10: Numpy Tableau ndarray

Contents

10: Numpy Tableau ndarray#

import numpy as np

1. ndarray** generators#

default generator : ndarray()
1D generator : np.linspace and np.arange()
ND generator: np.zeros(), np.ones(), np.random.randn() (these are the most useful)

A = np.array([1, 2, 3]) # default generator, which allows to convert lists (or other objects) into ndarray
A = np.zeros((2, 3)) # array of 0 with dimensions 2x3
B = np.ones((2, 3)) # array of 1 with dimensions 2x3
C = np.random.randn(2, 3) # random array (normal distribution) with dimensions 2x3
D = np.random.rand(2, 3) # random array (uniform distribution)

 E = np.random.randint(0, 10, [2, 3]) # array of random integers from 0 to 10 and of dimension 2x3

A = np.ones((2, 3), dtype=np.float16) # define the type and the space to occupy on the memory
B = np.eye(4, dtype=np.bool) # create an identity matrix and convert the elements into bool type.

/tmp/ipykernel_3860723/3272918426.py:2: DeprecationWarning: `np.bool` is a deprecated alias for the builtin `bool`. To silence this warning, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.
Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
  B = np.eye(4, dtype=np.bool) # create an identity matrix and convert the elements into bool type.

A = np.linspace(1,10, 10)
B = np.arange(0, 10, 10)

2. Important attributes#

size
shape

A = np.zeros((2, 3)) # create an array of shape (2, 3)
 
print(A.size) # the number of elements in the array A
print(A.shape) # the dimensions of the array A (as a Tuple)
 
print(type(A.shape)) # here is the proof that the shape is a tuple
 
print(A.shape[0]) # the number of elements in the first dimension of A

6
(2, 3)
<class 'tuple'>
2

3. Important methods#

reshape() : to resize an array
ravel()** : to flatten an array (make it one dimension only)
squeeze()** : when a dimension is equal to 1, this dimension disappears
concatenate() : assembles 2 arrays together along an axis (also exists in hstack and vstack)

A = np.zeros((2, 3)) # create a shape array (2, 3)
 
A = A.reshape((3, 2)) # resize the array A (3 rows, 2 columns)
A.ravel() # flatten array A (one dimension)
A.squeeze() # eliminate the "1" dimensions of A.

array([[0., 0.],
       [0., 0.],
       [0., 0.]])

A = np.zeros((2, 3)) # create a shape array (2, 3)
B = np.ones((2, 3)) # create a shape array (2, 3)

np.concatenate((A, B), axis=0) # axis 0 : equivalent of np.vstack((A, B))

array([[0., 0., 0.],
       [0., 0., 0.],
       [1., 1., 1.],
       [1., 1., 1.]])

np.concatenate((A, B), axis=1) # axis 1 : equivalent of np.hstack((A, B))

array([[0., 0., 0., 1., 1., 1.],
       [0., 0., 0., 1., 1., 1.]])

4. exercise and solutions#

def initialization(m, n):
    # m : number of rows
    # n : number of columns
    # return a random matrix (m, n+1)
    # with a bias column (filled with "1") on the right

    return X

SOLUTION#