Introduction to NumPy

numpy is a powerful set of tools to perform mathematical operations of on lists of numbers. It works faster than normal python lists operations and can manupilate high dimentional arrays too.

Additional material:

Import NumPy



In [ ]:

    
import numpy as np

Why NumPy?

Python has powerful data types. Why do we need another?

Generate a list



In [ ]:

    
# Regular Python
%timeit python_list_1 = list(range(1, 1000))

python_list_1 = list(range(1, 1000))
python_list_2 = list(range(1, 1000))



In [ ]:

    
#Numpy
%timeit numpy_list_1 = np.arange(1, 1000)

numpy_list_1 = np.arange(1, 1000)
numpy_list_2 = np.arange(1, 1000)

Basic Operations



In [ ]:

    
%%timeit
# Regular Python
[(x + y) for x, y in zip(python_list_1, python_list_2)]
[(x - y) for x, y in zip(python_list_1, python_list_2)]
[(x * y) for x, y in zip(python_list_1, python_list_2)]
[(x / y) for x, y in zip(python_list_1, python_list_2)]



In [ ]:

    
%%timeit
#Numpy
numpy_list_1 + numpy_list_2
numpy_list_1 - numpy_list_2
numpy_list_1 * numpy_list_2
numpy_list_1 / numpy_list_2

np.array

Numpy Arrays are central to Numpy. They ca be created in several ways.
np.arange([start,] stop[, step,], dtype=None)



In [ ]:

    
np.arange(10)



In [ ]:

    
np.arange(3, 10)



In [ ]:

    
np.arange(1, 10, 0.5)



In [ ]:

    
np.arange(1, 10, 2, dtype=np.float64)

np.array(object, dtype=None, ...)



In [ ]:

    
np.array([1,2,3])



In [ ]:

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



In [ ]:

    
a = np.array([[1, 1], [1, 1]])
a



In [ ]:

    
a = np.array([[[1, 1], [1, 1]], [[1, 1], [1, 1]]])
a



In [ ]:

    
a.shape

np.linspace(start, stop, num=50, dtype=None, ...)



In [ ]:

    
np.linspace(1, 10)



In [ ]:

    
np.linspace(1, 10, num=12)

Examing np arrays



In [ ]:

    
ds = np.arange(1, 10, 2)
ds.ndim



In [ ]:

    
ds.shape



In [ ]:

    
ds.size



In [ ]:

    
len(ds)



In [ ]:

    
ds.dtype



In [ ]:

    
ds

Common methods

zeros(shape, dtype=float, ...)



In [ ]:

    
np.zeros((3, 4))



In [ ]:

    
np.zeros((3, 4), dtype=np.int64)

np.linspace(start, stop, num=50, endpoint=True)



In [ ]:

    
np.linspace(1, 5)



In [ ]:

    
np.linspace(0, 2, num=4)



In [ ]:

    
np.linspace(0, 2, num=4, endpoint=False)

random_sample(size=None)



In [ ]:

    
np.random.random((2,3))



In [ ]:

    
np.random.choice(np.arange(10), 3)

Statistics



In [ ]:

    
data_set = np.random.random((3,4))
data_set



In [ ]:

    
np.max(data_set)



In [ ]:

    
np.max(data_set, axis=0)



In [ ]:

    
np.max(data_set, axis=1)



In [ ]:

    
np.min(data_set)



In [ ]:

    
np.mean(data_set)



In [ ]:

    
np.median(data_set)



In [ ]:

    
np.std(data_set)



In [ ]:

    
np.sum(data_set)



In [ ]:

    
np.argmax(data_set)



In [ ]:

    
np.argmin(data_set)



In [ ]:

    
np.argsort(data_set)

Reshaping



In [ ]:

    
np.reshape(data_set, (4, 3))



In [ ]:

    
np.reshape(data_set, (12, 1))



In [ ]:

    
np.reshape(data_set, (12))



In [ ]:

    
np.ravel(data_set)

Slicing



In [ ]:

    
data_set = np.random.randint(0, 10, size=(5, 10))
data_set



In [ ]:

    
data_set[1]



In [ ]:

    
data_set[1][0]



In [ ]:

    
data_set[1, 0]

Range



In [ ]:

    
data_set[2:4]



In [ ]:

    
data_set[2:4, 0]



In [ ]:

    
data_set[2:4, 0:2]



In [ ]:

    
data_set[:,0]

Stepping



In [ ]:

    
data_set[2:4:1]



In [ ]:

    
data_set[::]



In [ ]:

    
data_set[::2]



In [ ]:

    
data_set[2:4,::2]



In [ ]:

    
data_set[2:4,::-2]