NumPy Data Types and Shape

Every NumPy array has one data type for all its items, and a shape that describes its layout.

The dtype Attribute

dtype tells you the type of the values in an array:

import numpy as np

a = np.array([1, 2, 3])
b = np.array([1.5, 2.0])
c = np.array(["Ali", "Sara"])
d = np.array([True, False])
print(a.dtype)
print(b.dtype)
print(c.dtype)
print(d.dtype)

int64
float64
<U4
bool

<U4 means text (Unicode) of up to 4 characters. On some Windows systems you may see int32 instead of int64.

Common Data Types

Choosing and Converting Types

Give a dtype when you create the array, or convert later with astype():

import numpy as np

a = np.array([1, 2, 3], dtype="float64")
print(a)

prices = np.array([99.9, 150.5, 20.2])
whole = prices.astype(int)
print(whole)
print(whole.dtype)

[1. 2. 3.]
[ 99 150  20]
int64

astype() returns a new array. Like Python's int(), converting to int cuts off the decimal part.

Shape and Size

import numpy as np

sales = np.array([
    [1200, 1500, 900, 1100],
    [1300, 1700, 1000, 1250],
    [1100, 1400, 950, 1050]
])
print(sales.shape)
print(sales.ndim)
print(sales.size)

(3, 4)
2
12

Shape (3, 4) means 3 rows and 4 columns.

Reshaping

reshape() gives the same data a new shape. The total number of items must stay the same.

import numpy as np

arr = np.arange(1, 13)
print(arr.reshape(3, 4))
print(arr.reshape(2, 2, 3).shape)
print(arr.reshape(2, -1))

[[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]]
(2, 2, 3)
[[ 1  2  3  4  5  6]
 [ 7  8  9 10 11 12]]

Use -1 for one dimension and NumPy works it out for you. Here 12 items in 2 rows means 6 columns.

Trying an impossible shape gives an error:

import numpy as np
np.arange(10).reshape(3, 4)

ValueError: cannot reshape array of size 10 into shape (3,4)

Flattening

To turn any array back into 1-D, use flatten() (a copy) or ravel() (a view when possible):

import numpy as np

grid = np.array([[1, 2], [3, 4]])
print(grid.flatten())
print(grid.T)

[1 2 3 4]
[[1 3]
 [2 4]]

.T gives the transpose: rows become columns and columns become rows.

Exercise