The precision of the argument function is less than working precision

Questions
, ,
1

Dear community,
I am trying to numerically integrate a hypergeometric FoxH function. However, I am getting the message:

NIntegrate::precw: The precision of the argument function ((1-Erf[1/(2 Sqrt[s])]) (HypergeometricPFQ[{},{1/2,3/4},-0.390405 s^4]/Gamma[3/4]-1.78387 s HypergeometricPFQ[{},{3/4,5/4},-0.390405 s^4]+1.37869 s^2 HypergeometricPFQ[{},{5/4,3/2},-0.390405 s^4])) is less than WorkingPrecision (150.`)

when I try to plot the result.

How can I resolve this issue? I have attached the example notebook with this post.

Kind Regards,
Ashish

FoxH_func_Nintegrate_Ashish.nb (4.86 KB)

2

What defintion is p[x, s] ? without it nothing can be done.

3

Deleted

4

Dear Mariusz,
I have attached the updated notebook including the p[x,s] term

Open in Wolfram Cloud ↗
5
Open in Wolfram Cloud ↗
6

Dear Mariusz,
Thank you so much for the corrections and changes in the code. I had one small additional question. Can I use the same method for the attached notebook without exponential term?

Open in Wolfram Cloud ↗

Kind Regards,
Ashish

7

Second method is faster.

Open in Wolfram Cloud ↗
8

Dear Mariusz,
Thank you so much :slight_smile: