Introduction to NumPy and Matplotlib
Chapter 4: Array reshaping
Reshape an array
Excellent, oh young one. You have accomplished your first mission. It gets harder now!
Mission 2: A makeover!
I have a (3 \times 4) np.array, but I need it reshaped it to be a (6 \times 2) np.array instead.
Again, the solution is staring right in front of you in the official documentation. And again, you only need a single statement!
x = np.array([[0, 1, 2, 3],
[4, 5, 6, 7],
[-6, -5, -4, -3]
])
reshaped_x = ????
assert np.all(reshaped_x == np.array([[ 0, 1],
[ 2, 3],
[ 4, 5],
[ 6, 7],
[-6, -5],
[-4, -3]]))
Again, no peeking until you have tried it!
This one is pretty straightforward, although there are multiple variants to the solutions!
reshaped_x = x.reshape((6, 2))
# This also works
reshaped_x = x.reshape(6, 2)
# You can make one of the dimensions -1.
# The dimension will be inferred automatically
reshaped_x = x.reshape((-1, 2))
reshaped_x = x.reshape((6, -1))
# There is also a function version if you are not into OOP.
# This only takes a tuple as the second parameter.
reshaped_x = np.reshape(x, (6, 2))
There is also a .resize() method. This is a mutable version of .reshape() which modifies the original array directly.
>>> x = np.array([[0, 1, 2, 3],
... [4, 5, 6, 7],
... [-6, -5, -4, -3]
... ])
...
>>> x.resize((6, 2))
>>> print(x)
[[ 0 1]
[ 2 3]
[ 4 5]
[ 6 7]
[-6 -5]
[-4 -3]]