• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

The Math Help Thread

Status
Not open for further replies.
Here's the question
6nP1R4R.png

And (what I think is) the relevant theorem
ddQlwtB.png


I thought about making it an upper triangular matrix, but it seems like the calculations required are beyond what should just be an intuition, specifically to see what properties of the matrix make its determinant = 0.
 
I know the following identity: (sin(x))^2 + (cos(x))^2 = 1.

Is it also true for (sin(2x))^2 + (cos(2x))^2 = 1? What matters is that the inside of the sine and cosine are the same, right?
 
I know the following identity: (sin(x))^2 + (cos(x))^2 = 1.

Is it also true for (sin(2x))^2 + (cos(2x))^2 = 1? What matters is that the inside of the sine and cosine are the same, right?

Yeah, of course.

It is probably more obvious if you state it like this:

x = 2y

sin(2y)^2 + cos(2y)^2 = sin(x)^2 + cos(2y)^2 = 1
 
Here's the question
And (what I think is) the relevant theorem

I thought about making it an upper triangular matrix, but it seems like the calculations required are beyond what should just be an intuition, specifically to see what properties of the matrix make its determinant = 0.

A matrix has a 0 determinant if one of its rows is nothing but 0s. By Theorem (c) a matrix whose rows aren't linearly independent will also have a 0 determinant. You want to show that this is the case - that one row can be expressed as a linear combination of the other rows. Notice what happens if you just sum all the rows together.
 
A matrix has a 0 determinant if one of its rows is nothing but 0s. By Theorem (c) a matrix whose rows aren't linearly independent will also have a 0 determinant. You want to show that this is the case - that one row can be expressed as a linear combination of the other rows. Notice what happens if you just sum all the rows together.
Ahh, then there's a row of 0s. Appreciate the help! =)
 
I'm not sure what you mean by the y and z planes, but the definition of P tells you that any point in the plane must have the same y and z coordinates. So your parametrization is forced to look like (a(t), b(t), b(t)). Are you familiar with any parametrizations of the sphere? You could think of this problem as asking you to find the intersection of the sphere of radius 2 with the plane P and work from there.

I had a go, I got the parameterization (2cos t, -2sin t, 2sin t) for 0<t<pi, which I plotted in Matlab and it looks right...
 
I had a go, I got the parameterization (2cos t, -2sin t, 2sin t) for 0<t<pi, which I plotted in Matlab and it looks right...

This gives you r^2 = x^2 + y^2 + z^2 = 4(cos^2 t + sin^2 t + sin^2 t) = 4(1+sin^2 t), so your radius is dependent on t. You want r^2 = 4, independent of the value of t.
 
This gives you r^2 = x^2 + y^2 + z^2 = 4(cos^2 t + sin^2 t + sin^2 t) = 4(1+sin^2 t), so your radius is dependent on t. You want r^2 = 4, independent of the value of t.

Oh i see, its not a semicircle then. the issue i'm having is the "the semicircle lies in the plane given by y=z >=0". i'm a bit lost - for intersecting a sphere with the plane, the equation of the plane is y=z and the sphere is x^2+y^2+z^2=4? Sorry for being slow

e/ had another look, is it (2cos t, 2^0.5 sin t, 2^0.5 sin t)? gives x^2+y^2+z^2=4
 
Oh i see, its not a semicircle then. the issue i'm having is the "the semicircle lies in the plane given by y=z >=0". i'm a bit lost - for intersecting a sphere with the plane, the equation of the plane is y=z and the sphere is x^2+y^2+z^2=4? Sorry for being slow
No matter how you find your parametrization, you need the y- and z-coordinates to be equal since y=z in the plane. So you could try something like (2cos t, 2sin t, 2sin t). But then you get r^2 = 4(1+sin^2 t) again, so maybe a different choice of coefficients would be better? Look at (a cos t, b sin t, b sin t), and find values of a and b that do give you a semicircle of radius 2, i.e. that give you (a cos t)^2 + (b sin t)^2 + (b sin t)^2 = 4 = r^2.
 
is it (2cos t, 2^0.5 sin t, 2^0.5 sin t)? gives x^2+y^2+z^2=4

Yes, that's one possibility. Since you're taking linear algebra, let's come up with an explanation using matrices and vectors, rather than guessing the parametrization.


