How to append elements to a numpy array

Question

I want to do the equivalent to adding elements in a python list recursively in Numpy, As in the following code

matrix = open('workfile', 'w')
A = []
for row in matrix:
    A.append(row)

print A

I have tried the following:

matrix = open('workfile', 'w')
A = np.array([])
for row in matrix:
    A = numpy.append(row)

print A

It does not return the desired output, as in the list.

Edit this is the sample code:

mat = scipy.io.loadmat('file.mat')
var1 = mat['data1']
A = np.array([])
for row in var1:
    np.append(A, row)

print A

This is just the simplest case of what I want to do, but there is more data processing in the loop, I am putting it this way so the example is clear.

That looks more iteratively than recursively to me... also... have you looked at just using np.loadtxt to load data from files? — Jon Clements
– Jon Clements, Commented Mar 9, 2015 at 14:06
For what it's worth, none of your question has anything to do with recursion. You're just using looping, which is different from recursion. Beyond that, there's a lot of goofiness in your second example. You're trying to put strings of potentially arbitrary length into a Numpy array. — Iguananaut
– Iguananaut, Commented Mar 9, 2015 at 14:07
the variable "matrix" means that the file only contains numbers — user3025898
– user3025898, Commented Mar 9, 2015 at 14:13
A file containing only numbers can be loaded with np.loadtxt or np.genfromtxt. Appending to a NumPy array is slow. Avoid doing this if you can. — unutbu
– unutbu, Commented Mar 9, 2015 at 14:14
This is only an example, I know how to load files to numpy arrays and I know that it is better, the question is how to append values to numpy arrays in cases where I have to iterate as in a for loop. — user3025898
– user3025898, Commented Mar 9, 2015 at 14:18

user3590169 · Accepted Answer · 2015-03-09 14:42:10Z

23

You need to pass the array, A, to Numpy.

matrix = open('workfile', 'w')
A = np.array([])
for row in matrix:
    A = numpy.append(A, row)

print A

However, loading from the files directly is probably a nicer solution.

edited Mar 9, 2015 at 14:42

answered Mar 9, 2015 at 14:11

user3590169

3962 silver badges4 bronze badges

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

5 Comments

user3025898 Over a year ago

Please see the sample code provided, running your code returns an empty array []

user3590169 Over a year ago

Are you sure your input file is begin read correctly / has appropriate data? for i in range(5): np.append(A, i) Gives: array([ 0.]) array([ 1.]) array([ 2.]) array([ 3.]) array([ 4.]) So I believe this works.

user3025898 Over a year ago

this is the program I am running import numpy as np A = np.array([]) for i in range(5): np.append(A, i) print A it returns []

user3025898 Over a year ago

Thanks it is working now. Is there a way to save i as either a column or a as a row in your program?

Poik Over a year ago

@user3025898 Use the axis keyword argument for numpy.append. It will require getting the axes aligned with a transpose, i.e. A = numpy.append(A, row.T, axis=1)

Collectives™ on Stack Overflow

How to append elements to a numpy array

1 Answer 1

5 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related