Introduction to NumPy

Overview

Learning outcomes

• Students will be able to explain what NumPy is and why it is used for numerical work.
• Students will be able to state the characteristics of an array and compare a NumPy array with a Python list.
• Students will be able to import NumPy using the alias np.
• Students will be able to create 1-D arrays from lists and 2-D arrays from nested lists using np.array().
• Students will be able to predict the data type of an array created from a list of mixed values (type promotion).
• Students will be able to set the data type of a new array using the dtype argument.
• Students will be able to identify and correct common errors while creating arrays.
• Students will be able to use basic array attributes such as shape, ndim and dtype to check an array (enrichment).

Lesson plan

Notes

import numpy as np

print("NumPy version:", np.__version__)
marks = np.array([78, 85, 62, 90])
print(marks)
print(type(marks))

NumPy version: 2.5.4
[78 85 62 90]
<class 'numpy.ndarray'>

import numpy as np

price_list = [20, 35, 50]
price_arr = np.array([20, 35, 50])

print("list * 2  :", price_list * 2)     # repeats the list
print("array * 2 :", price_arr * 2)      # doubles every element
print("array + 5 :", price_arr + 5)
print("array / 5 :", price_arr / 5)

list * 2  : [20, 35, 50, 20, 35, 50]
array * 2 : [ 40  70 100]
array + 5 : [25 40 55]
array / 5 : [ 4.  7. 10.]

import numpy as np

runs = [45, 12, 78, 33]            # a Python list
a = np.array(runs)                 # 1-D array from the list
print(a)
print(a.dtype)

temps = np.array([31.5, 29.0, 33.2])   # list written directly inside array()
print(temps)
print(temps.dtype)

[45 12 78 33]
int64
[31.5 29.  33.2]
float64

import numpy as np

a = np.array([4, 7.5, 9])          # int and float mixed
print(a, a.dtype)

b = np.array([1, 0, True])         # bool with int
print(b, b.dtype)

c = np.array([5, -7.4, 'R', 3])    # a string is present
print(c)
print(c.dtype)

[4.  7.5 9. ] float64
[1 0 1] int64
['5' '-7.4' 'R' '3']
<U32

import numpy as np

# marks of 3 students in 4 subjects: each inner list is one row
marks = [[78, 85, 62, 90],
         [55, 67, 71, 80],
         [92, 88, 95, 91]]
m = np.array(marks)
print(m)
print("Rows and columns:", m.shape)

mixed = np.array([[1.5, 2], [3, 4]])   # ints promoted to float
print(mixed)

[[78 85 62 90]
 [55 67 71 80]
 [92 88 95 91]]
Rows and columns: (3, 4)
[[1.5 2. ]
 [3.  4. ]]

import numpy as np

a = np.array([10, 20, 30], dtype=float)
print(a, a.dtype)

b = np.array([2.9, 7.1, -3.6], dtype=int)    # decimal part is cut off
print(b, b.dtype)

c = np.array([1, 2, 3], dtype=str)
print(c, c.dtype)

[10. 20. 30.] float64
[ 2  7 -3] int64
['1' '2' '3'] <U1

import numpy as np

a = np.array([3, 6, 9])
back = a.tolist()          # array -> Python list
print(back, type(back))

t = np.array((11, 22, 33))  # a tuple also works
print(t)

[3, 6, 9] <class 'list'>
[11 22 33]

import numpy as np

try:
    a = np.array(1, 2, 3)        # WRONG: values not inside a list
except TypeError as e:
    print("TypeError:", e)

a = np.array([1, 2, 3])          # CORRECT: one list argument
print(a)

TypeError: array() takes from 1 to 2 positional arguments but 3 were given
[1 2 3]

import numpy as np

try:
    r = np.array([[1, 2, 3], [4, 5]])   # rows of unequal length
except ValueError:
    print("ValueError: rows must have the same length")

