索引超出范围错误：Python

Question

I have a set of data that is not smooth. Hence I'm trying to interpolate it with the generated mesh.

import numpy as np
data=np.loadtxt('test.txt')
x=data[:,0] #len(x) = 730
y=data[:,1]

Nx= len(x)
Ny=len(y)
del_x= 0.5
xn = np.linspace(0,Nx,2000)
yn = np.linspace(0,Ny,2000)

#loop 
for i in range(0,Nx-1):
 if x[i] > xn[i] and x[i] < xn[i+1]:
  new_x= (i + (xn[i] - x[i])/(xn[i]-xn[i+1]))*del_x
print new_x

I would like to perform the loop that basically does the operation ie : if my original data x[i] is in between the two grid points xn[i],xn[i+1], then compute new_x.

But I get the following error

Traceback (most recent call last):
 File "new.py", line 27, in <module>
  if x[i] > xn[i] and x[i] < xn[i+1]:
IndexError: index out of bound

Can someone help me out ?

Answer 1

If len(x) = 730 than its too short for the loop. Your loop iterates 1000 times, i would be at some point bigger than 730 and there are no values like x[730] , x[731] etc. You will get an out of bound error because you are accessing not existing values in an array.

Answer 2

您可以这样更改for循环：

for i in range(0, len(x)):

Answer 3

First some basics

1. Python is "zero indexed", meaning that the first element in the list is indexed as 0 .
2. Ranges are given as: [start, stop[ , meaning that the last element is not included.

Just see what happens if you try:

a = [1, 2, 3]
print("a has %d elements:" %(len(a)))
for i in range(len(a)):
  print("a[%d] = %d" %(i, a[i]))

The output is:

a has 3 elements:
a[0] = 1
a[1] = 2
a[2] = 3

If you try to access a[3] you will see:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
IndexError: list index out of range

Now to your problems

Numpy "linspace" takes the following input: numpy.linspace(start, stop, N) , resulting in a closed range [start, stop] (notice it includes the "stop") with N elements.

You want the number of elements, or len(xn) to be 730, and I am guessing you want the "stop" value to be max(x) :

xn = numpy.linspace(0, max(x), len(x))

for i in range(len(x)):
  ...

Tip: If you are uncertain about the signature of a function you can always ask for help , try: help(numpy.linspace)

Answer 4

numpy linspace official documentation

xn = np.linspace(0,Nx,2000)

in here 0 is start value, Nx is end value and 2000 is the number of values to be generated.

returned object of linspace is something like :

[0, 0.5, 1, ... , ] #2000 elements ie a row containing 2000 elements

x=data[:,0] x is something like :

[[1],[2],[3],[4], .... ,[]] #column vector with number of elements in text data

now row size of x can not be greater than 1 ie index has to be less than 1 or equal to 0

and you can't compare these two arrays

[ ... ] and [[], [], [], [], ... ]

are incomparable (you can compare the first element though.)

you could use this : reshape

refer to this answer : What does -1 mean in numpy reshape?

row_x = x.reshape(-1)

will give you a row vector.

and you can convert to column format by:

col_x = row_x.reshape(-1,1)

I would appreciate edits as I am currently at work and thus wrote the answer in hurry.