I have a differential equation which I think is pretty much unsolvable analytically and numerically, as I do not know $y(c_0)$. $y(c_0)$ is the value of $y(x)$ at $c_0$, however I do not know this value.

$y'(x)=x y(x)^n+\frac{1}{x}(y(c_0)-y(x))^{-q}-\alpha y(x)$

Can I deduce from this equation that $y(x)$ will be decreasing in $\alpha$? If yes, why? Can I even deduce whether $y(x)$ is decreasing or increasing in $x$?

  • $\begingroup$ What does $y(c_0)$ represent, is it related to an initial condition? $\endgroup$ – WalterJ Aug 2 at 9:24
  • $\begingroup$ Basically yes, but unfortunately I do not know the value, but $y'(x)$ depends on $y(c_0)$ $\endgroup$ – Paul Aug 2 at 9:37

If there are other specifications that allow you to show that $y'$ is always positive or negative for all $x$, then you can infer that $y$ Is monotone increasing or decreasing with $x$. But to determine how $y'$ changes with $\alpha$, you'd need to know the magnitude of the two other terms relative to $\alpha y$.


Just for the sake of experimentation, this MATLAB code integrates some version of your equation (note, I took $y(c_0)=y_0$). For this example $y\to 0$, but this is not generally true!

%% ODE example
clear all; clc; close all; 
t0 = 1;
t1 = 100; 
tspan = [t0 t1];
y0 = -100;
[t,y] = ode23s(@(t,y) t*(y^n)+(1/t)*(y0-y)^(-q)-a*y, tspan, y0);
grid on 
  • 1
    $\begingroup$ Thank you Amy and walter $\endgroup$ – Paul Aug 2 at 14:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.