ValueError: rows must have the same length

import numpy as np

a = np.array([10, 20, 30, 40])
m = np.array([[1.5, 2.0, 3.5],
              [4.0, 5.5, 6.0]])

print("ndim    :", a.ndim, m.ndim)
print("shape   :", a.shape, m.shape)
print("size    :", a.size, m.size)
print("dtype   :", a.dtype, m.dtype)
print("itemsize:", a.itemsize, m.itemsize)

ndim    : 1 2
shape   : (4,) (2, 3)
size    : 4 6
dtype   : int64 float64
itemsize: 8 8

import numpy as np

print(np.zeros((2, 3)))
print(np.ones((2, 2), dtype=int))
print(np.arange(5))
print(np.arange(10, 31, 5))
print(np.linspace(0, 1, 5))

[[0. 0. 0.]
 [0. 0. 0.]]
[[1 1]
 [1 1]]
[0 1 2 3 4]
[10 15 20 25 30]
[0.   0.25 0.5  0.75 1.  ]

import numpy as np

m = np.array([[78, 85, 62, 90],
              [55, 67, 71, 80],
              [92, 88, 95, 91]])

print(m[0, 2])        # row 0, column 2
print(m[2])           # whole row 2
print(m[-1, -1])      # last row, last column
print(m[0:2, 1:3])    # rows 0-1, columns 1-2
print(m[:, 0])        # first column of every row
a = np.array([5, 10, 15, 20, 25, 30])
print(a[1:5:2])
print(a[::-1])

62
[92 88 95 91]
91
[[85 62]
 [67 71]]
[78 55 92]
[10 20]
[30 25 20 15 10  5]

import numpy as np

a = np.array([[2, 4], [6, 8]])
b = np.array([[1, 2], [3, 4]])

print(a + b)
print(a * b)          # element-wise, not matrix product
print(a ** 2)
print(a // b)
print(a @ b)          # matrix multiplication
print(a.T)            # transpose
print(np.sort(np.array([40, 10, 30, 20])))

[[ 3  6]
 [ 9 12]]
[[ 2  8]
 [18 32]]
[[ 4 16]
 [36 64]]
[[2 2]
 [2 2]]
[[14 20]
 [30 44]]
[[2 6]
 [4 8]]
[10 20 30 40]

import numpy as np

x = np.array([[1, 2], [3, 4]])
y = np.array([[5, 6]])
print(np.concatenate((x, y), axis=0))

a = np.arange(1, 7)
r = a.reshape(2, 3)
print(r)

parts = np.split(np.arange(9), 3)
print(parts[0], parts[1], parts[2])

[[1 2]
 [3 4]
 [5 6]]
[[1 2 3]
 [4 5 6]]
[0 1 2] [3 4 5] [6 7 8]

import numpy as np

m = np.array([[78, 85, 62],
              [55, 67, 71],
              [92, 88, 95]])

print("Max         :", m.max())
print("Min per col :", m.min(axis=0))
print("Sum per row :", m.sum(axis=1))
print("Mean        :", round(m.mean(), 2))
print("Std dev     :", round(m.std(), 2))

Max         : 95
Min per col : [55 67 62]
Sum per row : [225 193 275]
Mean        : 77.0
Std dev     : 13.25

import numpy as np

# create a small CSV file first (normally it already exists)
with open("scores.csv", "w") as f:
    f.write("RollNo,Maths,Science,IP\n")
    f.write("1,78,85,90\n")
    f.write("2,55,67,80\n")
    f.write("3,92,88,95\n")

data = np.loadtxt("scores.csv", delimiter=",", skiprows=1, dtype=int)
print(data)
print("IP column:", data[:, 3])

with open("gaps.csv", "w") as f:
    f.write("1,78,85\n2,,67\n")
g = np.genfromtxt("gaps.csv", delimiter=",", filling_values=-1)
print(g)

[[ 1 78 85 90]
 [ 2 55 67 80]
 [ 3 92 88 95]]
IP column: [90 80 95]
[[ 1. 78. 85.]
 [ 2. -1. 67.]]

Key terms

NumPy

Numerical Python: a library for fast numerical computing with arrays.

Library / package

A ready-made collection of modules and functions that can be imported into a program.

Array

An ordered collection of elements of the same data type stored contiguously in memory.

ndarray

NumPy's n-dimensional array object; the official name of a NumPy array.

np.array()

NumPy function that creates an array from a list, tuple or nested list.

1-D array

An array with a single row of elements; one axis.

