Appending lists, arrays, matrices

You have somehow ended up with a list containing a mixture of single-element matrices and ordinary numbers. If you cannot avoid this, then you could always clean up the list for printing using something like the following:

def extract_number(x):
    try:
        return x.item(0)    # returns first item in x, if x is a matrix
    except AttributeError:
        return x            # if x is not a matrix or array, just return x

print([extract_number(x) for x in MSEV])

This will output the list in the standard way that Python displays lists

[1421, 7, 1184, 3, 925, 2, 1077, 4]

To display it more nicely you can look into string formatting and do something like this:

print("""MSEa     Ca MSEb     Cb 
{:4}{:6}{:6}{:7}
{:4}{:6}{:6}{:7}""".format(*msev))

I’m a little (or a lot) confused about how and when to use lists, arrays, and matrices and how to convert one to the other

Since you tagged it numpy, I am going to assume you mean numpy.array and numpy.matrix. Then, in short, when to use

• matrices: basically, never. np.matrix is an endless source of confusion. Better use arrays, use dot for matrix multiplication, or @ operator if you can use python 3.5+ only.

• arrays: use them when you know the dimensions and sizes (shape, in numpy parlance) in advance.

• lists: use lists when you need to append/remove elements.

As to how to convert things: given a list, lst, np.asarray(lst) is a numpy array. Given an array, arr, arr.tolist() is a list. Given a matrix, np.asarray conveerts it into an array, and if your matrix is 1-by-1, m[0, 0] is a scalar.

I'm not sure where your matrix comes from, and why the mix of matrix and numbers, but I'll try to process it.

For ease of copy-n-paste I'll define matrix (I've loaded numpy as np). I'm working in a Ipython session:

In [373]: matrix=np.matrix

In [375]: alist=[matrix([[ 1421]]), 7, matrix([[ 1184]]), 3, matrix([[ 925]]), 2, matrix([[ 1077]]), 4]

In [376]: alist
Out[376]: 
[matrix([[1421]]),
 7,
 matrix([[1184]]),
 3,
 matrix([[925]]),
 2,
 matrix([[1077]]),
 4]

I'll run it through a list comprehension with a ifelse expression that extracts the element from the matrices. I could also have defined a simple helper function as Stuart did:

In [379]: newlist=[x[0,0] if isinstance(x,np.matrix) else x for x in alist]
In [380]: newlist
Out[380]: [1421, 7, 1184, 3, 925, 2, 1077, 4]

now turn it into an array - and use reshape to make it 2 rows.

In [381]: Marray=np.array(newlist).reshape(2,-1)
In [382]: Marray
Out[382]: 
array([[1421,    7, 1184,    3],
       [ 925,    2, 1077,    4]])

np.savetxt is the numpy function to write csv style files:

In [386]: np.savetxt('test.txt',Marray)
In [387]: cat test.txt
1.421000000000000000e+03 7.000000000000000000e+00 1.184000000000000000e+03 3.000000000000000000e+00
9.250000000000000000e+02 2.000000000000000000e+00 1.077000000000000000e+03 4.000000000000000000e+00

Oops, default format is float; change that to integer:

In [388]: np.savetxt('test.txt',Marray, '%d')

and look at the resulting file (just like doing cat in a linux shell)

In [389]: cat test.txt
1421 7 1184 3
925 2 1077 4

and with a header line:

In [392]: np.savetxt('test.txt',Marray, '%d',header='MSEa Ca MSEb Cb')

In [393]: cat test.txt
# MSEa Ca MSEb Cb
1421 7 1184 3
925 2 1077 4

Format can be refined, but that gives the idea.