Python Array is read-only, can't append values

Question

I am new to Python. The following code is causing an error when it attempts to append values to an array. What am I doing wrong?

import re
from array import array

freq_pattern = re.compile("Frequency of Incident[\(\)A-Za-z\s]*\.*\s*([\.0-9]*)")
col_pattern = re.compile("([-\.0-9]+)\s+([-\.0-9]+)\s+([-\.0-9]+)\s+([-\.0-9]+)\s+([-\.0-9]+)")
e_rcs = array('f')

f = open('example.4.out', 'r')

for line in f:
    print line,

    result = freq_pattern.search(line)
    if result:
        freq = float(result.group(1))

    cols = col_pattern.search(line)
    if cols:
        e_rcs.append = float(cols.group(2))

f.close()

Error

Traceback (most recent call last):
File "D:\workspace\CATS Parser\cats-post.py", line 31, in e_rcs.append = float(cols.group(2)) AttributeError: 'list' object attribute 'append' is read-only attributes (assign to .append)

Answer 1

Do you want to append to the array?

e_rcs.append( float(cols.group(2)) )

Doing this: e_rcs.append = float(cols.group(2)) replaces the append method of the array e-rcs with a floating-point value. Rarely something you want to do.

Answer 2

append is a method. You're trying to overwrite it instead of calling it.

e_rcs.append(float(cols.group(2)))

Answer 3

You are assigning to the append() function, you want instead to call .append(float(cols.group(2))).

Answer 4

Try this instead:

import re

freq_pattern = re.compile("Frequency of Incident[\(\)A-Za-z\s]*\.*\s*([\.0-9]*)")
col_pattern = re.compile("([-\.0-9]+)\s+([-\.0-9]+)\s+([-\.0-9]+)\s+([-\.0-9]+)\s+([-\.0-9]+)")
e_rcs = [] # make an empty list

f = open('example.4.out', 'r')

for line in f:
    print line,

    result = freq_pattern.search(line)
    if result:
        freq = float(result.group(1))

    cols = col_pattern.search(line)
    if cols:
        e_rcs.append( float(cols.group(2)) ) # add another float to the list

f.close()

In Python you would only use array.array when you need to control the binary layout of your storage, i.e. a plain array of bytes in RAM.

If you are going to be doing a lot of scientific data analysis, then you should have a look at the NumPy module which supports n-dimensional arrays. Think of NumPy as a replacement for FORTRAN in doing mathematics and data analysis.