You had the right idea initially, where you tried to visualize the plane P = {(x, y, z) | y = z} being half-way in the y-z plane (in red lines in my drawing above).

For simplicity, let's forget about the z-axis and just imagine a circle of radius 2 on the x-y plane (in black lines). This circle can be parametrized by,

[x(t); y(t); z(t)] = [2*sin(t); 2*cos(t); 0], for 0 <= t <= 2*pi.

We see that, if we rotate this x-y plane by 45 degrees about the x-axis, we will get our plane P, and the circle that we want for our problem.


So the question just boils down to what is A, the 3 x 3 matrix that will transform/rotate the x-y plane to the plane P?

Recall that the entries of a matrix come from how a linear map acts upon the basis vectors. From the drawing, we know that we want the following to occur:

A * [2; 0; 0] = [2; 0; 0]

A * [0; 2; 0] = [0; sqrt(2); sqrt(2)]

A * [0; 0; 2] = [0; -sqrt(2); sqrt(2)].


Hence,

A * (2*I) = [2 0 0; 0 sqrt(2) -sqrt(2); 0 sqrt(2) sqrt(2)].

=>

A = 1/2 * [2 0 0; 0 sqrt(2) -sqrt(2); 0 sqrt(2) sqrt(2)].


Now that we know the transformation matrix, we can multiply A and the vector [x(t); y(t); z(t)] above to get the parametrization of the transformed circle.

We get,

A * [x(t); y(t); z(t)] = [2*sin(t); sqrt(2)*cos(t); sqrt(2)*cos(t)], for 0 <= t <= 2*pi.

Note that the transformed y and z values share the same value.
 
Here's the question
6nP1R4R.png

And (what I think is) the relevant theorem
ddQlwtB.png


I thought about making it an upper triangular matrix, but it seems like the calculations required are beyond what should just be an intuition, specifically to see what properties of the matrix make its determinant = 0.


You just need to show that one eigenvalue is zero. The eigenvalues vector is easy. It is just the vector with all ones.
 
Anyone care to help me with this problem.

"Sketch the region enclosed by the given curves and finds its area"

y=sec^2x, y=8cosx, -pi/3 &#8804;x&#8804;pi/3
 
Anyone care to help me with this problem.

"Sketch the region enclosed by the given curves and finds its area"

y=sec^2x, y=8cosx, -pi/3 &#8804;x&#8804;pi/3

Well first you need to draw out the graph and find the upper and lower function. The basic shape of the graph can be obtained easily with a graphing calculator so use that if you have one. If you don't online tools are a good resource. Once done the upper and lower functions in that interval should be determined on the graph. *****Correction****** For some silly reason I had sec^2 as 1/sin^2 in my calculator. 1/cos^2 gives the correct values and eliminates the gap in the middle I was talking about. Continuing you should find where the graphs intersect and integrate from those x bounds. Specifically addressing the integration of sec^2x as that might be a sticking point, one should look at the various trig functions and their derivatives. Tan specifically.

And my own question as a follow up for the thread.

"Find the general solution for the given differential equation"

That equation being y''+y'=tan(t)

The overall method of the chapter is variation of parameters with the relevant theorem
eq0027MP.gif


I have already solved the characteristic equation as y=c1cos(t)+c2sin(t) which gives a wronskian of cos^2t+sin^2t=1. Yay easy numbers! But I eventually end up, using that provided theorem, with an integral of sin(t)tan(t) and no idea how to proceed.

Edit: Dang, figured it out. sint*tant=sin^2t/cost=(1-cos^2t)/cos(t) and so on. Just needed to type it out.
 
