Numpy | Basics | Priority

Question 1

Pearson’s correlation r from a pandas dataframe.

Answer 1

corrmat = np.corrcoef(df.values.T)
corrmat = np.corrcoef(df.values, rowvar=False)

Machine Learning with PyTorch and Scikit-Learn Chapter 9 p277

Question 2

Function to count the number of occurrences of each value in array.

Answer 2

np.bincount()

Machine Learning with PyTorch and Scikit-Learn Chapter 3 p55

Question 3

Function to modify a sequence in-place by shuffling its contents.

Answer 3

… indices = np.arange(X.shape[0])
… np.random.shuffle(indices)

Raschka MLWPT Chapter 11 Implementing a Multilayer Artificial Neural Network from Scratch

Question 4

Create and use a random number generator.

Answer 4

“rng = np.random.RandomState(random_seed)
self.weight_h = rng.normal(
loc=0.0, scale=0.1, size=(num_hidden, num_features))”

Raschka MLWPT Chapter 11 Implementing a Multilayer Artificial Neural Network from Scratch

Question 5

Example of how to convert an array’s data type.

Answer 5

example: ary2d.astype(np.int32)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.1: NumPy Basics

Question 6

Example of how to initialize an array with 0s or 1s and a specific data type.

Answer 6

np.ones((3, 4), dtype=np.int)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.2: NumPy Array Construction and Indexing

Question 7

Example of how to create a particular number of evenly spaced values in a specified interval.

Answer 7

np.linspace(6., 15., num=10)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.2: NumPy Array Construction and Indexing

Question 8

Example of how to add 1 to every element in a 2D array.

Answer 8

ary + 1

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.3: NumPy Array Math and Universal Functions

Question 9

Examples of how to get the square root and exponentiated versions of an array.

Answer 9

np.sqrt(ary)

;

ary**2 OR np.power(ary, 2)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.3: NumPy Array Math and Universal Functions

Question 10

Idioms to sum out rows and sum out columns.

Answer 10

np.sum(axis=1) # i.e. sum along columns to get row sums; np.sum(axis=0) # i.e. sum along rows to get column sums

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.3: NumPy Array Math and Universal Functions

Question 11

Function to return a sorted copy of an array; Function to return the indices that would sort an array.

Answer 11

np.sort(); np.argsort()

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.3: NumPy Array Math and Universal Functions

Question 12

Slicing creates what? What is a way to avoid this behavior?

Answer 12

Views. ary.copy()

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.5: NumPy Advanced Indexing – Memory Views and Copies

Question 13

Fancy indexing always returns what? Is boolean indexing fancy indexing?

Answer 13

A copy of an array. Yes.

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.5: NumPy Advanced Indexing – Memory Views and Copies

Question 14

Example of how to select array elements that are greater than 3 and divisible by 2.

Answer 14

ary[(ary > 3) & (ary % 2 == 0)]

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.5: NumPy Advanced Indexing – Memory Views and Copies

Question 15

What is the difference between flatten() and ravel()? What is one more way to flatten an array?

Answer 15

flatten: copy; ravel: view; ary.reshape(-1)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.7: Reshaping NumPy Arrays

Question 16

What is the difference between flatten() and ravel()? What is one more way to flatten an array?

Answer 16

flatten: copy; ravel: view; ary.reshape(-1)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.7: Reshaping NumPy Arrays

Question 17

Idioms to stack along rows and along columns.

Answer 17

np.concatenate((ary, ary), axis=0); np.concatenate((ary, ary), axis=1)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.7: Reshaping NumPy Arrays

Question 18

Idiom to count how many elements in an array meet a certain condition.

Answer 18

mask = ary > 2; mask.sum()

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.8: NumPy Comparison Operators and Masks

Question 19

Function to assign values to specific elements in an array.

Answer 19

Example: np.where(ary, 1, 0)

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.8: NumPy Comparison Operators and Masks

Question 20

3 ways to reshape a one-dimensional array into a two-dimensional one.

Answer 20

row_vector..reshape(-1, 1); row_vector[:, np.newaxis]; row_vector[:, None]

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.9: Linear Algebra with NumPy

Question 21

3 ways to compute the dot product of a row_vector.

Answer 21

np.matmul(row_vector, row_vector); np.dot(row_vector, row_vector); row_vector @ row_vector

Scientific Computing in Python: Introduction to NumPy and Matplotlib 4.9: Linear Algebra with NumPy