This is used to calculate limits of the form limx→cg(x)f(−x), where c is an extended real number, ie. c is either a real number or ±∞.
Special case: Suppose f(x) and g(x) are continuously differentiable functions in some interval I containing C.
Also, assume that f(c)=g(c)=0.
And, g′(c)=0.
Then,
g(x)f(x)=g(x)−g(c)f(x)−f(c)(∵f(c)=0=g(c))=x−cg(x)−g(c)x−cf(x)−f(c) if x∈I,x=c.
Now,
limx→cx−cf(x)−f(c)=f′(c)
∴ and x→climx−cg(x)−g(c)=g′(c)=0x→climx−cg(x)−g(c)(x−cf(x)−f(c))=g′(c)f′(c)
∴limx→cg(x)f(x)=g′(c)f′(c)=limx→cg′(x)limx→cf′(x)
because f′(x) & g′(x) are assumed to be continuous at x=c.
Hence, limx→cg(x)f(x)=limx→cg′(x)f′(x)
The above rule, which is valid for more general cases, is known as the L'Hôpital's rule.
L'Hopital's : French mathematician (pronounced is lopital)
Suppose limx→cf(x)=limx→cg(x)=0 or ±∞.
(i.e.limx→cg(x)f(x)is of the form00or∞∞)
Also, assume that limx→cg′(x)f′(x) exists and g′(x)=0∀x in the interval I except possibly at x=c.
Then, limx→cg(x)f(x) exist and
limx→cg(x)f(x)=limx→cg′(x)f′(x).
Solution:
limx→0dxdxdxdsinx=limx→01cosx=1cos0=1.
By the L'Hopital's rule,
limx→0xsinx=1.
Note that here we used the fact that
dxd(sinx)=cosx.
x→0limx2ex−1−x ( 00 form) Solution:=x→0lim2xex−1 (by L’Ho¨pital) =x→0lim2ex (by L’Hopital’s) =2e0=21.
Remark: We might have to apply L'Hópitals rule several times in order to calculate the limit.
limx→∞e−x−e−xex+e−x(∞∞ form)
L′H=limx→∞ex+e−xex−e−x ( ∞∞ form)
L′H=limx→∞ex−e−xex+e−x which is the original limit itself.
So we wont be able to calculate the limit by applying L'Hopital's rule directly.
However, if we put ex=y, the as x→∞,y→∞.
e−x−e−xex+e−x=y−y1y+y1=y2−1y2+1∴x→∞limex−e−xe−x+e−x=y→∞limy2−1y2+1=y→∞lim1−y11+y21=1−01+0=1.
Or, we could use L'Hopital's rule to write
limy→∞y2−1y2+1=L′Hlimy→∞2y2y=1.
x→∞limx−x1x+x1(∞∞ form) =L′Hx→∞lim2x1+21x−3/22x1+2−1x−3/2(00from)x→∞lim−41x−3/2−43x−5/2−41x−3/2+43x−5/2(00 form) )
=... becomes more & more complicated.
However, x→∞limx−x1x+x1=x→∞limx−1x+1=L′Hx→∞lim11=1.
L'Hôpital's Rule can be used for other indeterminate forms like
0.∞,∞−∞, 1∞,0∘, etc. by somehow changing into 00 a ∞∞ form.
Examples: (1) limx→∞x2e−x ( ∞.0 form)
=x→∞limexx2 ( ∞∞ form) =L′Hx→∞limex2x ( ∞∞ form) =L′Hx→∞limex2=0. &
More generally, we can show that
limx→∞xn⋅e−x=0 for any positive integer n
limx→0+xlnx
(here we are taking limx→0+ because lnx is defined only for x>0)
As x→0+,lnx→−∞ y=lnx
Let's write xlnx=x−1lnx ( ∞−∞ form y)
limx→0+xlnx=limx→0+x−1lnx=L′Hlimx→0+−1/x21/x=limx→0+(−x)=0.
limx→0+xx(00 form)
let f(x)=xx. Then lnf(x)=xlnx
We have seen that limx→0+xlnx=0
∴limx→0+lnf(x)=0.
Now, f(x)=elnf(x)
∴limx→∞f(x)=limx→∞elnf(x)=elimx→0+lnf(x)
=e0=1. (∵ex is continuous)
Remark: Note that if limx→cg′(x)f′(x) does not exist, we cannot conclude that limx→c∂(x)f(x) does not exist.
Example: f(x)=x+sinx;g(x)=x
x→∞limf(x)=∞=x→∞limg(x)f′(x)=1+cosx,g′(x)=1
limg′(x)f′(x)=limx→∞(1+cosx), which does not exist.
However,
e. x→∞limg(x)f(x)=x→∞limxx+sinx=x→∞lim[1+xsinx]0≤xsinx≤x1→0 is x→∞.
By the Sandwich the, limx→∞xsinx=0
∴limx→∞g(x)f(x)=1+0=1.