Having trouble reading what this question is giving me for information.

HF0y0MK.png


From how I'm reading it, Sn multiples x to make up each column in the matrix A, with n changing per column. I don't know what i and j are supposed to be in the delta function though. I assumed it was row and column but trying out a random example I'm not getting a circulant matrix so I'm sort of stuck as to what the question is even saying.
 
Having trouble reading what this question is giving me for information.

HF0y0MK.png


From how I'm reading it, Sn multiples x to make up each column in the matrix A, with n changing per column. I don't know what i and j are supposed to be in the delta function though. I assumed it was row and column but trying out a random example I'm not getting a circulant matrix so I'm sort of stuck as to what the question is even saying.

i and j are indeed row and column indices. The bottom line specifies the (i,j) entry of the matrix of the linear transformation S^n with respect to the standard basis. All the construction is trying to do is circularly permute the entries of the column vector x by successively greater amounts as you move from left to right. It should be clear why that would yield a circulant matrix.

Example (does not include the action on x, just showing the matrix of S):

Code:
>> sfun = @(i,j,n,N) mod(i-j-n,N) == 0;
>> [J,I] = meshgrid(0:4);

>> sfun(I,J,0,5)

ans =

     1     0     0     0     0
     0     1     0     0     0
     0     0     1     0     0
     0     0     0     1     0
     0     0     0     0     1

>> sfun(I,J,2,5)

ans =

     0     0     0     1     0
     0     0     0     0     1
     1     0     0     0     0
     0     1     0     0     0
     0     0     1     0     0

EDIT2: Whoops, had the order of output arguments to meshgrid switched. The fixed version gives I as the matrix of row indices and J as the matrix of column indices as originally intended:

Code:
>> [I J]

ans =

     0     0     0     0     0     0     1     2     3     4
     1     1     1     1     1     0     1     2     3     4
     2     2     2     2     2     0     1     2     3     4
     3     3     3     3     3     0     1     2     3     4
     4     4     4     4     4     0     1     2     3     4
 
Thank you. I understand what's going on now and made a function that outputs the circulant matrix. For some reason I didn't realize that Sn was a matrix which was confusing me. With that though I see how we get the circulant matrix out of it.

edit: Alright I have another question if someone could help.

O52He4S.png


x = [-1 2 -3 4]^T

I have the eigenvalues magnitudes of B, but theres multiplicity due to conjugation so I'm not sure if I include that. IE

[0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14]

Would 5 be the smallest or would 11?
 
Thank you. I understand what's going on now and made a function that outputs the circulant matrix. For some reason I didn't realize that Sn was a matrix which was confusing me. With that though I see how we get the circulant matrix out of it.

edit: Alright I have another question if someone could help.

O52He4S.png


x = [-1 2 -3 4]^T

I have the eigenvalues magnitudes of B, but theres multiplicity due to conjugation so I'm not sure if I include that. IE

[0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14]

Would 5 be the smallest or would 12?


It is the 12th smallest magnitude. You need to count from the back.

The 12 smallest eigenvalue magnitude is 3
 
It is the 12th smallest magnitude. You need to count from the back.

The 12 smallest eigenvalue magnitude is 3

Are you sure? Maybe I'm thinking about this wrong. wouldnt 0 be the first smallest value, 1 the second smallest, ect?

So I've got everything for this problem set up(found my eiegenvectors with 12th smallest eigenvalue). I'm now confused by how I'm plotting my real/images of eigenvectors. It says to plot against time and I'm not quite sure what that means when it comes to vectors.

edit: Figured it out.
 
Alright, just learned sequences and series in my calculus class and just came across these on the homework:

cPLUQn2.png


There was nothing the professor went over in class that I can think to apply to this, so any tips on where to start?
 
Alright, just learned sequences and series in my calculus class and just came across these on the homework:

cPLUQn2.png


There was nothing the professor went over in class that I can think to apply to this, so any tips on where to start?

