10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [ ]:  # pip install numpy

In [ ]:  ​

In [1]:  import numpy as np

1. define arrays (vecteurs et matrices)

In [79]:  A = np.array([1, 2, 3]) # générateur par défaut, qui permet de convertir des listes (ou autres objets) en tableau
B = np.zeros((2, 3)) # tableau de 0 aux dimensions 2x3
C = np.ones((2, 3)) # tableau de 1 aux dimensions 2x3
​
print(A); print('\n')
print(B); print('\n')
print(C); print('\n')

[1 2 3]

[[0. 0. 0.]
[0. 0. 0.]]

[[1. 1. 1.]
[1. 1. 1.]]

In [80]:  np.full((2,2), 659)

Out[80]: array([[659, 659],

[659, 659]])

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 1/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [81]:  np.full_like(B, 4)

Out[81]: array([[4., 4., 4.],

[4., 4., 4.]])

In [85]:  # Repeat an array

r1 = np.repeat([A], 3, axis=0)
print(r1)

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

In [1]:  import numpy as np

In [2]:  A = np.linspace(1, 10, 10)

​
print(A);

[ 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.]

Pas comme les listes:

numpy arrays ne peuvent contenir qu'un seul type

In [10]:  import numpy as np

​
mixed_array = np.array([1, "apple", 3.14])
​
print(mixed_array)

['1' 'apple' '3.14']

In [12]:  mixed_array.dtype

Out[12]: dtype('<U32')

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 2/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

2. size, dim, shape

In [23]:  A = np.zeros((2, 3)) # création d'un tableau de shape (2, 3)
A

Out[23]: array([[0., 0., 0.],

[0., 0., 0.]])

In [24]:  A.size

Out[24]: 6

In [26]:  A.ndim

Out[26]: 2

In [25]:  A.shape

Out[25]: (2, 3)

3. indexing
In [1]:  import numpy as np
​
a = np.array([[1,2,3,4,5,6,7],
[8,9,10,11,12,13,14]])
print(a)

[[ 1 2 3 4 5 6 7]
[ 8 9 10 11 12 13 14]]

In [2]:  a[1][1]

Out[2]: 9

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 3/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [3]:  # Get a specific element [r, c]

a[1][5] #or a[1, 5]

Out[3]: 13

In [5]:  a

Out[5]: array([[ 1, 2, 3, 4, 5, 6, 7],

[ 8, 9, 10, 11, 12, 13, 14]])

In [6]:  a[-1] #dernière ligne

Out[6]: array([ 8, 9, 10, 11, 12, 13, 14])

In [8]:  a[-1][-1] #dernière ligne #dernière colonne

Out[8]: 14

In [9]:  a

Out[9]: array([[ 1, 2, 3, 4, 5, 6, 7],

[ 8, 9, 10, 11, 12, 13, 14]])

In [11]:  a[0, 2 : len(a[0]): 2] # a[ligne, start : end : steps]

Out[11]: array([3, 5, 7])

In [12]:  a

Out[12]: array([[ 1, 2, 3, 4, 5, 6, 7],

[ 8, 9, 10, 11, 12, 13, 14]])

In [13]:  a[1,5] = 20 #modifier la valeur dans la ligne 1 (deuxième ligne) et la colonne 5 (6ème col)
a

Out[13]: array([[ 1, 2, 3, 4, 5, 6, 7],

[ 8, 9, 10, 11, 12, 20, 14]])

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 4/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [15]:  a[: , 2] = [1,2] #modifier la colonne 2

a

Out[15]: array([[ 1, 2, 1, 4, 5, 6, 7],

[ 8, 9, 2, 11, 12, 20, 14]])

4. special functions
In [24]:  A = np.zeros((2, 3)) # création d'un tableau de shape (2, 3)
A

Out[24]: array([[0., 0., 0.],

[0., 0., 0.]])

In [25]:  A = A.reshape((3, 2)) # redimensionne le tableau A (3 lignes, 2 colonnes)

A

Out[25]: array([[0., 0.],

[0., 0.],
[0., 0.]])

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 5/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [42]:  A = np.zeros((2, 3))

B = np.ones((2, 3))
​
C = np.concatenate((A, B), axis=0) #concatener au niveau des lignes
​
print(A); print('\n')
print(B); print('\n')
print(C); print('\n')

[[0. 0. 0.]
[0. 0. 0.]]

[[1. 1. 1.]
[1. 1. 1.]]

[[0. 0. 0.]
[0. 0. 0.]
[1. 1. 1.]
[1. 1. 1.]]

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 6/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [43]:  A = np.zeros((2, 3))

B = np.ones((2, 3))

C = np.concatenate((A, B), axis=1) #concatener au niveau des colonnes

​
print(A); print('\n')
print(B); print('\n')
print(C); print('\n')

[[0. 0. 0.]
[0. 0. 0.]]

[[1. 1. 1.]
[1. 1. 1.]]

[[0. 0. 0. 1. 1. 1.]
[0. 0. 0. 1. 1. 1.]]

