Data book - Numpy

Question 1

How do you make a 2x3x2 empty np array?

Answer 1

np.empty((2,3,2))

Question 2

What is np’s equivalent to the range function?

Answer 2

Np.arange(n)

Out: array([0, 1, 2, …, n])

Question 3

How do you take an existing array a and make a new array with the same shape but all 1’s?

Answer 3

Np.ones_like(a)

Question 4

How do you make a square n x n matrix?

Answer 4

Np.eye(n)

Question 5

How do you change an np dtype from one type to another?

Answer 5

array_to_change.astype(dtype)

E.g

a = np.array([‘1’])
a = a.astype(np.float64)

Question 6

Does slicing an ndarray copy that section and return the copy, or zoom into that section?

Answer 6

Zoom in

Question 7

How do you make a copy of an ndarray slice?

Answer 7

arr = np.array([…])

barr = arr[5:8].copy()

Question 8

How do you select the top-left most element of the 2d array arr2d?

Answer 8

arr2d[0,0]

Question 9

How do you generate randomly distributed data 7 rows by 4 columns?

Answer 9

Np.random.randn(7, 4)

Question 10

How do you do both a boolean and an array slice in one line?

Answer 10

data[names==‘Bob’, 2:]

Question 11

How do you select everything except ‘Bob’ in an ndarray?

Answer 11

names != ‘Bob’

Or

data[-(names == ‘Bob’)]

Question 12

How do you use and/or logic in ndarray slicing?

Answer 12

Mask = (names == ‘Bob’) |/& (names == ‘Will’)

Question 13

What is fancy indexing?

Answer 13

Indexing with integer arrays

E.g arr[4, 3, 0, 6] pops out those rows in that order

Question 14

Say you call arr[[1, 5, 7, 2], [0, 3, 1, 2]] on a big 2D ndarray. What happens?

Answer 14

Outputs a 1D array with elements selected by [[rows], [columns]] in the input.

Question 15

How do you convert two 1D ndarrays to an indexer that selects a square region?

Answer 15

arr[np.ix_([], [])]

Question 16

How do you transpose?

Answer 16

arr.T

Question 17

How to compute the dot product of two arrays?

Answer 17

np.dot(a, b)

Question 18

How do you swap axis in general(e.g inn dimensions)?

Answer 18

arr.swapaxes(n1, n2, …, n)

Question 19

What is a universal function in numpy?

Answer 19

Funct that performs elementwise operations on data in ndarrays

Question 20

What are examples of ufuncs?

Answer 20

np. sqrt(arr)
np. exp(arr)

(Unary funcs - take one array)

Question 21

x = np.array([1, 3, 3,])

y = np.array([3, 1, 1])

What does np.maximum(x, y) output?

Answer 21

array([3, 3, 3)]

Also its a binary ufunc

Question 22

What does np.modf(arr) do?

Answer 22

Returns fractional and integral parts of an array

Question 23

What ufunc gives booleans corresponding to whether an element is NaN?

Answer 23

Np.isnan

Question 24

How do you make a 3 by 5 array filled with the integers 3.14?

Answer 24

np.full((3, 5), 3.14, dtype=float)

Question 25

How do you make a 3 by 3 array filled with random integers between 0 and 1?

Answer 25

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

Question 26

How do you make a 3 by 3 array with random values from a normal distribution with mean 0 and SD 1?

Answer 26

np.random.normal(0, 1, (3, 3))

Question 27

How do you make a 3 by 3 array with random values between 0 and 10?

Answer 27

np.random.randint(0, 10, (3, 3))

Question 28

How do you make a 3x3 identity matrix?

Answer 28

np.eye(3)

Question 29

What 3 methods are generally used for array concatenation?

Answer 29

np. concatenate
np. vstack
np. hstack

Question 30

How do you stack elements on the 3rd axis?

Answer 30

np.dstack

Question 31

What 3 methods are used for splitting?

Answer 31

np. split
np. hsplit
np. vsplit

Question 32

How do you apply a ufunc multiple times on an array x to remove a dimension?

Answer 32

np.ufunc.reduce(x)

Question 33

How do you apply a ufunc multiple times to an array x cumulatively, and save the result to an array?

Answer 33

np.ufunc.accumulate(x)

Question 34

What numpy function could you use to make an easy multiplication table in one line from an array x?

Answer 34

np.multiply.outer(x)