You should divide the factorials and see what you get first. One of the factorials has more terms, and many of the terms cancel out.
 
You should divide the factorials and see what you get first. One of the factorials has more terms, and many of the terms cancel out.

Yeah, just looked up how to divide factorials with variables (which I had no knowledge of before lol), and that makes these problems infinitely easier. Thanks for the help!
 
Does anyone know of any good resources for learning transcendental functions? Mainly general inverses, exponentials and logarithms, trig inverses, and the derivatives of all these? First year calculus is hard when the professor doesn't explain very well (and makes a ton of mistakes), and then not being able to read the textbook for extra help as it is written by the professor adds to the frustration.
 
Does anyone know of any good resources for learning transcendental functions? Mainly general inverses, exponentials and logarithms, trig inverses, and the derivatives of all these? First year calculus is hard when the professor doesn't explain very well (and makes a ton of mistakes), and then not being able to read the textbook for extra help as it is written by the professor adds to the frustration.

Khan academy and Paul's online math notes.

Just memorize the derivative rules for those functions, and if you are a STEM like me, those rules will stick in your brain for life.

i got a 97 on my calculus II midterm!!!!!!

Damn. Congrats sir.
 
Khan academy and Paul's online math notes.

Just memorize the derivative rules for those functions, and if you are a STEM like me, those rules will stick in your brain for life.

Paul's Notes seems pretty good from what I have read. He does all this for the general population, and his students, and my professor won't even respond to emails.

When I checked Khan Academy for general inverse stuff, not derivatives, I didn't think it went in-depth enough. I went searching for extra help when I got to this question in my homework:

f is a one-to-one function with inverse f^-1. Calculate the inverses of the given functions in terms of f^-1.

a) g(x) = f(x) - 2

I eventually figured out how to do these types of problems by asking my not so helpful TA and checking the solutions manual, but I don't think Khan Academy helped me here.

And yes, I am the E part of STEM.

Thanks for the help!
 
I'm giving a 6-minute talk at a Toastmasters meeting in a couple of weeks. The primary purpose of my talk is to use visual aids effectively, so I'm thinking of talking about infinity, a concept that is hard for all of us to grasp without pictures.

In particular, I think it will be interesting to talk about paradoxes involving infinity. Some examples that I know of are the pairing function (i.e. N^2 versus N), Gabriel's horn (Koch snowflake in 1D-2D), and Cantor's disappearing table.


Do you guys know of other examples involving infinity (paradoxical or not) that are easy to explain using pictures? I don't like the Achilles-tortoise example, as I can't see how to make an interesting picture out of it.

Also, if you are familiar with the history, can you tell me how people in the olden days didn't believe in infinity? A quick search returned me the name of Kronecker. Of course, that 0-1 guy. : )
 
0.999... = 1

not sure exactly how you'd visualize it, but maybe something like proceeding with the naive intuition that there exists some number "0.999...5" that is between "0.999..." and 1, and then sort of demonstrating that the definition of the ellipsis as infinitely repeating precludes this number from existing.
 
0.999... = 1

not sure exactly how you'd visualize it, but maybe something like proceeding with the naive intuition that there exists some number "0.999...5" that is between "0.999..." and 1, and then sort of demonstrating that the definition of the ellipsis as infinitely repeating precludes this number from existing.

Thanks for the suggestion, Stumpokapow. I'll think about whether the proof can be made interesting using pictures. I think, for that problem, it's easier to convince people using algebra though. I don't know, maybe it's hard to explain that 9.9... - 0.9... = 9.
 
I'm giving a 6-minute talk at a Toastmasters meeting in a couple of weeks. The primary purpose of my talk is to use visual aids effectively, so I'm thinking of talking about infinity, a concept that is hard for all of us to grasp without pictures.

