How to create an array in Python

Question

Trying to write my first python program. In a working sample program (script), some array of data is defined like this:

x_data = np.random.rand(100).astype(np.float32)

And when I subsequently type "x_data" in a Python console, it returns

    >>> x_data
        array([ 0.16771448,  0.55470788,  0.36438608, ...,  0.21685787,
        0.14241569,  0.20485006], dtype=float32)

and the script works.

Now I want to use my own data sets instead. I'm trying a statement like this

my_data = [1,2,3,4,5]

and replace use of x_data with my_data, but then the program doesn't work. I notice that when I type "my_data" in the Python console, it returns

>>> my_data
    [1, 2, 3, 4, 5]

which is missing the parts that say "array" and "dtype=float32". I'm guessing that difference is related to the problem.

How can I declare a dataset my_data that would be treated like x_data so I can feed my own data into the program?

I think it's irrelevant, but here's the full sample script I started from (which works):

import tensorflow as tf
import numpy as np

# Create 100 phony x, y data points in NumPy, y = x * 0.1 + 0.3
x_data = np.random.rand(100).astype(np.float32)
y_data = x_data * 0.1 + 0.3

# Try to find values for W and b that compute y_data = W * x_data + b
# (We know that W should be 0.1 and b 0.3, but TensorFlow will
# figure that out for us.)
W = tf.Variable(tf.random_uniform([1], -1.0, 1.0))
b = tf.Variable(tf.zeros([1]))
y = W * x_data + b

# Minimize the mean squared errors.
loss = tf.reduce_mean(tf.square(y - y_data))
optimizer = tf.train.GradientDescentOptimizer(0.5)
train = optimizer.minimize(loss)

# Before starting, initialize the variables.  We will 'run' this first.
init = tf.global_variables_initializer()

# Launch the graph.
sess = tf.Session()
sess.run(init)

# Fit the line.
for step in range(201):
    sess.run(train)
    if step % 20 == 0:
        print(step, sess.run(W), sess.run(b))

# Learns best fit is W: [0.1], b: [0.3]

Answer 1

I'm guessing you're coming from Matlab?

Python square-bracket notation doesn't give you any type of array by default: it gives you a simpler object of the built-in type list . numpy , the ubiquitous third-party package, is what you want to use for arrays. Clearly you already have it available.

The second line below converts your variable from a list to a numpy array with the same data type as your other array x_data :

my_data = [1,2,3,4,5]
my_data = np.array(my_data, dtype=np.float32)

Answer 2

numpy will construct an array from a list if you just use np.array :

import numpy as np
arr = np.array([1, 2, 3, 4])

Note that you can specify the datatype as well:

arr_int32 = np.array([1, 2, 3, 4], dtype=np.int32)
arr_float = np.array([1, 2, 3, 4], dtype=np.float64)

Also note that there are times when you might be working with an object that might be a list or it might be a numpy array. np.array will copy an array if you pass that in as input. That is sometimes not desireable for performance reasons. If you find yourself in that situation, you can use np.asarray which will convert non-arrays into arrays but it'll return arrays untouched.