Array of arrays (Python/NumPy)

Question

I am using Python/NumPy, and I have two arrays like the following:

array1 = [1 2 3]
array2 = [4 5 6]

And I would like to create a new array

Answer 1

a=np.array([[1,2,3],[4,5,6]])

a.tolist()

tolist method mentioned above will return the nested Python list

Answer 2

If the file is only numerical values separated by tabs, try using the csv library: http://docs.python.org/library/csv.html (you can set the delimiter to '\t')

If you have a textual file in which every line represents a row in a matrix and has integers separated by spaces\tabs, wrapped by a 'arrayname = [...]' syntax, you should do something like:

import re
f = open("your-filename", 'rb')
result_matrix = []
for line in f.readlines():
    match = re.match(r'\s*\w+\s+\=\s+\[(.*?)\]\s*', line)
    if match is None:
        pass # line syntax is wrong - ignore the line
    values_as_strings = match.group(1).split()
    result_matrix.append(map(int, values_as_strings))

Answer 3

It seems strange that you would write arrays without commas (is that a MATLAB syntax?)

Have you tried going through NumPy's documentation on multi-dimensional arrays?

It seems NumPy has a "Python-like" append method to add items to a NumPy n-dimensional array:

>>> p = np.array([[1,2],[3,4]])

>>> p = np.append(p, [[5,6]], 0)

>>> p = np.append(p, [[7],[8],[9]],1)

>>> p
array([[1, 2, 7], [3, 4, 8], [5, 6, 9]])

It has also been answered already...

From the documentation for MATLAB users:

You could use a matrix constructor which takes a string in the form of a matrix MATLAB literal:

mat("1 2 3; 4 5 6")

or

matrix("[1 2 3; 4 5 6]")

Please give it a try and tell me how it goes.

Answer 4

You'll have problems creating lists without commas. It shouldn't be too hard to transform your data so that it uses commas as separating character.

Once you have commas in there, it's a relatively simple list creation operations:

array1 = [1,2,3]
array2 = [4,5,6]

array3 = [array1, array2]

array4 = [7,8,9]
array5 = [10,11,12]

array3 = [array3, [array4, array5]]

When testing we get:

print(array3)

[[[1, 2, 3], [4, 5, 6]], [[7, 8, 9], [10, 11, 12]]]

And if we test with indexing it works correctly reading the matrix as made up of 2 rows and 2 columns:

array3[0][1]
[4, 5, 6]

array3[1][1]
[10, 11, 12]

Hope that helps.