In particular, I think it will be interesting to talk about paradoxes involving infinity. Some examples that I know of are the pairing function (i.e. N^2 versus N), Gabriel's horn (Koch snowflake in 1D-2D), and Cantor's disappearing table.


Do you guys know of other examples involving infinity (paradoxical or not) that are easy to explain using pictures? I don't like the Achilles-tortoise example, as I can't see how to make an interesting picture out of it.

Also, if you are familiar with the history, can you tell me how people in the olden days didn't believe in infinity? A quick search returned me the name of Kronecker. Of course, that 0-1 guy. : )

Lots of ways to visualize the golden ratio. Can find probably 10 with a quick google search
 
Anyone good with homogenous linear recursives? I understand the distinct formulas used if you have either 1 distinct root or 2 roots but it's a little messy when you get into it.
 
Math gaf, I have a homework question.

I'm asked to prove that the LHS <= RHS.

Base case: LHS = RHS

Induction step: I proved LHS < RHS.

Then can I conclude LHS <= RHS for all n that belongs to the well-ordered set S?
 
Math gaf, I have a homework question.

I'm asked to prove that the LHS <= RHS.

Base case: LHS = RHS

Induction step: I proved LHS < RHS.

Then can I conclude LHS <= RHS for all n that belongs to the well-ordered set S?

Could you be more specific, Kieli? It has been a while since I had to work with well-orderedness, heh. What are the set S and the expressions on the LHS and the RHS?
 
I have three questions which I hope you guys can help me out on

Utility function is U(m) = -115 + 22.21*log(m+150000) where m is the amount.

Question 1) I can buy a ticket to a lottery for $1 and the payout is $50 million. What is the probability of winning at which it makes sense to play?

So far I have this:

ExpectedUtility(Lottery) = (Probability of Winning)(Utility Of Winning) + (Probability of Losing)(Utility Of Losing)

(Utility Of Winning) = U($50 Million) + U(- $1) = 56.0230 - .0391 = 55.9839
(Utility Of Losing) = U(- $1) = -.0391

P(55.9839 + (-.0391)) = 0
P(55.9448) = 0

Could anyone explain to me how I can get the probability given that I know the utility value? Or am I going about the problem the wrong way?

Question 2) Probability is 1/52,000,000. How big does the lottery pot have to be to make sense to play?

1/52,000,000(U(x) + (.0391)) = 0

How can I get the value of x?

Question 3) I can pay $149,999 to play a lottery with a 85% of winning $100 Million. Should I play?

U(-$149,999) = -115
U($100 Million) = 62.6945

EU(Lottery) = (.85)(62.6945) + (.15)(-115)
EU(Lottery) = 53.290325 + (-17.25) = 36.040325

Since expected utility is greater than 0, I should play. Is this correct?

I appreciate any help that you guys can give.
 
ExpectedUtility(Lottery) = (Probability of Winning)(Utility Of Winning) + (Probability of Losing)(Utility Of Losing)

(Utility Of Winning) = U($50 Million) + U(- $1) = 56.0230 - .0391 = 55.9839
(Utility Of Losing) = U(- $1) = -.0391

P(55.9839 + (-.0391)) = 0
P(55.9448) = 0

The math you did here isn't the math you set up in the first line. What are you doing with probabilities? From where are you deriving 0?

EU = P(Win)U(Win) + (1-P(Win))U(Lose)

U(Win) = -115 + 22.21*log(-1+50000000+150000) = 56.023
U(Lose) = -115 + 22.21*log(-1+150000) = -0.039
EU = P(Win) (56.023) + (1-P(Win))(-0.039)
EU = 56.023P(Win) - 0.039 + 0.039P(Win)
EU = 56.062P(Win) - 0.039

The breakeven point occurs when EU=0 -- the function is obviously linearly increasing and the border cases are clearly negative and positive respectively so the IVT allows us to be confident that this breakeven point is unique:
0 = 56.062P(Win) - 0.039
0.039 = 56.062P(Win)
P(Win) = 0.039/56.062

Question 2) Probability is 1/52,000,000. How big does the lottery pot have to be to make sense to play?

