2
$\begingroup$

How can I calculate the integral- $\int_{S}^{T} \Phi \Big(\frac{l_0-f(t)}{\sigma}\Big)g(t)dt$ where $f(t)= \mu_0+\mu_1 e^{-\gamma (7+logt)^\delta}$ and $g(t)= \frac{1}{(t+h)^k}$ .

Here $\Phi \Big(\frac{l_0-f(t)}{\sigma}\Big)= \frac{1}{\sigma \sqrt{2 \pi} } \int_{-\infty}^{l_0} e^{\frac{(m-f(t))^2}{2 \sigma^2}} dm$; and $l_0,\mu_0,\mu_1,\gamma,\delta,h,k$ are constants.

$\endgroup$

migrated from stats.stackexchange.com Mar 3 at 12:36

This question came from our site for people interested in statistics, machine learning, data analysis, data mining, and data visualization.

  • $\begingroup$ Are the limits of the integral $(-\infty,\infty)$? If so, and $k=1$, then I sense an application of the Cauchy residue theorem. $\endgroup$ – Peter Leopold Mar 1 at 21:59
  • $\begingroup$ @PeterLeopold, No it's a definite integral. $\endgroup$ – gultu Mar 1 at 22:00
  • 1
    $\begingroup$ Welcome to the site. Could you explain how this integral arises? The question has been flagged as off-topic, presumably since the connection to statistics is unclear. $\endgroup$ – Juho Kokkala Mar 2 at 11:57
0
$\begingroup$

Use the integrate() command in R.

I = function(mu0, mu1, L0, gamma, d, h, k, sigma, S, T) 
{
  f = function(t) mu0 + mu1*exp(-gamma*(7 + log(t))^d) 
  g = function(t) 1/( (t+h)^k ) 
  z = function(t) dnorm( (L0-f(t))/sigma ) * g(t) 
  return( integrate(z, S, T)$val ) 
}

I(0,1,0,1,1,0,1,1,1,2)

Make sure the arguments result in defined function values for the integrand (e.g. S$\geq$0).

$\endgroup$
  • $\begingroup$ Thanks but I need analytical result. Not R command. $\endgroup$ – gultu Mar 1 at 21:55
  • $\begingroup$ Good luck with that @gultu. Again, what is the statistical question? $\endgroup$ – beta1_equals_beta2 Mar 1 at 21:57
  • $\begingroup$ @gultu If you need an analytical result, why did you include the numerical-integration tag on your post? $\endgroup$ – Sycorax Mar 1 at 22:08
  • $\begingroup$ @Sycorax, sorry just noticed. Deleting right now $\endgroup$ – gultu Mar 1 at 22:09

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.