2-D array

An array with rows and columns, created from a nested list; two axes.

Nested list

A list whose elements are themselves lists.

dtype

The data type of the elements of an array, e.g. int64, float64, <U32.

Type promotion (upcasting)

Automatic conversion of all elements to the most general type present in the list.

Contiguous memory

Elements stored next to each other in memory without gaps.

Element-wise operation

An operation applied to each element of an array separately.

Zero-based indexing

Numbering positions from 0, so the first element has index 0.

Alias

A short alternative name for an imported module, e.g. np for numpy.

shape

Attribute giving the size of an array in each dimension, as a tuple (enrichment).

Axis

A dimension of an array; axis 0 runs down the rows, axis 1 across the columns (enrichment).

MCQs

Tap an option to check it, or use Show answers above to reveal them all.

1. NumPy stands for:

   1. Number Python
   2. New Python
   3. Numeric Pandas
   4. Numerical Python
   Answer

   (d) Numerical Python

   NumPy is short for Numerical Python.

2. Which statement correctly imports NumPy with its usual alias?

   1. import numpy as np
   2. import NumPy as np
   3. include numpy as np
   4. from np import numpy
   Answer

   (a) import numpy as np

   The module name is lowercase numpy; np is the conventional alias.

3. Which function creates a NumPy array from a list?

   1. np.list()
   2. np.create()
   3. np.array()
   4. np.ndarray.make()
   Answer

   (c) np.array()

   np.array() converts a list, tuple or nested list into an ndarray.

4. Which of these is TRUE for every NumPy array?

   1. It can hold elements of different data types
   2. It is part of core Python
   3. All its elements have the same data type
   4. Its index starts from 1
   Answer

   (c) All its elements have the same data type

   An array is homogeneous; indexing starts from 0 and it comes from the NumPy library.

5. What is the output?

   import numpy as np
print(np.array([2, 4.5, 6]))

   1. [2 4.5 6]
   2. [2, 4.5, 6]
   3. [2. 4.5 6. ]
   4. Error
   Answer

   (c) [2. 4.5 6. ]

   Int and float are mixed, so every element is promoted to float and printed with a decimal point. Elements are separated by spaces.

6. What will be the dtype of np.array([3, 'a', 4.5])?

   1. Unicode string, e.g. <U32
   2. float64
   3. int64
   4. object
   Answer

   (a) Unicode string, e.g. <U32

   If even one string is present, every element is converted to a string.

7. If L = [5, 10] and A = np.array([5, 10]), what are L * 2 and A * 2?

   1. Both [10, 20]
   2. [10, 20] and [5 10 5 10]
   3. Both cause an error
   4. [5, 10, 5, 10] and [10 20]
   Answer

   (d) [5, 10, 5, 10] and [10 20]

   * repeats a list but multiplies each array element.

8. What is the output of print(np.array([3.9, -2.7], dtype=int))?

   1. [4 -3]
   2. [3 -3]
   3. [4 -2]
   4. [3 -2]
   Answer

   (d) [3 -2]

   Converting to int truncates the decimal part: 3.9 becomes 3 and −2.7 becomes −2.

9. Which statement creates a 2-D array with 2 rows and 3 columns?

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

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

   A nested list with 2 inner lists of 3 elements each gives 2 rows and 3 columns. Option C gives 3 rows and 2 columns.

10. Elements of a NumPy array are stored:

    1. Contiguously in memory
    2. At random locations
    3. Only on disk
    4. As separate objects with their own types
    Answer

    (a) Contiguously in memory

    Contiguous storage is one reason arrays are faster than lists.

11. Which statement about lists and NumPy arrays is FALSE?

    1. Arrays support element-wise operations
    2. Arrays usually take less memory than lists for the same numbers
    3. Lists can store mixed data types
    4. A list must be imported from the NumPy library
    Answer

    (d) A list must be imported from the NumPy library

    Lists are built into Python; only arrays need NumPy.

12. What does np.array([True, 0, 5]) produce?

    1. [True False True]
    2. ['True' '0' '5']
    3. [1 0 5]
    4. Error
    Answer

    (c) [1 0 5]

    bool mixed with int is promoted to int, so True becomes 1.

13. Which command installs NumPy?

    1. pip install numpy
    2. python numpy install
    3. install numpy
    4. import numpy
    Answer

    (a) pip install numpy

    pip install numpy is run once at the command prompt; import numpy is used inside a program.

14. The index of the last element of np.array([7, 14, 21, 28, 35]) is:

    1. 5
    2. 35
    3. -5
    4. 4
    Answer

    (d) 4

    With zero-based indexing, 5 elements have indexes 0 to 4. (Its negative index is −1.)

15. (Enrichment) For m = np.array([[1, 2, 3], [4, 5, 6]]), what is m.shape?

    1. (2, 3)
    2. (3, 2)
    3. 6
    4. (2,)
    Answer

    (a) (2, 3)

    2 rows and 3 columns.

16. (Enrichment) For the same m, what is m.size?

    1. 2
    2. 3
    3. 6
    4. 5
    Answer

    (c) 6

    size is the total number of elements: 2 x 3 = 6.

17. (Enrichment) What is the output of print(np.arange(2, 11, 3))?

    1. [2 5 8 11]
    2. [2 5 8]
    3. [3 6 9]
    4. [2 4 6 8 10]
    Answer

    (b) [2 5 8]

    Start at 2, step 3, and stop before 11: 2, 5, 8.

18. (Enrichment) Which function creates an array filled with 0s?

    1. np.empty()
    2. np.zeros()
    3. np.null()
    4. np.ones()
    Answer

    (b) np.zeros()

    np.zeros(shape) fills an array of the given shape with 0.0.

19. (Enrichment) For a = np.array([10, 20, 30, 40, 50]), what is a[1:4]?

    1. [20 30 40]
    2. [10 20 30 40]
    3. [20 30 40 50]
    4. [10 20 30]
    Answer

    (a) [20 30 40]

    The slice includes index 1 up to, but not including, index 4.

20. (Enrichment) Which function reads a CSV file with missing values into an array?

    1. np.loadtxt()
    2. np.genfromtxt()
    3. np.savetxt()
    4. np.readcsv()
    Answer

    (b) np.genfromtxt()

    genfromtxt() can fill missing values; loadtxt() expects complete data.

Assertion–Reason

1. Assertion (A): np.array([4, 5.5, 6]) creates an array of floating-point numbers.

   Reason (R): All elements of a NumPy array must be of the same data type, so integers are promoted to float.

   Answer

   (A)

   Both are true and R explains why the integers become floats.

2. Assertion (A): Arithmetic on NumPy arrays is generally faster than the same work done on lists with a loop.

   Reason (R): NumPy array elements are stored contiguously and all have the same type.

   Answer

   (A)

   Contiguous, same-type storage is what allows NumPy's fast element-wise operations.

3. Assertion (A): np.array(10, 20, 30) creates an array of three elements.

   Reason (R): The values to be stored must be passed to np.array() as a single list or tuple.

   Answer

   (D)

   A is false; this statement raises an error. R is true and gives the correct usage.

4. Assertion (A): A Python list can store an integer, a string and a float together.

   Reason (R): A NumPy array can also store an integer, a string and a float without changing them.

   Answer

   (C)

   A is true. R is false: in an array the numbers would be converted to strings.

5. Assertion (A): np.array([2.8, 5.9], dtype=int) gives [2 5].

   Reason (R): Converting a float to int in NumPy rounds it to the nearest whole number.

   Answer

   (C)

   A is true because the decimal part is truncated. R is false: there is no rounding.

Short answer

1. 2 marks

   What is NumPy? Write the statement to import it with the alias np.

   Answer

   NumPy (Numerical Python) is a Python library for scientific and numerical computing. Its main object is the n-dimensional array (ndarray), which stores same-type data and supports fast element-wise operations.

   import numpy as np

2. 2 marks

   State any two differences between a Python list and a NumPy array.

   Answer

   • A list can hold elements of different data types; all elements of an array are of the same type.
   • A list does not support element-wise arithmetic (L * 2 repeats it); an array does (A * 2 doubles each element).

   (Also accepted: arrays are stored contiguously and use less memory; a list is built into Python while an array needs NumPy.)

3. 2 marks

   Write a program to create a NumPy array from the list [88, 92, 79, 95] and print it with its data type.

