QNA > I > La Radice Quadrata Di N È Divergente O Convergente?
Domanda

La radice quadrata di n è divergente o convergente?

Risposte
01/29/2022
Dollie

A questa domanda si risponde molto facilmente una volta che impariamo il test dell'integrale per la convergenza delle serie. Esso afferma che se l'integrale della sequenza converge, converge anche la somma, e viceversa. Vediamo i risultati per [math]\dfrac{1}{n}[/math] e [math]\dfrac{1}{n^2}[/math]

[math]\displaystyle \int_{1}^{\infty}{dfrac{1}{n}}=\ln n|_{1}^{\infty}=\ln \infty -\ln 1=\infty -0 =\infty[/math]

[math]\displaystyle \int_{1}^{\infty}{\dfrac{1}{n^2}}=-\dfrac{1}{n}|_{1}^{\infty}=-\dfrac{1}{\infty}+1=0+1=1[/math]

Hence [math]\displaystyle \sum_{n=1}^{\infty} \dfrac{1}{n}[/math] diverge dove come [math]\displaystyle \sum_{n=1}^{\infty} \dfrac{1}{n^2}[/math] converge.

02/07/2022
Sair Hebard

I’m not sure if the right tools exist to answer this question. What decides about convergence or divergence is basically how often [math]\sin(n)[/math] comes “close” to [math]0[/math]. In other words we need to know how well [math]n[/math] approximates multiples of [math]\pi[/math]. Or, to put it another way, how well can we approximate [math]\pi[/math] by rational numbers?

The first thing that comes to mind is Dirichlet's approximation theorem. What this says is that there are infinitely many natural numbers [math]p[/math] and [math]q[/math] such that

[math]\displaystyle \left|\pi-\frac{p}{q} \right|<\frac{1}{q^2}.[/math]

What this implies is that

[math]\displaystyle |\sin(p)| =|\sin(p-q\pi)|\leq\frac{1}{q}<\frac{4}{p}[/math]

for infinitely many [math]p[/math]. The problem is that this result leads us nowhere. Sì, è bello sapere che [math]\sin(n)[/math] diventa arbitrariamente piccolo, ma anche se sapessimo che è esattamente [math]0[/math] infinitamente spesso, questo non ci aiuterebbe (potrebbe per esempio essere il caso solo se [math]n=m^2[/math]).

Quindi se volessimo dimostrare che la serie diverge, avremmo bisogno di risultati molto migliori del teorema di approssimazione di Dirichlet. Non vogliamo solo sapere che [math]\sin(n)[/math] diventa piccolo infinitamente spesso, dobbiamo sapere quanto frequentemente questo accade.

D'altra parte, cosa succede se cerchiamo di dimostrare che la serie converge? Questo significa che [math]n[/math] si allontana dai multipli di [math]\pi[/math] abbastanza spesso. Questo richiederebbe di avere qualcosa come il teorema di Dirichlet, ma nell'altra direzione. E abbiamo infatti tali risultati. Si può dimostrare che esiste [math]\alpha>2[/math] tale che

[math]\displaystyle \left|\pi-\frac{p}{q} \right|\geq\frac{1}{q^{\alpha}}[/math]

for all but finitely many pairs [math](p,q)[/math] (see Liouville number). Similar to the things above we then find that

[math]\displaystyle |\sin(p)|=|\sin(p-q\pi)|\geq \frac{K}{p^{\alpha-1}}[/math]

for some constant [math]K[/math] and almost all [math]p[/math]. This “almost all” sounds way better than just infinitely many when we speak about series, but it is still not enough. At this stage we have established that

[math]\displaystyle \frac{1}{n^{1+|\sin(n)|}} \leq \frac{1}{n^{1+Kn^{1-\alpha}}}[/math]

for almost all [math]n[/math]. This looks great for a comparison test. Well, if there weren’t the problem that

[math]\displaystyle \sum_{n=1}^{\infty} \frac{1}{n^{1+Kn^{\beta}}}[/math]

diverges for all [math]\beta<0[/math]. Questo può, per esempio, essere dimostrato con il test di condensazione di Cauchy.

