Help with Solving 5a & Understanding Recursion Relations

Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
16 replies · 4K views
cloud360
Messages
212
Reaction score
0
Would be really grateful if someone helped me with 5a, and explained the ideas behind recursion relations. I don't know what a recursion relation is, and how to apply for forum ales given

Homework Statement


[PLAIN]http://img839.imageshack.us/img839/2301/recurrencerelation.gif
http://img839.imageshack.us/img839/2301/recurrencerelation.gif


Homework Equations



none

The Attempt at a Solution




∫xne-x=-xne-x-∫(-nxn-1e-x)

but then i have to keep integrating forever


how do i do 5a, any help, any online references to learn?
 
Last edited by a moderator:
Physics news on Phys.org
hi cloud360! :smile:

(have an integral: ∫ :wink:)
cloud360 said:
integrate xne-x=-xne-x-|(-nxn-1e-x)

but then i have to keep integrating forever


but you have proved the main part of 5(a) ! :smile:

(and what's "forever" about going from n down to zero? :confused:)
 


tiny-tim said:
hi cloud360! :smile:

(have an integral: ∫ :wink:)


but you have proved the main part of 5(a) ! :smile:

(and what's "forever" about going from n down to zero? :confused:)


ok, so if i have to integrate something forever. what notation can i use to represent this.

i keep seeing the gamme function, but don't know what its used for

(my teachers like putting things on hwk that we never did in class)
 


ok, i heard soemthing about homegenous function which go down to 0?

but don't really know what to do
 
cloud360 said:
ok, i heard soemthing about homegenous function which go down to 0?

but don't really know what to do

no, homogeneous functions are irrelevant …

your recurrence relation is Kn = nKn-1

so for example K7 = 7K6 = 7*6K5, and so on down to K0, where it stops! :wink:
 


tiny-tim said:
no, homogeneous functions are irrelevant …

your recurrence relation is kn = nkn-1

so for example k7 = 7k6 = 7*6k5, and so on down to k0, where it stops! :wink:

k7 = 7k6 = 7*6k5
=7*6*5k4=7*6*5*4k3 = 7*6*5*4*3k2= 7*6*5*4*3*2k1= 7*6*5*4*3*2*1k0


I know very litle about recurssion relations. But know, if i have 2 squences with same recurssion relations. then sequence 1 - sqeuence 2 = 0

I have the solutions to the question i posted. but don't know why its the answer

[PLAIN]http://img231.imageshack.us/img231/3002/recussion2.gif
 
Last edited by a moderator:


tiny-tim said:
That's right! :smile:

Similarly, if you know K1/2, then you can find K3/2, then you can find K5/2, and so on. :wink:


Question: is this factorial llike thing... where you keep multiplying downwards e.g 5*4*3..e.t.c (does this happen for all recursion relations), or only this. does it depend on forumale they give us?

K3.5
K3.5=3.5*2.5*1.5*k0.5
K3.5=3.5*2.5*1.5*0.5*(π0.5)
K3.5=11.6317
 
Last edited:


Still. please can you tell me what i would do to work out 5a. i got an exam on this tomorrow. and 5a, and one like it are only things i don't know how to do
 
cloud360 said:
K3.5=3.5*2.5*1.5*0.5*(π0.5)
K3.5=11.6317

Yes, or you can write it K(2n+1)/2 = (2n+1)(2n-1)… 1/2n K1/2

= (2n+1)(2n)… 1/2n2n(2n-2)… 2 K1/2

= (2n+1)!/22n(2n)! K1/2 :smile:

(which is easier for large n because you can look up factorials in a book or on your calculator :wink:)
Question: is this factorial liek this where you keep mutipying downards e.g 5*4*3..e.t.c (does this happen for all recursion relations)?

No, it just happens to work for this one.
cloud360 said:
Still. please can you tell me what i would do to work out 5a. i got an exam on this tomorrow. and 5a, and one like it are only things i don't know how to do

ah, when I posted my last answer, I hadn't see this, which you edited in later:
cloud360 said:
I have the solutions to the question i posted. but don't know why its the answer

[PLAIN]http://img231.imageshack.us/img231/3002/recussion2.gif[/QUOTE]

which part of that is worrying you?
 
