is there a way to plot a function in excel other than just calculating a bunch of points along the function and sticking in a trend line or something?

/noob

10/18/2008 4:13:32 PM

nevermind, i gave up on that so i'm trying to learn matlab

in a unity lab, how do i change matlab's current directory to one where i can save stuff (ie. somewhere on the K: drive)?

10/18/2008 5:02:30 PM

post the function so I know what you are working with

10/18/2008 5:12:52 PM

If all you need to do is plot a function, there are easier ways than matlab.

10/18/2008 5:17:12 PM

ok, i kinda got this figured out. this is my first time using matlab unfortunately.

here's my code:

w=0:.0001:200;
k=4090000
c=22800
m=2000
h=k*((k^2+(c*w)^2)^0.5)/(((k-m*w^2)^2+(c*w)^2)^0.5)
plot(w,h)
title('Frequency Response Magnification Factor')
xlabel('Frequency (rad/s)')
ylabel('Magnification Factor')
grid on

i get this error:

>> run E:\magfac

k =

4090000


c =

22800


m =

2000

??? Error using ==> mpower
Matrix must be square.

Error in ==> magfac at 5
h=k*((k^2+(c*w)^2)^0.5)/(((k-m*w^2)^2+(c*w)^2)^0.5)

Error in ==> run at 57
evalin('caller', [s ';']);

10/18/2008 5:24:09 PM

this code worked fine for the other function:

t=0:.0001:.5;
Y=0.1
k=4090000
c=22800
m=2000
w=101.81
p=atan((m*c*w^3)/(k*(k-m*w^2)+(c*w)^2))
x=(Y*(k^2+(c*w)^2)^0.5)/((k-m*w^2)^2+(c*w)^2)^0.5*cos(w*t-p)
plot(t,x)
title('Steady State Motion of Landing Gear')
xlabel('Time (s)')
ylabel('Vertical Displacement (m)')
grid on


[Edited on October 18, 2008 at 5:34 PM. Reason : ^^i need to learn matlab anyway]

10/18/2008 5:26:40 PM

link

[Edited on October 18, 2008 at 5:58 PM. Reason : ]

10/18/2008 5:34:37 PM

i'd rather just find out why the matlab code isn't working

[Edited on October 18, 2008 at 5:42 PM. Reason : i appreciate your help though]

10/18/2008 5:41:39 PM

I googled the error, and it seems you need to use .^

Quote :

"Technically, you only need dots in front of ^ operators for which the first argument might be a vector, and * and / operators for which the quantities on each side of the operator could both be vectors. The MATLAB function quad gives a vector argument to the function it is integrating, so vectors may crop up when you don't expect them. The safest thing is to put a dot in front of every * /, and ^ operator in all your functions. If you forget the dots in front of operators you may get hard-to-understand error messages. If somewhere you have written a^b instead of a.^b, you may get ??? Error using ==> mpower Matrix must be square. "

- http://www.physics.wustl.edu/~alford/p217/matlab_hints.html

.^ just tells matlab, IIRC, to process corresponding values together regardless of if the set of those values lines up.


like if you have a set of values a=[1,2,3,4] and b=[1,2,3,4,5,6] and use just did a^b it'd throw an error because they don't match. a.^b would just do the first 4 because that's all it has data for.

note: this info is based on very vague recollection of working with matlab



[Edited on October 18, 2008 at 5:51 PM. Reason : obama '08]

10/18/2008 5:45:09 PM

yeah, i found the .^ thing too, but it still wasn't working. lemme try again.

10/18/2008 5:49:13 PM

If you want, you can try your original code with w=0:.5:200 you don't need a .00001 resolution from 0 to 200, that's a LOT of values.

10/18/2008 5:50:40 PM

i didn't even know what the .0001 stood for. someone told me to go with that.

10/18/2008 5:54:49 PM

all i need to do is graph h as a function of w based on the function in line 5

code:

w=0:.5:200;
k=4090000
c=22800
m=2000
h=(k*((k^2+(c*w).^2).^0.5)/(((k-m*w.^2).^2+(c*w).^2).^0.5))
plot(w,h)
title('Frequency Response Magnification Factor')
xlabel('Frequency (rad/s)')
ylabel('Magnification Factor')
grid on

output a graph with constant h and gave me this in the command window:

>> run E:\magfac

k =

4090000


c =

22800


m =

2000


h =

4.8747e+005



that value for h is the value it had on the graph

10/18/2008 5:58:20 PM

Matlab rocks.. I get paid to write matlab code

10/18/2008 5:59:24 PM

^can you help me out? i'd be most grateful.

[Edited on October 18, 2008 at 6:00 PM. Reason : it doesn't seem difficult, this is just my first time ever even seeing it]

10/18/2008 6:00:22 PM

hmm... maybe try this as your h, i noticed you left out some .^ and you want to put them in front of the multiplies too

h=(k.*((k.^2+(c.*w).^2).^0.5)/(((k-m.*w.^2).^2+(c.*w).^2).^0.5))

10/18/2008 6:01:25 PM

i didn't think i should do it for k since it's a constant, not a matrix

10/18/2008 6:03:12 PM

i'm just guessing... your exponents are constants too and it throws an error. It looks like you're trying to use the .5^2 to replace absolute value, maybe you should see what matlab's absolute value function is?

