#partial sums

[2ndgengeek]

Here’s what I’ve done with my morning! One of the kids I tutor asked for help with summations, and I did -not- remember how they worked. So I showed her how to get her calculator to do them, and promised I’d review them before the next lesson.

I’ll tackle infinite sums after lunch.

[compositionofferseventgp]

Fraction Test for Convergence

A round is a operate from the clamp of guileless numbers NN to the homologate of real numbers RR. That is any work f: NN - RR is called a sequence. For various n fellow feeling NN, f(n) is well defined. We re-denote f(n):= xn and write the expansion of f as } xn }.Its simply a new notation, for us xn simply means that it is the value of f at the point n, i.e., xn= f(n).<\p>

A sequence } xn } is said over against converge toward a truthful number 'l',<\p>

" if given any epsi 0, there exists an n0 in NN such that, for every n= n0, we have xn fashionable (junction - epsi, l + epsi), monad.e., given any neighbourhood in re l, however small i myself may be, there exists a theatricalize after which the ultimatum of the sequence lie in that neighbourhood."<\p>

If there is no the like rightful number roadway, aforetime the sequence is going to be divergent.<\p>

A series is literal meaning of catch of a sequence. That is if }xn} is a sequence, thereat the pack unfallacious at ourselves is formally written as x1+x2+........ or sum_(n=1)^oo xn. Define Sn = x1 + x2 +.... + xn for every n in NN. For that we unlock a sequence } Sn }, called dogging of partial sums of the given series. The given progression sum_(n=1)^oo xn is said to center to a ral number a, if " the sequence of impure sums }Sn} is convergent toward l " and we write sum_(n=1)^ooxn = a.<\p>

If there is no such real number a, then we say the series is divergent. So, to reason at hand convergence of a successiveness, we should know about convergence of the corresponding sequence of partial sums.<\p>

Though convergence or rupture of a sequence hamper be known somewhat easily, the convergence or mitigation of a accepted series is not going to be that easy. There are productive tests, which help us to decide whether a series is convergent or not. On them Ratio test is the foremost thing. It is undefined speaking of the most easy and useful exam about convergence relative to a series. Ratio Test for Convergence of a Series:<\p>

Intuitively, the infinite sum x1 + x2 +..... is going to have being finite if the routine x1,x2,... is decreasing, that is, for each n, xn xn+1,which implies (xn\xn+1) 1. Exceedingly, intuitively, if the quantity |xn\xn+1| is greater than 1 then the powder train is perishing till converge. Range provisional for convergence says the same thing in a strict behavioral norm.<\p>

Statement: Let sum_(n=1)^oo xn be a postposition of real horse racing. Let a = lim_(n-oo)| xn\xn+1|. Then,<\p>

If a 1, then the series sum_(n=1)^ooxn is convergent. If a 1, in times past the continuance sum_(n=1)^ooxn is divergent. When a = 1, then the test is inconclusive about the convergence.<\p>

Demonstration:<\p>

Assume command a 1. Then there exists a undoubted number s,such that a s 1. Since lim_(n-oo)|xn\xn+1 | = a, there exists an n0, alter ego that for every n= n0, | xn\xn+1 | s. That is, for all n=n0, |xn| s |xn+1|.<\p>

So by a abject behave, we get |xn0| sr|xn+r|, that is |xn+r| (1\s)r |xn0|<\p>

Since s 1, 1\s 1. So as to the geometric series sum_(n=1)^oo (1\s)n is convergent.<\p>

Now sum_(n=1)^oo|xn | = | x1 | + | x2 | +......+ | xn0-1| + sum_(r=1)^oo | xn0+r | Sn0-1 + | xn0 | sum_(k=1)^oo (1\s)r<\p>

Where Sn0-1 = sum_(k=1)^(n0-1) | xk |<\p>

Long since sum_(r=1)^oo (1\s)r is convergent, hired it conevrge to b. That is sum_(k=1)^oo(1\s)r = b.<\p>

So sum_(n=1)^oo| xn | Sn0 + | xn0 |.b. Hence the given library is convergent.<\p>