Quindi, in ogni caso, penso che abbiamo bisogno di risultati molto più forti sulle approssimazioni razionali di [math]\pi[/math]. Non sono a conoscenza di nulla che possa aiutare qui. Ma forse qualcun altro lo è.

02/17/2022
Boone Haselhuhn

Se una serie [math]\\sum\,a_{n} \è convergente,[/math]allora [math]\,\sum\,a_{n} ^{2} \,, \,[/math] la serie dei quadrati dei termini corrispondenti non deve necessariamente essere convergente.

Per esempio, la serie[math]\,\sum\,a_{n} \,[/math]dove [math]\,a_{n} =\frac{(-1)^{n+1}}}{sqrt{n}}},[/math]è convergente, ma [math]\,\,\sum\,a_{n} ^{2} \è divergente.

Ma se una serie [math]\sum,\sum,a_{n} [/math]di termini positivi è convergente, allora [math]\\,\,\sum\,a_{n} ^{2} \,, \,[/math] è anche convergente poiché

[math]\,\,\sum_{j=1}^{j=n}\,a_{j} ^{2} \le\le\le, s_{n}.s_{n} \ che è delimitata in alto per qualsiasi [math]\,n\,,,[/math]dove [math]\,s_{n} \è l'ennesima somma parziale di [math]\,\sum\,a_{n} \\sum\,[/math].

02/24/2022
Randie

Il dottor Hartig ha dato la risposta corretta, cioè che questo è falso. Il suo controesempio è [math]\sum_{n=1}^\infty \frac{1}{n^2} [/math] che converge e tuttavia [math]\sum_{n=1}^\infty \sqrt{frac{1}{n^2}} [/math] diverge. Per qualche ragione i bot di Quora dicono che questo deve essere migliorato. Quindi questo subirà lo stesso destino? Senza dubbio.

P.S. Radice quadrata di una serie convergente non significa nulla, ma possiamo indovinare che questo è ciò che l'OP sta chiedendo.

03/10/2022
Esmerelda Batara

Hi there!!

First of all you did not mentioned weather it's a series or a sequence. This will not matter in this case as both sequence and series converges or diverge together but in other cases such as 1/n sequence converge but the series diverge.

Now, coming to your question

  • Lim (n)^1/2 multiplying both numerator and denominator with (n)^1/2 will gives us n/(n)^1/2 .
  • Now as n tends to infinity we know that numerator(n) will grow faster than denominator(n^1/2) i.e the term will be infinite and is divergent by nth term test for a series.
  • When it comes to sequence lim x=0 for it to be convergent and here it's not.

So, both series and sequence are divergent :)

03/23/2022
Storz Somdevi

Vuoi dire perché il limite come n va all'infinito del ∑ da x = 1 a n di [math]1/x[/math] diverge? Ecco perché:

1 + 1/2 + 1/3 + 1/4 + 1/5 + 1/6 + 1/7 + 1/8 + 1/9 + 1/10 + 1/11 + ... + 1/x

Il primo termine è maggiore di 1/2. Il secondo è 1/2.

I due termini successivi sono 1/3 e 1/4. La loro somma è maggiore di 1/2.

1/3 + 1/4 > 1/2

Ora prendiamo i prossimi quattro termini, 1/5, 1/6, e 1/7 e 1/8. La loro somma è maggiore della metà.

1/5 + 1/6 + 1/7 + 1/8 > 1/2

I prossimi otto termini sono 1/9, 1/10, 1/11, 1/12, 1/13, 1/14, 1/15 e 1/16. Anche la loro somma è più di 1/2.

1/9 + 1/10 + 1/11 + 1/12 + 1/13 + 1/14 + 1/15 + 1/16 > 1/2

La somma dei 16 termini successivi è più di 1/2. La somma dei 32 termini da 1/33 a 1/64 è maggiore di 1/2. E così via: Non esauriremo mai il doppio degli ultimi 2^n valori della sequenza, quindi possiamo dire che possiamo sempre aggiungere 1/2 all'ultima serie di valori della serie. C'è una fornitura eterna di 1/x da aggiungere nella nostra prossima somma, non importa quanto grande sia il numero x. Questi ultimi termini della serie originale si comportano sempre come ∑ 1/2, e questo significa che il limite man mano che n si avvicina all'infinito di ∑ 1/x da x=1 o x=n non va mai a zero, e diverge.

