ndarray.shape
Returns a tuple of array dimensions. It can also be used to "reshape" the array, as long as this would not require a change in the total number of elements.
1>>> a.shape
2(2, 3)
3>>> x.shape
4(3,)
5>>> a.shape = (3,2)
6>>> a
7array([[ 1., 2.],
8 [ 3., 4.],
9 [ 5., 6.]])
10
ndarray.size
Returns the number of elements in the array. This can also be calculated with np.prod(a.shape), i.e., the product of the array’s dimensions.
1>>> a
2array([[ 1., 2.],
3 [ 3., 4.],
4 [ 5., 6.]])
5>>> a.size
66
7>>> np.prod(a.shape)
86
9
NumPy array: default types
NumPy types are typically numeric types such as integers and floats of different precisions.
1>>> import numpy as np
2>>> qty = np.array([5,7,4,10,16])
3>>> qty.dtype
4dtype('int32')
5>>> price = np.array([5,8,12.7,89.6,12.9,5.4])
6>>> price.dtype
7dtype('float64')
8>>>
9
There are also some other types:
| NumPy type | Example |
|---|
| Int | myarr = np.array( [1, 5, 0, 3 ] ) |
| Float | myarr = np.array( [1, 5, 0, 3.1] ) |
| Complex | myarr = np.array( [1+2j, 3+4j, 6+7j] ) |
| Bool | myarr = np.array( [True, False, True] ) |
| String | myarr = np.array( ['Jon', 'Georg', 'Jose'] ) |
Casting Array Data Types
An array with one dtype can be converted or casted into another dtype using ndarray’s astype method.
1>>> import numpy as np
2>>> qty = np.array([5,7,4,10,16])
3>>> qty.dtype
4dtype('int32')
5>>> float_qty = qty.astype(np.float64)
6>>> float_qty.dtype
7dtype('float64')
8>>>
9
Casting can be specifically useful if you have an array of strings representing numeric data (numbers in quotes).
1>>> prices = np.array(['5.1','8.6','12.7','89.6','12.9','5.4'])
2>>> prices
3array(['5.1', '8.6', '12.7', '89.6', '12.9', '5.4'],
4 dtype='<U4')
5>>> prices_num = prices.astype(float)
6>>> prices_num
7array([ 5.1, 8.6, 12.7, 89.6, 12.9, 5.4])
8>>> prices_num.dtype
9dtype('float64')
10>>>
11