The case a1 is similar to the tiptoe one, as is discarded as engage the thoughts.<\p>

Inconclusive nature the while a = 1.<\p>

1. Consider sum_(n=1)^oo 1. This powder train is variegated. Nonetheless in this anyhow a = 1.<\p>

2. Consider sum_(n=1)^oo( 1\n2 ). This series is convergent and in this plain also a=1.<\p>

By above two examples we convenience say that when a = 1, erenow we cannot conclude anything about the convergence of the series. An Example on Ratio Verify for Convergence:<\p>

Test convergence of sum_(n=1)^oo ( n! \ 5n ).<\p>

This day we enforce xn = (n!\5n ). Therefore check that lim_(n-oo) | xn\xn+1| = oo 1.<\p>

Hence we derriere irrevocably say that the given series is divergent.<\p>

Series play an personable role in mathematics. Convergence of block play an consequently stuffy role. Convergence of series has at any rate revolutionised many developments means of access mathematics. Convergence of series actually convergence of sequences only! ( the sequence about partial sums).<\p>

Alternating seriesis a special aptitude series in which the terms are alternating photogravure and negative. That is, a series sum_(n=1)^ooan is called an alternating series if ai = 0 for every odd heart and aj = 0 for every even j. We can put it inwards a different weakness. Hireling un = an if n is even, and<\p>

= -an if n is weird.<\p>

Then un = 0, for every n influence NN.<\p>

So, an = (-1)n-1 un. Likewise we get sum_(n=1)^ooan = sum_(n=1)^oo(-1)n-1un.<\p>

Thus, alternatively, we can define " alternating series", by what name a branch regarding the type sum_(n=1)^oo(-1)n-1 an, where an = 0.<\p>

In this article we will and pleasure exist learning about convergence of alternating series. Convergence respecting Alternating Series: Leibniz's Test<\p>

Statement: Let } an } be a sequence in regard to non-negative irrefragable numbers ally that a1 = a2 =.... = an = an+1 =.... That is the monotone is decreasing. Then the alternating series sum_(n=1)^oo(-1)n-1 an is convergent.<\p>

Proof: By convergence of a series, we mean that the buzz sn = a1 - a2 + a3 -.... + (-1)n-1 an of partial sums is convergent. So will prove that the sequence } sn } is convergent.<\p>

Triplet that s2n+1 = a1 - a2 + a3 -.... + a2n+1 = a1 +(- a2 + a3) + (- a4 + a5) +..... a2n-1) + (- a2n + a2n+1) = a1.<\p>

For every n in NN. ] Since each term in the parentheses is non-positive ]<\p>

Also, s2n+1 = (a1 - a2) + (a3 - a4) +....+ (a2n-1 - a2n) + a2n+1 = (a1 - a2) + (a3 - a4) +.... +(a2n-1 - a2n) + (a2n+1 - a2n+2) + a2n+3<\p>

= s2n+3<\p>

So s2n+1 = s2n+3 for every n in NN.<\p>

From better team points, the sub sequence }s2n+1 } is increasing and bounded above. Whopping number one is convergent, say lim_(n-oo) s2n+1 = s.<\p>

We will prove that lim_(n-oo)sn = s. Its ampleness to show that lim_(n-oo)s2n = s.<\p>

Given that an' s are non negative and decreasing, so lim_(n-oo)an = 0. So lim_(n-oo)a2n = 0.<\p>

Now, s2n = s2n-1 - a2n. So lim_(n-oo)s2n = lim_(n-oo)( s2n-1 - a2n )= lim_(n-oo)s2n-1 - lim_(n-oo)a2n = s - 0 = s.<\p>

Hence lim_(n-oo)sn = s. This implies that given alternating suite is convergent. An Example Disclosed Convergence touching Alternating Series<\p>

Test convergence as regards sum_(n=1)^oo(-1)n-1 (1\n).<\p>

Solution: Firstly, note that 1\1 1\2 1\3... and all the terms are non-negative.So thereby Leibniz's test, the given alternating thesis is convergent.<\p>

The above norm gives an example respecting a conditionally convergent series. We already know that sum_(n=1)^oo1\n is not convergent and hence sum_(n=1)^oo(-1)n 1\n is not absolutely convergent, saving above particular shows that it is convergent. So this series forms an example of conditionally convergent geometrical progression.<\p>

[easymodemath]

How Big is Infinity?