   Answer

   Use np.array() and the dtype attribute.

   import numpy as np
   
   marks = [88, 92, 79, 95]
   arr = np.array(marks)
   print(arr)
   print(arr.dtype)

   Output

   [88 92 79 95]
   int64

4. 2 marks

   Predict the output:

   import numpy as np
a = np.array([1, 2, 3.5])
b = np.array([1, 'two', 3])
print(a)
print(b)

   Answer

   a has a float, so all values become float. b has a string, so all values become strings.

   import numpy as np
   a = np.array([1, 2, 3.5])
   b = np.array([1, 'two', 3])
   print(a)
   print(b)

   Output

   [1.  2.  3.5]
   ['1' 'two' '3']

5. 2 marks

   How is a 2-D array created from a list? Give an example of a 2 x 2 array.

   Answer

   Pass a nested list to np.array(); each inner list becomes one row, and all inner lists must have the same length.

   import numpy as np
   m = np.array([[10, 20], [30, 40]])
   print(m)

   Output

   [[10 20]
    [30 40]]

6. 2 marks

   What is the purpose of the dtype argument in np.array()? Give an example.

   Answer

   dtype sets the data type of the elements; every value in the list is converted to that type.

   import numpy as np
   a = np.array([1, 2, 3], dtype=float)
   print(a)

   Output

   [1. 2. 3.]

7. 2 marks

   Find the error and correct it:

   import Numpy as np
a = np.array(5, 10, 15)

   Answer

   1. The module name must be lowercase: import numpy as np.
   2. Values must be passed as a single list: a = np.array([5, 10, 15]).
8. 2 marks

   Why are NumPy arrays faster and smaller in memory than lists?

   Answer

   Array elements all have one data type and are stored contiguously in memory, so NumPy does not keep separate type information for each element and can process the whole block at once in fast compiled code. List elements are separate Python objects stored at different places, each carrying its own type information.

9. 2 marks

   What will be the output?

   import numpy as np
L = [3, 6, 9]
A = np.array(L)
print(L + L)
print(A + A)

   Answer

   + joins two lists, but adds two arrays element by element.

   import numpy as np
   L = [3, 6, 9]
   A = np.array(L)
   print(L + L)
   print(A + A)

   Output

   [3, 6, 9, 3, 6, 9]
   [ 6 12 18]

10. 2 marks

    Explain type promotion in NumPy with one example.

    Answer

    When a list has values of different types, NumPy converts all elements to the most general type present (bool -> int -> float -> string), because an array can hold only one data type. Example: np.array([5, 2.5]) gives [5. 2.5]; the integer 5 is promoted to float 5.0.

11. 2 marks

    (Enrichment) Write the output of:

    import numpy as np
m = np.array([[4, 5, 6], [7, 8, 9]])
print(m.ndim, m.shape, m.size)

    Answer

    2 dimensions, 2 rows x 3 columns, 6 elements.

    import numpy as np
    m = np.array([[4, 5, 6], [7, 8, 9]])
    print(m.ndim, m.shape, m.size)

    Output

    2 (2, 3) 6

12. 2 marks

    (Enrichment) Differentiate between np.zeros((2, 3)) and np.ones((3, 2)).

    Answer

    np.zeros((2, 3)) creates a 2-row, 3-column array filled with 0.0. np.ones((3, 2)) creates a 3-row, 2-column array filled with 1.0. Both give float arrays by default.

Long answer

1. 3 marks

   Compare a Python list and a NumPy array on any three points, giving a small code example for one difference.

   Answer

   1. Data type: a list can mix types ([1, 'a', 2.5]); an array holds one type, and mixed values are promoted.
   2. Operations: lists have no element-wise arithmetic; arrays apply + - * / to every element.
   3. Memory and speed: array elements are stored contiguously with a single type, so arrays use less memory and are faster; list elements are separate objects.

   Example of point 2:

   import numpy as np
   L = [2, 4, 6]
   A = np.array(L)
   print(L * 3)
   print(A * 3)

   Output

   [2, 4, 6, 2, 4, 6, 2, 4, 6]
   [ 6 12 18]

2. 3 marks

   Write a program that stores the runs scored by a team in 5 overs, [8, 12, 4, 15, 9], in a NumPy array. Print the array, its data type, and the array after adding 2 bonus runs to every over.