03/31/2022
Bernadette

[math]\displaystyle \lim_{n\to\infty} \dfrac{1}{n\sqrt{\ln(n)}}=\lim_{n\to\infty} \dfrac{1}{n}\times\lim_{n\to\infty} \dfrac{1}{{sqrt{\ln(n)}}=0\times 0 = 0[/math]

[math]n{sqrt{\ln(n)}[/math] aumenta all'aumentare di [math]n[/math].

Quindi la sequenza [math]s_n==dfrac{1}{n\sqrt{\ln (n)}}[/math] è una sequenza monotona decrescente.

Siccome il limite di [math]s_n[/math] è [math]0[/math] ed è monotono decrescente, quindi la somma

[math]\displaystyle \sum_{n=2}^{\infty} \dfrac{(-1)^n}{n\sqrt{\ln(n)}[/math] converge per il test della serie alternata (AST)

04/06/2022
Esbenshade Nusom

There seems to be some confusion or uncertainty in the question concerning the placement of the factorial symbol. So for the sake of generality I will consider three cases.

a) The case where we have [math]\displaystyle \sum _{n=0}^{\infty } \frac{1}{\sqrt{n!}}.[/math]

In this case it can be shown that the series converges by the ratio test.

A numerical solution to the infinite sum can be found with Mathematica by typing the code:

  1. NSum[1/Sqrt[n!], {n, 0, Infinity}, WorkingPrecision -> 30]

The result or answer obtained is:

[math]\displaystyle \boxed{\sum _{n=0}^{\infty } \frac{1}{\sqrt{n!}} \approx 3.469506314521047562476}[/math]

b) The case where we have [math]\displaystyle \frac{1}{(\sqrt{n})!}.[/math]

In this case we have to deal with non integers or with irrational numbers, so it is best to write the sum in the form [math]\displaystyle \sum _{n=0}^{\infty } \frac{1}{\Gamma \left(\sqrt{n}+1\right)},[/math] since [math]n! = \Gamma(n+1).[/math]

It can be shown that this series converges by the comparison test.

A numerical solution to this infinite sum can be found with Mathematica by typing the code:

  1. SetPrecision[Sum[1/Gamma[Sqrt[n] + 1], {n, 0, Infinity}], 30]

The result or answer obtained is:

[math]\displaystyle \boxed{\sum _{n=0}^{\infty } \frac{1}{\Gamma \left(\sqrt{n}+1\right)} \approx 6.050864944678733075929218044803507662787}[/math]

c) The (unlikely) case where we have [math]\displaystyle \left(\frac{1}{\sqrt{n}}\right)!.[/math]

For [math]n = 0[/math] we get an infinite expression, so it is best to write the sum in the form [math]\displaystyle \sum _{n=1}^{\infty } \Gamma \left(\frac{1}{\sqrt{n}}+1\right).[/math]

It can be shown by the limit test that the series diverges and it can be verified with Mathematica that the infinite sum does not converge.

04/12/2022
Pompea

[math]lim_{n\rightarrow\infty}( \sqrt{n-1} - \sqrt{n})[/math]

[math]=lim_{n\rightarrow\infty}\frac{(\sqrt{n-1} - \sqrt{n})(\sqrt{n-1} + \sqrt{n})}{sqrt{n-1}+\sqrt{n}} [/math]

[math]=lim_{n\rightarrow\infty}\frac{n-1-n}{\sqrt{n-1}+\sqrt{n}} [/math]

[math]=lim_{n\rightarrow\infty}\frac{-1}{\sqrt{n-1}+\sqrt{n}} [/math]

[math]=0 [/math]

Dare una risposta
Le piante grasse possono essere legnose? :: Qual è il valore della radice 19?
Link utili