RuntimeWarning: overflow encountered in double_scalars

Question

I get the error RuntimeWarning: overflow encountered in double_scalars and I am not sure what it means (internet research did not help) and what double_scalars are.

I am not dividing by zero, so could it be that the warning/error arrises as I have too many decimal points or because the numbers are to large? I do get inf for a lot of them.

Is there a way I can get around this error?

This part ist just importing and defining variable:


## IMPORT PACKAGES
import numpy as np
import math
import matplotlib.pyplot as plt


## DEFINITION AND VALUE OF PARAMETERS
A = 10**-16       
rho = 910                              
g = 9.81                                           
a = 1                               
n = 3
fact = A * ((rho * g)**n) 
t_end = 10 
dt = 1 
dx = 100 
L = 300000
b = 0
H_0 = 1000

This part should be ok


## Solve equation numerically

# initialization
xint = np.arange(-L, L+dx, dx) 
H_old = []

for xind in range(len(xint)):
    x = xint[xind]
    H_old.append(H_0 * (1 - abs(x)/L))
H_new = H_old   
result = [H_old]
print(len(xint),len(H_old))

I think that this part causes troubles

# Loops through the time-span
for t in range(0, t_end+dt, dt):
    print(t,min(H_old),max(H_old)) 

    # Loops through the x-range
    for xind in range(1,len(xint)):
        x = xint[xind]
        if x < L and x > -L:
            dH1 = (H_old[xind]-H_old[xind-1])/dx
            H1 = H_old[xind]
            dH2 = (H_old[xind+1]-H_old[xind])/dx
            H2 = H_old[xind+1]

            v1 = - (2/4)/(n+1) * fact * ((abs(dH1))**(n-1)) * dH1 * H1**4 
            v2 = - (2/4)/(n+1) * fact * ((abs(dH2))**(n-1)) * dH2 * H2**4
            dH_dt = - (H2*v2-H1*v1)/dx + a 
            H_new[xind] = H_old[xind] + dt * dH_dt 

        else:                           
            H_new[xind] = 0.0

    H_old = H_new   
    result.append(H_old)




Thanks

Visit this question - https://stackoverflow.com/questions/43405777/what-are-the-causes-of-overflow-encountered-in-double-scalars-besides-division-b/43406360 and you should post the code that causes the problem so others can help you effectively — Prathamesh, Jun 09 '20 at 13:36
apparently some of the variables were approximated to either `inf` or `nan` — Anwarvic, Jun 09 '20 at 16:59
@Anwarvic I thought that that might have happened. Is there a way around it, without changing the numbers (there are part of a model so should not be changed)? — random__human, Jun 10 '20 at 06:12

RuntimeWarning: overflow encountered in double_scalars

0 Answers0