In [16]:  L = [1, 1, 1, 1, 3, 4, 5, 6, 6, 6, 6]
​
uniques, counts = np.unique(L, return_counts=True)

In [17]:  uniques

Out[17]: array([1, 3, 4, 5, 6])

In [18]:  counts

Out[18]: array([4, 1, 1, 1, 4], dtype=int64)

5. math operations (matrices)

all can be found here: https://numpy.org/doc/stable/reference/routines.math.html (https://numpy.org/doc/stable/reference/routines.math.html)

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 7/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [19]:  a = np.array([1,2,3,4])
print(a)

[1 2 3 4]

In [20]:  a + 2

Out[20]: array([3, 4, 5, 6])

In [21]:  a - 2

Out[21]: array([-1, 0, 1, 2])

In [22]:  a * 2

Out[22]: array([2, 4, 6, 8])

In [23]:  a / 2

Out[23]: array([0.5, 1. , 1.5, 2. ])

In [24]:  # b = np.array([1,0,1,0,1]) #error

b = np.array([1,0,1,0])
a + b

Out[24]: array([2, 2, 4, 4])

In [25]:  a ** 2

Out[25]: array([ 1, 4, 9, 16])

In [26]:  np.cos(a)

Out[26]: array([ 0.54030231, -0.41614684, -0.9899925 , -0.65364362])

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 8/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [27]:  np.sum(a)

Out[27]: 10

In [28]:  np.prod(a) #product of all elements

Out[28]: 24

In [29]:  np.cumsum(a)

Out[29]: array([ 1, 3, 6, 10])

In [30]:  np.cumprod(a)

Out[30]: array([ 1, 2, 6, 24])

In [31]:  np.diff(a) #2-1, 3-2, 4-3

Out[31]: array([1, 1, 1])

In [32]:  np.exp(a)

Out[32]: array([ 2.71828183, 7.3890561 , 20.08553692, 54.59815003])

In [33]:  np.log(a)

Out[33]: array([0. , 0.69314718, 1.09861229, 1.38629436])

In [34]:  np.log(np.exp(a))

Out[34]: array([1., 2., 3., 4.])

In [35]:  # --------------------------------------------------------------

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 9/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [36]:  a = np.array([1,2,3,4])
b = np.array([2,4,6,8])

In [37]:  print(a)

[1 2 3 4]

In [38]:  print(b)

[2 4 6 8]

In [39]:  print(a.shape)
print(b.shape)

(4,)
(4,)

In [40]:  np.add(a, b)

Out[40]: array([ 3, 6, 9, 12])

In [41]:  np.subtract(a, b)

Out[41]: array([-1, -2, -3, -4])

7. Matrix multiplication options:

1. np.multiply(a,b) ----------- + sum ----------- or a * b +sum
2. np.dot(a,b)
3. np.matmul(a, b) ----------- or @

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 10/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

1. for 1D arrays

In [44]:  a = np.array([1,2,3,4])
b = np.array([2,4,6,8])

In [45]:  a

Out[45]: array([1, 2, 3, 4])

In [46]:  b

Out[46]: array([2, 4, 6, 8])

In [47]:  np.multiply(a,b) #+ sum or a * b +sum

Out[47]: array([ 2, 8, 18, 32])

In [48]:  1*2 , 2*4 , 3*6 , 4*8

Out[48]: (2, 8, 18, 32)

In [49]:  np.dot(a, b)

Out[49]: 60

In [50]:  np.matmul(a, b)
# a@b

Out[50]: 60

In [51]:  1*2 + 2*4 + 3*6 + 4*8

Out[51]: 60

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 11/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

2. for 2D arrays and more

In [52]:  c = np.array([[1, 2, 3],

[4, 5, 6]])
c.shape

Out[52]: (2, 3)

In [53]:  d = np.array([[6, 2, 1],

[8, 9, 7]])
d.shape

Out[53]: (2, 3)

In [54]:  np.multiply(c,d) #element wise

Out[54]: array([[ 6, 4, 3],

[32, 45, 42]])

In [55]:  np.matmul(c, d) #error

---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
Cell In[55], line 1
----> 1 np.matmul(c, d) #error

ValueError: matmul: Input operand 1 has a mismatch in its core dimension 0, with gufunc signature (n?,k),(k,m?)->
(n?,m?) (size 2 is different from 3)

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 12/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [56]:  np.dot(c, d) #error

---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
Cell In[56], line 1
----> 1 np.dot(c, d) #error

File <__array_function__ internals>:200, in dot(*args, **kwargs)

ValueError: shapes (2,3) and (2,3) not aligned: 3 (dim 1) != 2 (dim 0)

In [57]:  d

Out[57]: array([[6, 2, 1],

[8, 9, 7]])

In [58]:  dt = d.T #Transposed or reshape

dt

Out[58]: array([[6, 8],

[2, 9],
[1, 7]])

In [59]:  dr = d.reshape(3,2) #not like transposed !! but works

dr

Out[59]: array([[6, 2],

[1, 8],
[9, 7]])

In [60]:  np.dot(c, dt)

Out[60]: array([[ 13, 47],

[ 40, 119]])

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 13/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [61]:  np.matmul(c, dt)

Out[61]: array([[ 13, 47],

[ 40, 119]])

In [62]:  np.dot(c, dr)

Out[62]: array([[35, 39],

[83, 90]])

7. read and write txt files

In [66]:  # 1. load text file
​
f = np.loadtxt('test.txt')
f

Out[66]: array([1., 2., 3., 4., 5., 6.])

In [67]:  f.shape

Out[67]: (6,)

In [68]:  # 2. modify
​
f[0] = 999
f

Out[68]: array([999., 2., 3., 4., 5., 6.])

In [69]:  # 3. save
​
np.savetxt('test3.txt', f, fmt='%d') #delimiter=',')

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 14/15

10/12/23, 7:37 PM Cours_numpy - Jupyter Notebook

In [ ]:  ​

localhost:8888/notebooks/séance 3/Cours_numpy.ipynb 15/15