Fonctions usuelles:
a = np.random.rand(2,3)
print(a)
'''
[[0.04703464 0.930676 0.17006579]
[0.74658188 0.23148728 0.86145293]]
'''
print(a.sum()) # 2.9872985242505363
print(a.mean()) # 0.4978830873750894
print(np.median(a)) # 0.4890345803558861
print(a.std()) # 0.3566134819519723
print(a.var()) # 0.12717317550990967
print(a.min()) # 0.04703464248321543
print(a.max()) # 0.9306760015352592
print(a.ptp()) # 0.8836413599999999 - range
print(np.percentile(a, 95)) # 0.9133702324999999
Somme cumulée
print(a.cumsum())
# [0.04703464 0.97771064 1.14777643 1.89435831 2.12584559 2.98729852]
Produit
print(np.prod(np.array([1, 2, 3, 4])))
# 24
Produit cumulé
print(np.cumprod(np.array([1, 2, 3, 4])))
# [ 1 2 6 24]
Toutes ces fonctions peuvent prendre axis pour paramètre, ce qui aura pour effet d’appliquer la fonction sur chaque ligne (1) ou colonne (0)
print(a.sum(axis=0))
'''
[0.79361652 1.16216328 1.03151872]
'''
print(a.sum(axis=1))
'''
[1.14777643 1.83952209]
'''
print(a.sum(axis=1, keepdims=True))
'''
[[1.14777643]
[1.83952209]]
'''
print(a.cumsum(axis=0))
'''
[[0.04703464 0.930676 0.17006579]
[0.79361652 1.16216328 1.03151872]]
'''
Mode
Il n’existe pas de méthode pour trouver le mode directement, il faut le faire manuellement
a = np.array([1,2,2,3,4,1,2])
print(np.bincount(a).argmax()) # 2
x = np.array([[0.2, 0.5, 1.2]])
# À l'entier en-dessous
print(np.floor(x)) # [[0. 0. 1.]]
# À l'entier au-dessus
print(np.ceil(x)) # [[1. 1. 2.]]
# À l'entier le plus proche
print(np.round(x)) # [[0. 0. 1.]]
print(np.floor(x + 0.5)) # [[0. 1. 1.]]
# Inner product
x = np.array([[1,2],[3,4]])
y = np.array([[10,20],[30,40]])
print(x.dot(y))
print(np.dot(x, y))
print(np.matmul(x, y))
'''
[[ 70 100]
[150 220]]
'''
# Outer product
print(np.outer(x, y))
'''
[[ 10 20 30 40]
[ 20 40 60 80]
[ 30 60 90 120]
[ 40 80 120 160]]
'''
print(np.pi) # 3.141592653589793
print(np.e) # 2.718281828459045
print(np.euler_gamma) # 0.5772156649015329
print(np.PINF) # inf
print(np.NINF) # -inf
print(np.NAN) # nan
print(np.PZERO) # 0.0
print(np.NZERO) # -0.0
np.negative(x) # 1-X
np.square(x) # Carré
np.sqrt(x) # Racine carrée
np.abs(x) # Valeur absolue
np.exp(x) # Exponentielle
np.log(x) # Log exp(x)
np.log1p(x) # Log exp(x)-1
np.log10(x) # Log 10
np.log2(x) # Log 2
np.sin(x) # Sinus
np.cos(x) # Cosinus
np.tan(x) # Tangente
np.hypot(x1, x2) # Hypothénuse
np.arcsin(x)
np.arccos(x)
np.arctan(x)
np.arctan2(x1, x2)
np.sinh(x)
np.cosh(x)
np.tanh(x)
np.arcsinh(x)
np.arccosh(x)
np.arctanh(x)
np.degrees(x) # Radian -> degres
np.radians(x) # Degres -> radians
# Greatest Common Denominator
print(np.gcd(6, 9)) # 3
print(np.gcd.reduce(np.array([6, 9, 15]))) # 3
# Lowest Common Multiple
print(np.lcm(6, 9)) # 18
print(np.lcm.reduce(np.array([6, 9, 15]))) # 90