Disable warnings originating from scipy

Question

When I integrate a function with different parameter values, I often encounter errors, for example:

lsoda-- warning..internal t (=r1) and h (=r2) are such that in the machine, t + h = t on the next step (h = step size). solver will continue anyway 

While I understand the technical aspects of the error, I am at a loss with how I can print it from stdout . I looked at numpy.seterr() , but since this error message comes from the base C / Fortran code, it does not work. And although the effectiveness of python script.py > /dev/null effective, it obviously also hides other useful information that can be printed.

Is there a way to disable these warnings and lsoda errors or otherwise prevent them from printing to stdout ?

Answer 1

Using JF Sebastian stdout_redirected , you can redirect stdout at the file descriptor level to os.devnull.

 with stdout_redirected(): soln2 = integrate.odeint(f, y2, t2, mxstep = 5000) 

Outside of with-statement , stdout is still printing. Inside with-suite stdout is suppressed.

For example, here is an example of Caeiro odeint , whose lsoda warnings are suppressed by stdout_redirected .

 import os import sys import contextlib import numpy as np import scipy.integrate as integrate from numpy import pi def fileno(file_or_fd): fd = getattr(file_or_fd, 'fileno', lambda: file_or_fd)() if not isinstance(fd, int): raise ValueError("Expected a file (`.fileno()`) or a file descriptor") return fd @contextlib.contextmanager def stdout_redirected(to=os.devnull, stdout=None): """ /questions/7962/redirect-stdout-to-a-file-in-python/57093#57093 (JF Sebastian) """ if stdout is None: stdout = sys.stdout stdout_fd = fileno(stdout) # copy stdout_fd before it is overwritten #NOTE: `copied` is inheritable on Windows when duplicating a standard stream with os.fdopen(os.dup(stdout_fd), 'wb') as copied: stdout.flush() # flush library buffers that dup2 knows nothing about try: os.dup2(fileno(to), stdout_fd) # $ exec >&to except ValueError: # filename with open(to, 'wb') as to_file: os.dup2(to_file.fileno(), stdout_fd) # $ exec > to try: yield stdout # allow code to be run with the redirected stdout finally: # restore stdout to its previous value #NOTE: dup2 makes stdout_fd inheritable unconditionally stdout.flush() os.dup2(copied.fileno(), stdout_fd) # $ exec >&copied #Constants and parameters alpha=1/137. k=1.e-9 T=40. V= 6.e-6 r = 6.9673e12 u = 1.51856e7 #defining dy/dt's def f(y, t): A, B, C, D, E = y # the model equations f0 = 1.519e21*(-2*k/T*(k - (alpha/pi)*(B+V))*A) f1 = ((3*B**2 + 3*C**2 + 6*B*C + 2*pi**2*B*T + pi**2*T**2)**-1 *(-f0*alpha/(3*pi**3) - 2*r*(B**3 + 3*B*C**2 + pi**2*T**2*B) - u*(D**3 - E**3))) f2 = u*(D**3 - E**3)/(3*C**2) f3 = -u*(D**3 - E**3)/(3*D**2) f4 = u*(D**3 - E**3)/(3*E**2) + r*(B**3 + 3*B*C**2 + pi**2*T**2*B)/(3*E**2) return [f0, f1, f2, f3, f4] t = np.linspace(1e-15, 1e-10, 1000000) # time grid y2 = [2e13, 0, 50, 50, 25] t2 = np.linspace(1.e-10, 1.e-5, 1000000) with stdout_redirected(): soln2 = integrate.odeint(f, y2, t2, mxstep = 5000) 

The warnings that lsoda writes to stdout are suppressed.