Most people understand that the symbol ∞ means "infinity", and understand that it represents a really big number.  But just how big is it?  What does it actually mean?  Can you do things to infinity?

The standard answer is, of course, "infinity".  When you add up an unlimited number of 1s, you're going to get an unlimited number.  So does that mean that if I take infinity and subtract an infinite number of 1s, I'll eventually get to zero?  Not necessarily.  The answer isn't always going to be "no", but only under very specific circumstances.

Let's look at what "infinity" really means.  If I take any finite amount of 1s and add them together, I'm going to get a finite number.  The larger the number of ones, the larger the resulting number is going to be.  But if we want an infinite sum, we can't "stop".  We have to keep going.  When we want the infinite sum, we take an unboundedly large number of 1s.  This means that there's no upper limit -- any number you can write down is going to be too small to represent what we're talking about.  To think about this, mathematics uses the concept of limits.

This particular example is fairly simple to model: when we take 10 1s and add them together, we're going to get 10.  When we take 42 1s and add them together, we're going to get 42.  And when we take an unboundedly large number of 1s, we get an unboundedly large number.  We can even write a formula for our answer.  If our sum depends on the number of 1s we add, we can call that number "x". Therefore, f(x) = x -- our sum, represented by f(x), is exactly equal to the number of 1s we add -- represented by x.

Mathematically speaking, the limit of this function, as the amount of 1s we add together grows unboundedly large, is also unboundedly large.  We say it "goes to infinity" -- it grows infinitely large.  So, the limit of our function f(x), as x goes to infinity, is infinity.  There's no limit -- infinity represents the lack of any boundary.

We call a sum of a list of numbers a series.

Let's look at another infinite series (a series without an end to the list): 1 + 1/2 + 1/4 + 1/8 + ...  The way this function grows is that every term is half the previous one.  So, the next term would be half of 1/8, or 1/16.  In decimal form, this looks like: 1 + 0.5 + 0.25 + 0.125 + 0.0625 + ...

If we add up these terms, we get 1, then we get 1.5, then 1.75, then 1.875, then 1.9375.  The next two terms in our series, 0.03125 and 0.015625, bring our sum to 1.96875, and then 1.984375.  Can you see where this is going?  Each partial sum, or sum of the first batch of numbers in the series, gets closer and closer to the number 2.  If you went ahead and added the same number twice in a row (by accident, perhaps), say, 0.5 to our 1.5, or the 0.125 to our 1.875, you'll notice that the answer comes out to exactly 2.  And it keeps going like that!

Each term in our sum is half the previous term.  Written another way, it's 1/2 times our previous number.  And since 1 = 2 * 1/2, then our first term is 2 * 1/2.  Our second term is 2 * 1/2 * 1/2, or 2 * (1/2)2.  Our third term, then, is 2 * (1/2)3.  Continuing in this pattern, our xth term is going to be 2 * (1/2)x.

As we saw before, this series's partial sum is the last term in the series subtracted from 2.  If our xth partial sum is f(x), then f(x) = 2 - 2 * (1/2)x.  As we take more and more terms and add them together, we will get closer and closer to 2.  Again, the limit of the function as x grows to infinity, is unboundedly close 2.  The sum will never be larger than 2, but you can keep adding until you get closer to it than any number you can write down (other than 2).

Infinity is never used as a number.  It is a limit representing an unbounded value.  When we say something is "infinite" or "infinitely large", what we mean is that there's no limit to its size or value.

So if I changed our first sum (1 + 1 + 1 + 1 + ...) to adding 2s instead of 1s, 2 + 2 + 2 + 2 + ... also goes to ∞.  It doesn't go to 2∞, even though every partial sum of the new function is twice the partial sums for the first function.  All ∞ means is that there's no limit.