1/52,000,000(U(x) + (.0391)) = 0

How can I get the value of x?

Set EU to 0; calculate the utility of losing and probability of losing terms; move them to the other side of the equation; multiply the result by the probability of winning; sub in the formula for the utility of winning for U(Win) in the equation; solve algebraically.
 
The math you did here isn't the math you set up in the first line. What are you doing with probabilities? From where are you deriving 0?

EU = P(Win)U(Win) + (1-P(Win))U(Lose)

U(Win) = -115 + 22.21*log(-1+50000000+150000) = 56.023
U(Lose) = -115 + 22.21*log(-1+150000) = -0.039
EU = P(Win) (56.023) + (1-P(Win))(-0.039)
EU = 56.023P(Win) - 0.039 + 0.039P(Win)
EU = 56.062P(Win) - 0.039

The breakeven point occurs when EU=0 -- the function is obviously linearly increasing and the border cases are clearly negative and positive respectively so the IVT allows us to be confident that this breakeven point is unique:
0 = 56.062P(Win) - 0.039
0.039 = 56.062P(Win)
P(Win) = 0.039/56.062



Set EU to 0; calculate the utility of losing and probability of losing terms; move them to the other side of the equation; multiply the result by the probability of winning; sub in the formula for the utility of winning for U(Win) in the equation; solve algebraically.

Thank you.
 
Last problem on my homework has me completely me stumped. I need to determine whether this sum converges or not:

g2awD2Z.png


Also, does anyone have a general progression of what laws/tests they use in which order when trying to solve series problems? I feel like there's a better way than what I'm doing which is just randomly trying laws until one works.
 
Last problem on my homework has me completely me stumped. I need to determine whether this sum converges or not:

g2awD2Z.png


Also, does anyone have a general progression of what laws/tests they use in which order when trying to solve series problems? I feel like there's a better way than what I'm doing which is just randomly trying laws until one works.


Maybe there's an easier way than what I show below.


I noted that every other term is equal to 0, since cos(pi/2), cos(3 * pi/2), ... = 0.

Hence, we can re-write the sum as,

\sum_{n = 0}^{\infty} 3 / log(n + 4) * cos(n * pi/2)

= \sum_{m = 0}^{\infty} 3 / log(2*m + 4) * cos(m * pi)

= \sum_{m = 0}^{\infty} 3 / log(2*m + 4) * (-1)^m.


(Sorry, my reasoning after this was incorrect. I'll let you work out the rest.)
 
Last problem on my homework has me completely me stumped. I need to determine whether this sum converges or not:

g2awD2Z.png


Also, does anyone have a general progression of what laws/tests they use in which order when trying to solve series problems? I feel like there's a better way than what I'm doing which is just randomly trying laws until one works.
It's been like a year since i've taken calc 2, but the very first test you should do is the divergence test. If the function goes to 0 then it may not diverge, if it doesn't go to 0 then it diverges and you stop there. If you're having trouble knowing what test to use next then i'd use this guy's videos: Professor Leonard. Best online calc 2 teacher i know of, he helped me get an A in calc 2 when i took it
From what i remember the progression was:
Divergence test
p-series
geometric
limit comparsion
ratio test(best test but takes a bit)
Then your other tests for more obvious cases(power series, alternating series, etc)
 
Last problem on my homework has me completely me stumped. I need to determine whether this sum converges or not:

g2awD2Z.png


Also, does anyone have a general progression of what laws/tests they use in which order when trying to solve series problems? I feel like there's a better way than what I'm doing which is just randomly trying laws until one works.

It is a convergent series.

Simply apply the alternating series test.

In an alternating series (the sign changes with each term), if the absolute value of each term monotonically decreases, then the series converges. This is the alternating series test.

It is easy to check that this is true for the series you want.
 
Status
Not open for further replies.
Top Bottom