numpy array: accessing element in a loop gives an error

Question

This code throws an exception:

"list index out of range"

at the line marked below.

col_sig_squared = [np.zeros(shape=(1,6), dtype=int)]

def calculate_col_sigma_square(matrix):
    mx = np.asarray(matrix)
    for(x,y), value in np.ndenumerate(matrix):
    if(x > 4):
            continue
    else:
        val = matrix[x][y] - x_bar_col[x]
        val = val**2
        EXCEPTION-->print col_sig_squared[y]

Why is this a problem? col_sig_squared is an array with indices. Why can't I access it like this. Tried a bunch of things but not sure why this syntax is wrong. I'm new to Python and its intricacies, any help would be appreciated.

Thanks

Generally don't use matrix[x][y] use matrix[x,y] there is a big difference if x is not a scalar. — seberg
– seberg, Commented Oct 24, 2012 at 21:18

kindall · Accepted Answer · 2012-10-24 19:19:20Z

1

Well, the message tells you what is wrong pretty clearly. y is bigger at that point than the number of items in col_sig_squared. This doesn't surprise me because col_sig_squared is defined as a list with one item, a NumPy array:

col_sig_squared = [np.zeros(shape=(1,6), dtype=int)]

This means that only col_sig_squared[0] is valid.

Maybe you meant:

col_sig_squared = np.zeros(shape=(1,6), dtype=int)

Now col_sig_squared is a NumPy array.

answered Oct 24, 2012 at 19:19

kindall

185k36 gold badges291 silver badges321 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Community · Accepted Answer · 2017-05-23 12:12:12Z

0

It is more common to express NumPy array vectors using shape = (N,). For instance:

>>> col_sig_1byN = np.zeros(shape=(1,6), dtype=int)
>>> col_sig_N = np.zeros(shape=(6,), dtype=int)
>>> print col_sig_1byN
[[0 0 0 0 0 0]]
>>> print col_sig_N
[0 0 0 0 0 0]

You can index col_sig_N as col_sig_N[p], but for col_sig_1byN you have to do col_sig_1byN[0,p] - note that [x,y] is the way to index into a multidimensional NumPy array.

To index a whole row/column you can do [x,:]/[:,y].

And, as kindall said, you should not embed your array in a list.

edited May 23, 2017 at 12:12

CommunityBot

11 silver badge

answered Oct 24, 2012 at 20:55

mtrw

35.3k7 gold badges66 silver badges73 bronze badges

Collectives™ on Stack Overflow

numpy array: accessing element in a loop gives an error

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related