)}(c)\frac{(x-a)^{n+1}}{(n+1)!}\right|\]
Yes, so why do I only have to put absolute value bars around the numerator?
1+2x+4x^2+.... -------------------- 1-2x | 1 1-2x 2x 2x-4x^2 4x^2 ---------------------------- \[\frac{1}{1-2x}=\sum_{n=0}^{\infty}~2^nx^n\] \[\sum_{n=7}^{\infty}~2^nx^n\le\left|\frac{2^7x^7}{1-2x}\right|\]
1-2x < 0 2x > 1 x > 1/2 becasue the bottom is never negative 1-2(-1/2) = 2 1-2(1/2) = 0 the bottom is from 0 to 2
2^7 is never negative ... so ... whats that leave?
spose we have x=-1/3 the top goes negative right?
for x = -1/2 to 0, the top goes negative, the bottom stays positive ...
-1/2 < x < 1/2 -1 < -2x < 1 0 < 1-2x < 2 notice the bottom is always positive in this interval but x^7 is negative from -1/2 < x < 0 ------------------------------- -1/2 < x < 1/2 -1/2^7 < x^7 < 1/2^7 |x|^7 <= 1/2
I'm really confused
tellme what about this explanation is not making sense
I just don't understand how you are proving that the bottom doesn't need abs val bars
when is 1-2x less than zero? when is it negative?
When x>2
x>1/2
sorry
x is between -1/2 and 1/2 right? now work x into 1-2x given: -1/2 < x < 1/2 , multiply by -2 -2/2 < -2x < 2/2 , add 1 1-1 < 1-2x < 1+1 0 < 1-2x < 2 since the bottom can only be values from 0 to 2 .... its ALWAYS positive for a given x value in its domain. right?
Ooooh ok.
if A and B are always postive then \[\left|\frac{Ax}{B}\right|\implies \frac{A|x|}{B}\]
if k is even: x^k is always positive, but 7 is not an even number :)
Oooh ok. That makes much more sense! So I technically could put the bars around the whole thing?
yes, it wouldnt make a difference either way
IF the stated domain made the bottom negative at any point ... then we would keep it covered in | | but theres no real need for it in this case
I see! Thanks so much!
youre welcome
Join our real-time social learning platform and learn together with your friends!