Numpy

Question 1

IMPORTING

What is the statement needed to import the numpy module.

Answer 1

import numpy as np

Question 2

ARRAY CREATION

a = [1, 2, 3, 4]
b = [5, 6, 7, 8]

(a) Make a 1d array from list a[ ].
(b) Make a 2d array from lists a[ ] and b[ ].

Answer 2

array1 = np.array(a)
array2 = np.array([a,b])  
array2 = np.array((a,b))

Question 3

DIMENSIONS

a = np.array([ [1,2,3], [4,5,6] ])

How can you check the dimensións and size of the array?

Answer 3

a = np.array([ [1,2,3], [4,5,6] ])

print(a.ndim)
print(a.shape)
print(a.size)
print(a.itemsize)

Question 4

DATA TYPE

a = np.array([ [1,2,3], [4,5,6] ])

(a) How can you check the data type of the new array?
(b) Can you change the type from int to float?

Answer 4

a = np.array([ [1,2,3], [4,5,6] ])
print(a.dtype)

a = a.astype(‘float32’)
print(a.dtype)

Question 5

ARRAY DEFINITION

Create a 10x10 array with numbers from 1 to 100.

Answer 5

import numpy as np

b = np.arranage(1,101)
b.reshape(10,10)

Question 6

ARRAY DEFINITION

(a) Create a 5x3 array of zeros.
(b) Reshape the new array to be 3x5.
(c) Change the type of the array from float to int

Answer 6

import numpy as np

z = np.zeros((5,3))
print(z)

z = z.reshape((3,5))
print(z)

z = z.astype(‘int16’)
print(z)

Question 7

ARRAY DEFINITION

(a) Create a 5x3 array of ones (dtype -> int).
(b) Reshape the new array to be 3x5.

Answer 7

import numpy as np

z = np.ones((5,3), dtype = np.int32)
print(z)

z = z.reshape((3,5))
print(z)

Question 8

ARRAY INDEXING

a = np.array( [ [5,7,8,6], [5,9,7,8], [5,4,9,2], [5,3,9,6] ] )

(a) Select the second row (index = 1).
(b) Select the second and fourth colums.
(c) Get the rows where there is at least one mulltiple of three.

Answer 8

(a) a[1]
(b) a[ [1, 3] ]
(c) a[ np.any( a % 3 == 0, axis = 1) ]

Question 9

SUM

a = np.array( [ [5,7,8], [5,9,7], [5,4,9], [6,3,9] ] )

(a) Compute the sum of all the elements in the array.
(b) Compute the sum of the elements in each row.
(c) Compute the sum of the elements in each column.

Answer 9

a = np.array( [ [5,7,8], [5,9,7], [5,4,9], [6,3,9] ] )

# (a) ---------------
a.sum( )
# (b) ---------------
a.sum(axis = 1)
# (c) ---------------
a.sum(axis = 0)

Question 10

MAX, MIN, MEAN

a = np.array( [ [5,7,8], [5,9,7], [5,4,9], [6,3,9] ] )

How can you compute the min, max and mean of the elements in the array?

Answer 10

array-wise

a. max()
a. min()
a. mean()

Question 11

COPYNG SHAPE AN DIMS

import numpy as np
a = np.arange(1,101).rashape(10,10)

(a) Create a zeros array with the same shape and dtype of a.
(b) Create a full array with the same shape and dtype of a.

Answer 11

import numpy as np
a = np.arange(1,101).rashape(10,10)

z = np.zeros_like(a)
f = np.full_like(a, 5)

Question 12

ARITHMETIC

prices = np.array([100, 25, 40, 33, 60])

(a) Compute the taxes corresponding to every price, consider a 16% tax.
(b) Compute the total with and without taxes

Answer 12

taxes = prices * 1.16
total_1 = prices.sum() 

taxes_inc = prices + taxes
total_2 = taxes_inc.sum()

Question 13

ROUNDING

prices = np.array([50, 53.5, 58.9, 62.57])

(a) Round prices to the next integer
(b) Round prices to the previous integer
(c) Round prices to the nearest intger

Answer 13

r1 = np.ceil(prices)
r2 = np.floor(prices)
r3 = np.rint(prices)

Question 14

COMPARING ARRAYS

prices_1 = np.array([50, 53.5, 58.9, 62.57])
prices_2 = np.array([48, 55.5, 57, 63])

prices_1 and prices_2 correspond to prices of the same products (product 1:4) in two different places. Find the cheapest and most expensive price for every product.

(Bonus: How would you solve the same problem if there were three or more lists of prices? — prices_3 = np.array([47, 60.5, 56, 67]) — )

Answer 14

cheapest = np,minimum(prices_1, prices_2)
most_expensive = np.maximum(prices_1, prices2)

Bonus: use vstack() to create a prices array with one row per list, then use prices.min(axis = 0) / prices.max(axis = 0) .

Question 15

ARRAY RESHAPING

# Consider the 2x2x3 array
a = np.array([[[1,2,3],[4,5,6]],[[8,9,10],[12,13,14]]])

How can you get a 1xn array from a?

Answer 15

a = a.flatten()

Question 16

ARRAY SORTING

arr = np.random.randn(3, 5)

How can you reorder the array by its first row?

Answer 16

arr = np.random.randn(3, 5)

result = array[ : , arr[0].argsort() ]

Question 17

ARRAY CREATION

(a) Create a 3x5 array with random numbers
(b) Create a 3x5 array with random ints between 0 and 100

Answer 17

arr_1 = np.random.randn(3,5)
arr_2 = np.random.randint(100, size = (3,5))

Question 18

BOOLEAN MASKING

arr = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])

How can you get all the entries divisible by 3?

Answer 18

arr[arr % 3 ==0]

Question 19

SLICING

arr = np.random.randn(10,10)

Create an arr2 with all the elements in the midle of arr.

(“in the middle” means any element which is not in the edges of the array)

Answer 19

arr2 = arr[1:-1,1:-1]

Question 20

BOOLEAN SLICING

arr = np.random.randint(10, size = (10,10))

Get all the rows in which at least 2 entries are equal to 1.

Answer 20

arr = np.random.randint(10, size = (10,10))

one_counter = (arr == 1).sum(axis = 1)
indexer = one_counter >= 2

arr[indexer]

Question 21

RESHAPING

a = np.zeros( ( 10,10) )

reshape array a to be a -row vector-, -column vector-, 2 x k matrix.

Answer 21

b = a.reshape(-1,1)
c = a.reshaper(1,-1)
d = a.reshape(2,-1)