NumPy被定义为np，但仍然得到“NameError:np未定义”

Question

我这里有一个脚本，它调用一个自定义函数并在for循环中运行它。假设函数查找函数的根（此处定义为f(X)），并使用容差值拒绝任何假根。问题是，我一直得到错误：

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-17-10016a66492e> in <module>
     17 start = tictoc()
     18 for i in range(0,n):
---> 19     fsolvetol(f,x0,q,i,tol)
     20 finish = tictoc() - start
     21 print('Elapsed time is {0} seconds'.format(finish))

~\OneDrive - University of Massachusetts Dartmouth\EAS520\Project1\hw3\customsolver.py in fsolvetol(func, guesses, roots, i, Tol)
      7 def fsolvetol(func,guesses,roots,i,Tol):
      8     import numpy as np
----> 9     roots[i] = fsolve(func,guesses[i])
     10 #     if func(roots[i]) > Tol or func(roots[i]) < -Tol:
     11 #         roots[i] = 1

NameError: name 'np' is not defined

尽管在我的两个脚本中都将numpy定义为np。下面是我的主脚本和自定义函数

#!/usr/bin/env python3

import numpy as np
from numpy import pi
from time import perf_counter as tictoc
from customsolver import *
import matplotlib.pyplot as plt

# Organizing Inputs #
def f(x): return np.sin(3*pi*np.cos(2*pi*x)*np.sin(pi*x))
a = -3; b = 5; n = 4**4
x0 = np.linspace(a,b,n)
q = np.zeros(np.shape(x0))
tol = 10**(-13)
#####################

start = tictoc()
for i in range(0,n):
    fsolvetol(f,x0,q,i,tol)
finish = tictoc() - start
print('Elapsed time is {0} seconds'.format(finish))

# Processing Outputs #
q = np.unique(q) # keep roots with unique values only.
q = q[~np.isnan(q)] # removes the nan value

我将调用此处编写的自定义函数：

#!/usr/bin/env python3

from scipy.optimize import fsolve
import numpy as np

# Flags points that fsolve incorrectly assigns as roots within tolerance (now a function!)
def fsolvetol(func,guesses,roots,i,Tol):
    roots[i] = fsolve(func,guesses[i])
    if func(roots[i]) > Tol or func(roots[i]) < -Tol:
        roots[i] = np.nan
        
if __name__ == "__main__":
    fsolvetol()

尽管

Answer 1

你忘了添加np。在你的根[i]前面，它应该是np.根[i]。或者你可以导入根

from scipy.optimize import fsolve
import numpy as np
from numpy import roots // Just do this and it should be fine

# Flags points that fsolve incorrectly assigns as roots within tolerance (now a function!)
def fsolvetol(func,guesses,roots,i,Tol):
    roots[i] = fsolve(func,guesses[i])
    if func(roots[i]) > Tol or func(roots[i]) < -Tol:
        roots[i] = np.nan
        
if __name__ == "__main__":
    fsolvetol()