how do you find the equation of a tangent line through a point outside a circle.
Suppose the point is (6,-5) and the circle is
(x-1)^2 + (y^2) = 4 , but i want to find a general formula.
we know that two tangents can always be drawn to a circle from any point outside the circle

- anonymous

- schrodinger

I got my questions answered at brainly.com in under 10 minutes. Go to brainly.com now for free help!

At vero eos et accusamus et iusto odio dignissimos ducimus qui blanditiis praesentium voluptatum deleniti atque corrupti quos dolores et quas molestias excepturi sint occaecati cupiditate non provident, similique sunt in culpa qui officia deserunt mollitia animi, id est laborum et dolorum fuga.
Et harum quidem rerum facilis est et expedita distinctio. Nam libero tempore, cum soluta nobis est eligendi optio cumque nihil impedit quo minus id quod maxime placeat facere possimus, omnis voluptas assumenda est, omnis dolor repellendus.
Itaque earum rerum hic tenetur a sapiente delectus, ut aut reiciendis voluptatibus maiores alias consequatur aut perferendis doloribus asperiores repellat.

Get this expert

answer on brainly

SEE EXPERT ANSWER

Get your **free** account and access **expert** answers to this

and **thousands** of other questions

- amistre64

so like this?

##### 1 Attachment

- anonymous

YESSSSSSSSS!!!!!!!!!!!!!!

- anonymous

in an xy plane

Looking for something else?

Not the answer you are looking for? Search for more explanations.

## More answers

- amistre64

since a tangent to the circle is always at 90 degrees to the radius.... this makes a right triangle.

- anonymous

hey how do you draw attachments like that

- amistre64

the hyp is the length of the center of the cirlce to the point and one leg is the radius

- amistre64

i use paint in the windows accessories...save it to my desktop and retrieve it from there

- anonymous

you can draw circles?

- amistre64

##### 1 Attachment

- amistre64

yes; its the oval selection and hold down the shift key while you drag

- anonymous

yes
what program is it? windows paint?

- anonymous

ok so we have a point, lets call it (x0,y0) outside a circle

- amistre64

yes; windows generi paint program that comes in every installation of windows since 3.xx

- anonymous

ok i will check that

- anonymous

i want to actually find a general formula here

- anonymous

ok so we have (x0,y0) outside a circle (x-h)^2 + (y-k)^2 = r^2

- amistre64

the general formula is the pythag thrm... to find a vector

- amistre64

find a vector that is parallel to the tangent from the point....

- anonymous

you cant do this analytically

- amistre64

you can ; but then convert it :)

- anonymous

we have y - y0 = m(x-x0) ,

- amistre64

the slope of the tangent to the circle at any given point is f'(x)

- anonymous

also the line from the center of circle to (x,y) is y-k = -1/m (x-h)

- anonymous

the normal i mean

- anonymous

since (x,y) is unknown

- amistre64

given a standard point on the circle for say (Xc,Yc)...

- anonymous

oh then i messed up

- amistre64

the point has to be within a certain domain for an outside point to have a tangent right?

- anonymous

thats possible

- anonymous

there are 3 points given

- anonymous

(h,k) the center of the circle, (xo,yo) point outside circle
(x1,y1) the point on the circle and (x2,y2) the second point on circle that intersects tangent

- anonymous

so 4 points

- amistre64

f'(c) that passes thru P(x1,y1) in the field and P1(xc,yc) on the circle

- amistre64

if we view this from the origin...would it be simpler?

- amistre64

the angle formed by the center to outside point forms an angle [a] with the x axis; the angle between that line and the radius to tangent form [b] right

- anonymous

yes

- amistre64

like this so far

##### 1 Attachment

- amistre64

a' is in the wrong spot lol

- anonymous

so you can do this vectorially? or with derivatives?

- amistre64

you can, but lets get a clear picture first and then add the details :)

- amistre64

the perp lines slope we want is tan(a+b) right?

- anonymous

tan a + b ?

- amistre64

a = cos^-1(CP/r)

- amistre64

yes; from my picture I got a + b = the slope of the line that is perp to the tangent....

- amistre64

and tan(a+b) givs us that slope; so figure the angles... :)

- amistre64

does this makes sense?