   Answer

   import numpy as np
   
   runs = [8, 12, 4, 15, 9]
   arr = np.array(runs)
   print("Runs   :", arr)
   print("Type   :", arr.dtype)
   print("Bonus  :", arr + 2)

   Output

   Runs   : [ 8 12  4 15  9]
   Type   : int64
   Bonus  : [10 14  6 17 11]

3. 3 marks

   Predict the output of each statement and give a reason:

   (a) np.array([1, 0, 2.0])
   (b) np.array([False, 7])
   (c) np.array([4, '4'])

   Answer

   (a) Int and float are mixed, so all elements become float.
   (b) bool and int are mixed, so all become int (False -> 0).
   (c) A string is present, so all become strings.

   import numpy as np
   print(np.array([1, 0, 2.0]))
   print(np.array([False, 7]))
   print(np.array([4, '4']))

   Output

   [1. 0. 2.]
   [0 7]
   ['4' '4']

4. 3 marks

   The marks of 3 students in Physics, Chemistry and IP are given below. Write a program to create a 2-D NumPy array from a nested list, then print the array, the number of rows and columns, and the same array converted to float.

   Aarav: 72, 65, 88
   Ishita: 81, 79, 94
   Kabir: 58, 62, 70

   Answer

   import numpy as np
   
   marks = [[72, 65, 88],
            [81, 79, 94],
            [58, 62, 70]]
   m = np.array(marks)
   print(m)
   print("Rows:", m.shape[0], "Columns:", m.shape[1])
   print(np.array(marks, dtype=float))

   Output

   [[72 65 88]
    [81 79 94]
    [58 62 70]]
   Rows: 3 Columns: 3
   [[72. 65. 88.]
    [81. 79. 94.]
    [58. 62. 70.]]

5. 5 marks

   (a) What is an array? State its three characteristics. (b) Why is NumPy needed when Python already has lists? (c) Write the code to create a 1-D array of the daily rainfall (in mm) [12.5, 0, 33.2, 7] and explain the dtype of the result.

   Answer

   (a) An array is an ordered collection of elements under one name, where (i) every element has the same data type, (ii) elements are stored contiguously in memory, and (iii) each element is accessed by a zero-based index.

   (b) Lists are general-purpose and slow for numerical work: they cannot do element-wise arithmetic, take more memory and need loops. NumPy arrays are compact, fast, support element-wise operations without loops, and form the base for Pandas and Matplotlib.

   (c) The list has a float, so the integers 0 and 7 are promoted and the dtype is float64.

   import numpy as np
   rain = np.array([12.5, 0, 33.2, 7])
   print(rain)
   print(rain.dtype)

   Output

   [12.5  0.  33.2  7. ]
   float64

6. 5 marks

   Explain the use of np.array() to create (i) a 1-D array, (ii) a 2-D array, (iii) an array of a chosen data type. Write one example of each with its output. Also state one error a student may make while creating a 2-D array.

   Answer

   (i) Pass a flat list; (ii) pass a nested list whose inner lists all have the same length; (iii) pass dtype=. A common error: inner lists of different lengths, e.g. [[1, 2], [3]], which raises a ValueError in current NumPy.

   import numpy as np
   one = np.array([5, 10, 15])
   two = np.array([[1, 2], [3, 4]])
   typed = np.array([7, 8, 9], dtype=float)
   print(one)
   print(two)
   print(typed)

   Output

   [ 5 10 15]
   [[1 2]
    [3 4]]
   [7. 8. 9.]

7. 3 marks

   (Enrichment) Given t = np.array([[30, 32, 31], [27, 29, 28]]) (temperatures of two cities over three days), write statements to print: (a) the shape, (b) the temperature of the second city on the third day, (c) the first day's readings for both cities, (d) the maximum temperature in each row.