10/18/2008 6:04:49 PM

good point. i'm actually supposed to plot the abs. value of that h value anyway

i'll give it a try

10/18/2008 6:07:46 PM

this returns the same constant value stuff:

w=0:.5:200;
k=4090000
c=22800
m=2000
h=abs(k*(sqrt(k^2+(c*w).^2))/sqrt((k-m*w.^2).^2+(c*w).^2))
plot(w,h)
title('Frequency Response Magnification Factor')
xlabel('Frequency (rad/s)')
ylabel('Magnification Factor')
grid on

as does this:

w=0:.5:200;
k=4090000
c=22800
m=2000
h=abs(k.*(sqrt(k.^2+(c.*w).^2))/sqrt((k-m.*w.^2).^2+(c.*w).^2))
plot(w,h)
title('Frequency Response Magnification Factor')
xlabel('Frequency (rad/s)')
ylabel('Magnification Factor')
grid on

[Edited on October 18, 2008 at 6:13 PM. Reason : .....]

10/18/2008 6:12:01 PM

i'll buy a cookout tray tonight for whoever gets this thing working for me and explains it to me

10/18/2008 6:14:17 PM

hell, i tried this and get that same damn graph:

w=0:.5:200;
k=4090000
c=22800
m=2000
h=abs(k.*(sqrt(k.^2.+(c.*w).^2))/sqrt((k-m.*w.^2).^2.+(c.*w).^2))
plot(w,h)
title('Frequency Response Magnification Factor')
xlabel('Frequency (rad/s)')
ylabel('Magnification Factor')
grid on

10/18/2008 6:23:28 PM

ok, i've done some googling and it's because it's getting two values when it takes the sqrt. i can find out how to make it take only the positive value. if i do abs(sqrt(x)) i think it's still getting two values, they're just the same (ie. two positive values)

[Edited on October 18, 2008 at 6:32 PM. Reason : quintuple post y'all]

10/18/2008 6:31:41 PM

You do realize that you're doing in MATLAB what you didn't want to do in Excel, right?

10/18/2008 6:38:08 PM

google didn't tell me how to do it in excel. i had some code to go off of for matlab, so i figured i'd give it a shot.

they ought to replace that godforsaken fortran requirement with a matlab class so i wouldn't be SOL right now.

[Edited on October 18, 2008 at 6:40 PM. Reason : and the one i got to work was simpler than it would've been in excel]

10/18/2008 6:40:08 PM

you're going to slap yourself when you hear the problem...

you also have to put a period in front of the division symbol too

10/18/2008 6:42:31 PM

when would you like your cookout tray?

10/18/2008 6:44:13 PM

i'll PM you some time next week

Also, you realize that if you have a Mac or get Xwin32 for Windows you can do matlab from the comfort of your home, by remotely logging in to the linux cluster and running matlab from there?

10/18/2008 6:48:34 PM

yes, but my home internet is too slow for the remote stuff

it's free. you get what you pay for.


supposedly it's included in our rent, but i wish they'd lower the rent and make us just get roadrunner or something

[Edited on October 18, 2008 at 6:50 PM. Reason : it's wireless only]

10/18/2008 6:50:08 PM

i got another question. i'm using "atan" to find an angle as a function of frequency. i want it to graph the positive angle that results in that tangent value. here's the code:

w=0:.5:200;
k=4090000
c=22800
m=2000
p=atan((m.*c.*w.^3)./(k*(k-m.*w.^2)+(c.*w).^2))
plot(w,p)
title('Frequency Response Phase Angle')
xlabel('Frequency (rad/s)')
ylabel('Phase Angle (rad)')
grid on

i get a graph with a vertical asymptote where the value of "p" goes negative. pretty much what i want is all the values to the right of that point to have 180 degrees or pi added to them so i get a nice continuous curve. absolute value doesn't help because then it just gives the positive of the angle 180 degrees from the angle i really want. i know this is very unclear, but any ideas?

edit: in other words, "atan" gives a value between -pi/2 and pi/2 and i want values from 0 to pi

[Edited on October 18, 2008 at 7:03 PM. Reason : ya feel???]

10/18/2008 7:00:50 PM

oh shit, nevermind, i figured out how to do it with "atan2"

[Edited on October 18, 2008 at 7:10 PM. Reason : thanks for the help and for putting up with the quintuple posts. this is enough tech talk for me lo]

10/18/2008 7:09:26 PM

would

atan( f(x) - pi ) do what you want too?

[Edited on October 18, 2008 at 7:17 PM. Reason : ]

10/18/2008 7:17:14 PM

maybe abs(atan(x-pi)), but i'm not sure. i don't think that would've worked either.

i did:

vert=(numerator of stuff i had before)
horiz=(denominator of stuff i had before)
angle=atan2(vert,horiz)

"atan2" uses -pi to pi and is quadrant specific. it's useful for working with complex numbers in polar form.

10/18/2008 8:22:54 PM

Quote :

">> run E:\magfac"

i read that as 'macfag'.

10/18/2008 8:31:47 PM

I too, read the error as, "macfag" and thought to myself, "well there's his problem right there!."

10/18/2008 9:15:28 PM

^^^ Dude u got ur code working yet?

[Edited on October 19, 2008 at 12:00 AM. Reason : ^]

10/19/2008 12:00:24 AM

i got it figured out.

i was looking over everything and i realized i messed one of the functions up though. not a matlab problem, a not-knowing-vibrations-very-well problem. should be an easy 2 minute fix though.

10/19/2008 12:06:17 AM