- anonymous

no

- amistre64

we center the circle at the origin so we can examine it...

- anonymous

ok

- anonymous

i dont understand the a = arcos (CP/r) business, where does that come from

- amistre64

the angle made by the line from the center to the extPoint makes an angle [a] with the x axis...that angle is cos^-1[]

- anonymous

ok , the angle is ?

- amistre64

the angle is cos(a) = P.r/(|P||r|)...dot product of 2 vecotrs

- anonymous

cos a = Ca' / ( CP)

- amistre64

P being the outside point

- anonymous

what is r ?

- amistre64

cos(b) = T.P/(|P||r|)

- amistre64

radius = r.....

- anonymous

cos b =r / TP

- amistre64

to find the tangent..... we use the pythag thrm;
|T|=sqrt(|P|^2-r^2)

- anonymous

ok but we want the tangent line equation

- anonymous

|T| is infinity

- anonymous

but that line segment, right comes from pythangrean

- anonymous

and you were doing the inner product, im a bit rusty

- amistre64

back...

- anonymous

hey

- amistre64

so if we know tha toutside point ot not; we can assign a vector to it ..p we can also assign a vector to our radius at any given point [c]... r = .
when f'(c) is parallel with r-p..... we have our tangent line

- amistre64

the equation of a tangent line thru any point is given by:
x = x0 + at
y = y0 + bt
z = 0....since we are in the xy plane

- anonymous

oh i think i have it

- anonymous

if youre given a circle and a point outside a circle

- anonymous

nevermind

- amistre64

...... but i gotta mind lol

- amistre64

if we are examine it from the origin; like all good analysts; we move the point and the circle by -x and -y..... make sure to write that in a note so we know were to put it when were done...

- anonymous

i guess im confused on a lot of things. where did you get tan (a+b)

- anonymous

where did you get arcos (cP/r)

- amistre64

i was rummaging thru the thoughts in my head.... some were good, others not so good :)

- anonymous

where did point c come from ?

- anonymous

ok youre starting over

- amistre64

c = center; ..... or origin depending on where the circle was

- amistre64

since we dont know the point ot tangency, i figure vectors are a safe way to find it

- anonymous

yeah thats the trouble

- amistre64

we know the origin; we have vectors splayed out from there. the vector for the radius is simply where c is the point of tangency right?

- amistre64

p is the vector from the origin to the outside point..

- amistre64

we know that if we subtract p from r we get a new vector heading inthe direction of tangency...

- amistre64

it has to match f'(c) in order to determine if 'c' is a match..

- anonymous

one sec

- amistre64

when we find the right [c] we have the f'(c) as the slope for our line; and we have a point to place in it to figure out the b of y=f'(c)x + b

- anonymous

ok vector R - P has the same direction as the tangent line , i agree

- anonymous

because P + (R-P) = R , where R is that point (c,f(c)) or the vector to the point

- anonymous

so f ' (c) will , you mean the vector c ? or just

- amistre64

the derivative of the function itself.

- anonymous

so f ' ( < c , f(c) ) >

- anonymous

oh , the derivative of the equation of the circle

- amistre64

the circle is what i had in mind; but the vector might work just as good

- anonymous

2x + 2y * y' = 0

- anonymous

i prefer analytic, my vector skills are shaky :)

- anonymous

so y ' = -x/y

- amistre64

vectors are analytic; just useful when drawn as arrows lol

- anonymous

true

- anonymous

ok i prefer baby algebra ?

- amistre64

lets restrict our selves to half the circle..... for simplicity
y = sqrt(r^2 - x^2)

- anonymous

its confusing, because the points , like an n tuple, can be represented as a geometric point, or an arrow from the origin,

- anonymous

ok

- anonymous

so f ' (c) = -c / ( sqrt ( r^2 - c^2 )

- anonymous

you must be good at geometry, im better with algebra/symbols, that sort of thing

- anonymous

so we have x = tv + x0 or something

- anonymous

x = x0 + at , y = y0 + at , but then we will have to remove the t

- amistre64

the derivative of u^(1/2) = 1/[2u^(1/2)]
u = 1-x^2 ; du = -2x
the derivative of sqrt(r^2 - x^2)...so i agree with the derivative :)

- amistre64

the t is just a scalar..

