MHB Finding Restrictions for the Implicit Function Theorem

Click For Summary
The discussion revolves around determining the conditions on constants \(a\) and \(b\) for the implicit function theorem to apply to a system of equations involving \(u\) and \(v\). It is established that the determinant of a specific matrix must be non-zero, leading to the condition \(ab \neq 0\) or equivalently \(a^2 + b^2 \neq 0\). Participants explore the implications of setting \(u\) and \(v\) to zero, questioning if this satisfies the conditions. The conversation also emphasizes the need for continuous partial derivatives and suggests substituting specific values to derive necessary restrictions for \(a\) and \(b\). Ultimately, the focus is on ensuring that the derivatives are well-defined and continuous around the point \((1,0)\).
evinda
Gold Member
MHB
Messages
3,741
Reaction score
0
Hello! (Wave)

Which relation do the constants $a,b$ have to satisfy so that the implicit function theorem implies that the system of two equations

$$axu^2v+byv^2=-a \ \ \ \ bxyu-auv^2=-a$$

can be solved as for u and v as functions $u=u(x,y)$ and $v=v(x,y)$ with continuous partial derivatives of first order in some region of $(1,0)$ and with u(1,0)=1, v(1,0)=-1.

I have thought the following:$$\Delta=\begin{pmatrix}
au^2v & bv^2\\
byu & -bxu
\end{pmatrix}=\begin{pmatrix}
-a & b\\
0 & -b
\end{pmatrix}$$It should hold that $det(\Delta)\neq0 \Rightarrow ab \neq 0$.

So the condition is $a^2+b^2 \neq 0$. Am I right?
 
Physics news on Phys.org
evinda said:
Hello! (Wave)

Which relation do the constants $a,b$ have to satisfy so that the implicit function theorem implies that the system of two equations

$$axu^2v+byv^2=-a \ \ \ \ bxyu-auv^2=-a$$

can be solved as for u and v as functions $u=u(x,y)$ and $v=v(x,y)$ with continuous partial derivatives of first order in some region of $(1,0)$ and with u(1,0)=1, v(1,0)=-1.

I have thought the following:$$\Delta=\begin{pmatrix}
au^2v & bv^2\\
byu & -bxu
\end{pmatrix}=\begin{pmatrix}
-a & b\\
0 & -b
\end{pmatrix}$$It should hold that $det(\Delta)\neq0 \Rightarrow ab \neq 0$.

So the condition is $a^2+b^2 \neq 0$. Am I right?

Hey evinda! (Smile)

Suppose we pick $u(x,y)=v(x,y)=a=0$, wouldn't that satisfy all conditions? (Wondering)
 
I like Serena said:
Hey evinda! (Smile)

Suppose we pick $u(x,y)=v(x,y)=a=0$, wouldn't that satisfy all conditions? (Wondering)

Yes, that's right... Could you give me a hint how can we can find the desired condition? (Thinking)
 
evinda said:
Yes, that's right... Could you give me a hint how can we can find the desired condition? (Thinking)

Shouldn't we have partial derivative like:
$$
\pd{}x(axu^2v+byv^2)=au^2v + 2axuvu_x + axu^2v_x + 2buvv_x = 0
$$
and 3 more such equations?

We should be able to solve for $(u_x,u_y)$ and $(v_x,v_y)$, after which we need to check if each of them is a well defined , continuous vector function around $(x,y)=(1,0)$.
That means in particular that if we substitute $(x,y)=(1,0)$, $u(1,0)=1$, and $v(1,0)=-1$, that we should find that $(u_x,u_y)$ and $(v_x,v_y)$ are defined (not infinite).

We can probably make the substitution immediately, before solving for $(u_x,u_y)$ and $(v_x,v_y)$.
That should give us the required restrictions for a and b.
That is:
$$
\pd{}x(axu^2v+byv^2)\Big|_{(1,0)}=a\cdot -1 + 2a\cdot -1 \cdot u_x + av_x + 2b\cdot -1 \cdot v_x
= -a -2au_x+(a-2b)v_x = 0
$$
and the same for the other 3 equations.
Then solve for $(u_x,u_y)$ and $(v_x,v_y)$. (Thinking)
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
Replies
1
Views
1K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 16 ·
Replies
16
Views
2K
Replies
10
Views
2K