Last edited by a moderator:


tiny-tim said:
Yes, or you can write it K(2n+1)/2 = (2n+1)(2n-1)… 1/2n K1/2

= (2n+1)(2n)… 1/2n2n(2n-2)… 2 K1/2

= (2n+1)!/22n(2n)! K1/2 :smile:

(which is easier for large n because you can look up factorials in a book or on your calculator :wink:)


No, it just happens to work for this one.


ah, when I posted my last answer, I hadn't see this, which you edited in later:


which part of that is worrying you?
[PLAIN]http://img839.imageshack.us/img839/2301/recurrencerelation.gif

the first question


5a...(i)

∫uv'=uv-u'v
u=xn
dv=e-x

1. i don't understand why they are using limits to show the recursion relation holds for the given equation?
The way i would do it is by working out K0 , K1 , K2...And find a nth term formulae?
2. I don't know how to get K0 , K1 , K2 by direct integration> the question as for K0 by direct integration, what does this even mean

[PLAIN]http://img231.imageshack.us/img231/3002/recussion2.gif

in fact you said that i did the main part of the question when i posted this

∫xne-x=-xne-x-∫(-nxn-1e-x)

i then said i would need to integrate forever because of this ∫(-nxn-1e-x)
 
Last edited by a moderator:
cloud360 said:
in fact you said that i did the main part of the question when i posted this

∫xne-x=-xne-x-∫(-nxn-1e-x)

i then said i would need to integrate forever because of this ∫(-nxn-1e-x)

but you don't have to integrate forever, because the value of n keeps falling each time …

after n integrations, you get down to ∫e-x dx, which of course you can do immediately
 


tiny-tim said:
but you don't have to integrate forever, because the value of n keeps falling each time …

after n integrations, you get down to ∫e-x dx, which of course you can do immediately

So you get e^-x by using limits.

I don't understand the second line of the solutions. please can you explain

[PLAIN]http://img231.imageshack.us/img231/3002/recussion2.gif


why are they subbing infinity into the equation. and why sub infinity into :

-xne-x

where did this come from?

were did they get this from.
 
Last edited by a moderator:


this is the solution to part b (see first post and check black area)

[PLAIN]http://img709.imageshack.us/img709/7861/ans2k.gif


I don't understand why they don't use limits in this question to prove the recursion relations holds for the given forumale? but did for part a

i can't see any consistency. therefore i don't know the general way of solving/proving a recursion relation holds for a given equation
 
Last edited by a moderator:


2. I don't know how to get K0 for part a

can you help me with this. i don't even know what it means "by direct integration". i have seen the soluton but don't understand why they sub "0" into x, and not n?


[PLAIN]http://img231.imageshack.us/img231/3002/recussion2.gif
 
Last edited by a moderator:
cloud360 said:
… I don't understand the second line of the solutions. please can you explain

why are they subbing infinity into the equation. and why sub infinity into :

-xne-x

where did this come from?

were did they get this from.

oh i see … it's just the [xn(-e-x)]0 that's worrying you …

ok, that's ∞n(-e-∞) - 0n(-e-0), wchih is 0 - 0

except of course we can't really write the first part like that, we have to put a limn->∞ in front of it :wink:
cloud360 said:
… I don't understand why they don't use limits in this question to prove the recursion relations holds for the given forumale? but did for part a

i can't see any consistency. therefore i don't know the general way of solving/proving a recursion relation holds for a given equation

because the limits were there only because ∞ was one of the bounds of the integration, ∫0, so you got a […]0, which involved taking a limit

in most recurrence relations, there's either no integral at all, or there's an integral with finite bounds …

you do not normally get limits in a recurrence relation!

cloud360 said:
2. I don't know how to get K0 for part a

can you help me with this. i don't even know what it means "by direct integration". i have seen the soluton but don't understand why they sub "0" into x, and not n?

K0 by definition is just ∫0 e-x dx, = [-e-x]0 = 1.
 


tiny-tim said:
oh i see … it's just the [xn(-e-x)]0 that's worrying you …

ok, that's ∞n(-e-∞) - 0n(-e-0), wchih is 0 - 0

except of course we can't really write the first part like that, we have to put a limn->∞ in front of it :wink:


because the limits were there only because ∞ was one of the bounds of the integration, ∫0, so you got a […]0, which involved taking a limit

in most recurrence relations, there's either no integral at all, or there's an integral with finite bounds …

you do not normally get limits in a recurrence relation!



K0 by definition is just ∫0 e-x dx, = [-e-x]0 = 1.

i am so grateful for your help. thanks a lot !

the 0-0 part was the thing i couldn't picture