- amistre64

it can be 1 for all it cares

- anonymous

well i want the final solution to look like y = mx + b

- anonymous

the problem is that (c,f(c) is unknown

- amistre64

c has to meet the criteria set forth.... once it meets it; we know it :)

- radar

will it meet it today?

- amistre64

lol...depends on how it behaves lol

- amistre64

if anything ; f'(c) is the slope of the line passing thru points (c,f(c)) and (x,y)

- anonymous

ok

- amistre64

y = f'(c)x + B

- anonymous

ok

- amistre64

ive been playing some numbers here... see if this follows...
p =<6,5> ; r = 2
R=
f'(c) = -1/sqrt(2^2 - c^2)
R-p = = (c-6,-3)

- amistre64

-c -3
--------- = ----
sqrt(4-c^2) c-6

- amistre64

cross multiply...
-c(c-6) = -3(sqrt(4-c^2)) ; /-3
-c^2 +6c
-------- = sqrt(4-c^2 ; ^2 both sides..
-3

- amistre64

c^4 -12c^3 +36c^2
------------------ = 4-c^2
9

- amistre64

*9 both sides
c^4 -12c^3 +36c^2 = 36 - 9c^2

- amistre64

zero out the rhs...
c^4 -12c^3 +9c^2 + 36c^2 -36 = 0
c^4 -12c^3 + 45c^2 -36 = 0

- anonymous

how did you get this
-c -3 -3
--------- =-----------
sqrt(4-c^2) c-6

- anonymous

there are some algebraic approaches on yahoo answers, but im not sure they are correct

- anonymous

http://answers.yahoo.com/question/index?qid=20090721083008AA1VEba

- amistre64

the derivative at c = -c/sqrt(r^2-c^2) right?

- amistre64

at r = 2....

- amistre64

...... i see a hiccup tho; i used (c,2) instead of (c,f(c)).... gonna have to rewrite it al lol

- amistre64

-c sqrt(4-c^2)-5
---------- = -------------
sqrt(4-c^2) c-6

- amistre64

-c(c-6) = sqrt(4-c^2)[sqrt(4-c^2)-5]
-c^2 +6c = (4-c^2) -5sqrt(4-c^2)
6c -4
------ = sqrt(4-c^2)
-5

- amistre64

now ^2 both sides :)
6c^2 -48c +16 = 25(4-c^2)
6c^2 -48c +16 = 100 -25c^2

- amistre64

36c^2....

- amistre64

36c^2 +25c^2 -48c +16 -100 = 0
61c^2 -48c -84 = 0

- amistre64

solve for c

- anonymous

hmmm

- anonymous

well this does seem easier than the yahoo method , its insanely difficult

- anonymous

if you want to read it its here
http://answers.yahoo.com/question/index?qid=20090721083008AA1VEba

- anonymous

YOU GET MEDAL for persistence !!!

- amistre64

:) i get from teh wolfram graphing ; that c = -.844 and 1.631

- anonymous

one sec, let me draw this

- anonymous

wait

- anonymous

so you said the direction of R- P where R-P = say is b/a >

- anonymous

i think theres something wrong with that , oh, it only works in 2 dimensions

- amistre64

lets define p as (a,b) thats fine... but we dont want to confuse it for the quadratic....
Ac^2 +Bc + C
(a^2 -b^2) = A
(-2ra) = B
[r^2(1+b^2)] = C

- amistre64

...got some signs wrong lol

- anonymous

so basically you did f ' (c) = slope [ segment RP ]

- anonymous

and we know that f ' (c) from the implicit derivative -x/y

- amistre64

yes

- anonymous

oh well then you didnt really need to use vectors then , i guess

- amistre64

I used vectors to compare it inititally to get a formula for the rest...
(a^2 -b^2) = A
(-2ra) = B
[r^2(1+b^2)] = C

- anonymous

in higher dimensions we will need vectors

- amistre64

vector is used to calibrate the slope

- anonymous

but you took the slope of that vector? correct ?

- amistre64

(a^2 +b^2) = A
(-2ra) = B
[r^2(1-b^2)] = C

- amistre64

correct, the vector x and y form a slope.... so yes

- amistre64

c = thats stuf in the quadraitc formula lol

- anonymous

whats all this
Ac^2 +Bc + C (a^2 -b^2) = A (-2ra) = B [r^2(1+b^2)] = C

- amistre64

i took the numbers out that i was using and placed the (a,b) fromthe exteroir point back into it...

- amistre64

the point of tangency is at c..which is equal to the solution to the quadratic equation formed by ....
(a^2+b^2)c^2 -(2ra)c + [r^2(1-b^2)] = 0

- anonymous

youre making a general solution

- amistre64

yes :)...its no good if it only used for one point

- anonymous

ok start over, where did you A, B< and C

- amistre64

..... when i used the the point (6,5) and a radius of 2... i manipulated the algebra to get a final quadratic form right?

- amistre64

61c^2 -48c -84 = 0
[c] can be solved by teh quadratic formula now since we know a b and c

- amistre64

61 is from a^2 + b^2; 36 +26 = 61

- amistre64

-48 is fron -2ra; -2(2)(6).....gonna have to review tha tone lol
and -84 is from (r^2[1-b^2]): 4(1-25) = 4-100......murmur..

- anonymous

ok

- amistre64

paper is so much easier to work on....
Ac^2 +Bc +C
[a^2+b^2] = A
[-2a(r^2)] = B
[r^2(r^2-b^2)] = C

- amistre64

(a,b)
(6,5) ; r=2
(36+25) = 61 = A
-2(6)(2^2) = -12(4) = -48 = B
4(4-25) = 16 - 100 = -84 = C

- anonymous

ok starting with a point (a,b)

- anonymous

external to the circle, and the point tangent on the circle is (c,f(c))
and the radius is r , center , well we used (0,0)

- amistre64

to find the point [c] of tangency we get:
c = :
ar^2 sqrt[4a^2r^4 - 4(a^2+b^2)(r^2(r^2-b^2))]
------- ( +-) -----------------------------
a^2+b^2 a^2+b^2

- amistre64

if the center is not at the origin; then use transformations to get it there :)

- anonymous

actually this wont work

- anonymous

-c / sqrt ( r^2 -c^2) = ...

- anonymous

the right side could be b - f(c) or f(c) - b

- anonymous

we used (f(c) - b ) / ( c- a)

- anonymous

so you can say, if f(c) > b, then ...

- amistre64

its the same slope regardless of the way its pointing

- anonymous

oh right , duh

- anonymous

so then we get -c / sqrt ( r^2 -c^2) =(f(c) - b ) / ( c- a)

- anonymous

-c / sqrt ( r^2 -c^2) =(srt (r^2 - c^2) - b ) / ( c- a)

- amistre64

the whole point is to get to a point c to find the derivative to use in the equation to the tangent line lol

- anonymous

correct

- anonymous

but i get a simpler derivation

- amistre64

but we have the slope; and we have a point from the vectors...

- anonymous

-c / sqrt ( r^2 -c^2) =(srt (r^2 - c^2) - b ) / ( c- a)
multiply both sides by sqrt ( r^2 - c^2)

- anonymous

-c (c-a) =(r^2 - c^2) - b sqrt (r^2-c^2)

- anonymous

actually i cross multiplied , ok so far?

- amistre64

good so far

- anonymous

sqrt (r^2 - c^2) = r^2/b - ac/b

- amistre64

i think we broke openstudy... the fonts gotten weird on my end

- anonymous

you can start over, use www.openstudy.com , and click math

- anonymous

r^2 - c ^2 = ( r^2 - ac)^2 / b^2

- anonymous

-c (c-a) =(r^2 - c^2) - b sqrt (r^2-c^2)
sqrt (r^2 - c^2) =( r^2 - ac)/b ok so far?

- anonymous

oh i subtracted c^2 from both sides by the way

- amistre64

so did i :)

- anonymous

ok squaring both sides i get
r^2*b^2 - b^2 c^2 = r^4 -2r^2ac + a^2 *c^2

- anonymous

now r is given, b and a are given

- anonymous

i didnt get A = a^2 +b ^2 though

- anonymous

still there?

Looking for something else?

Not the answer you are looking for? Search for more explanations.