   Answer

   import numpy as np
   t = np.array([[30, 32, 31], [27, 29, 28]])
   print(t.shape)
   print(t[1, 2])
   print(t[:, 0])
   print(t.max(axis=1))

   Output

   (2, 3)
   28
   [30 27]
   [32 29]

Case-based

Practicals

import numpy as np

# weekly sales (in Rs) of a school canteen, Monday to Saturday
sales = [4200, 3850, 4600, 3975, 5100, 2800]
arr = np.array(sales)

print("Array      :", arr)
print("Type       :", type(arr))
print("Data type  :", arr.dtype)
print("Dimensions :", arr.ndim)
print("Elements   :", arr.size)
print("Sales + 10%:", arr * 1.1)

Array      : [4200 3850 4600 3975 5100 2800]
Type       : <class 'numpy.ndarray'>
Data type  : int64
Dimensions : 1
Elements   : 6
Sales + 10%: [4620.  4235.  5060.  4372.5 5610.  3080. ]

import numpy as np

# rows: students, columns: Term 1, Term 2, Term 3 marks
marks = [[67, 72, 80],
         [88, 91, 85],
         [54, 60, 58],
         [75, 70, 79]]
m = np.array(marks)

print(m)
print("Shape     :", m.shape)
print("Rows      :", m.shape[0])
print("Columns   :", m.shape[1])
print("Data type :", m.dtype)

f = np.array(marks, dtype=float)
print(f)

[[67 72 80]
 [88 91 85]
 [54 60 58]
 [75 70 79]]
Shape     : (4, 3)
Rows      : 4
Columns   : 3
Data type : int64
[[67. 72. 80.]
 [88. 91. 85.]
 [54. 60. 58.]
 [75. 70. 79.]]

import numpy as np

lists = [
    [10, 20, 30],
    [1.5, 2, 3],
    [1, 2.5, 'three'],
    [True, False, 1],
]
for L in lists:
    a = np.array(L)
    print(L, "->", a, a.dtype)

[10, 20, 30] -> [10 20 30] int64
[1.5, 2, 3] -> [1.5 2.  3. ] float64
[1, 2.5, 'three'] -> ['1' '2.5' 'three'] <U32
[True, False, 1] -> [1 0 1] int64

import numpy as np

temps = np.array([[31, 33, 35, 34, 30],      # Delhi
                  [28, 29, 29, 30, 31],      # Mumbai
                  [33, 35, 36, 37, 34]])     # Chennai
cities = ["Delhi", "Mumbai", "Chennai"]

for i in range(3):
    row = temps[i]
    print(cities[i], "max:", row.max(), "min:", row.min(),
          "avg:", row.mean())
print("Hottest reading overall:", temps.max())
print("Day-wise average:", temps.mean(axis=0).round(2))

Delhi max: 35 min: 30 avg: 32.6
Mumbai max: 31 min: 28 avg: 29.4
Chennai max: 37 min: 33 avg: 35.0
Hottest reading overall: 37
Day-wise average: [30.67 32.33 33.33 33.67 31.67]

import numpy as np

a = np.arange(1, 13)
print("Original :", a)
grid = a.reshape(3, 4)
print(grid)
print("Second row        :", grid[1])
print("Last column       :", grid[:, -1])
print("Middle 2x2 block  :")
print(grid[0:2, 1:3])
print("Transpose shape   :", grid.T.shape)

Original : [ 1  2  3  4  5  6  7  8  9 10 11 12]
[[ 1  2  3  4]
 [ 5  6  7  8]
 [ 9 10 11 12]]
Second row        : [5 6 7 8]
Last column       : [ 4  8 12]
Middle 2x2 block  :
[[2 3]
 [6 7]]
Transpose shape   : (4, 3)

Common mistakes

• Writing import Numpy as np or import NumPy. Correction: the module name is lowercase: import numpy as np.

• Passing values separately: np.array(1, 2, 3). Correction: pass one list: np.array([1, 2, 3]).

• Expecting [1, 2.5, 3] to stay a mix of int and float. Correction: all elements become float: [1. 2.5 3. ].

• Assuming dtype=int rounds 7.8 to 8. Correction: conversion to int truncates, so 7.8 becomes 7.

• Creating a 2-D array from rows of different lengths. Correction: every inner list must have the same number of elements.

• Thinking list * 2 and array * 2 do the same thing. Correction: the list is repeated; each array element is multiplied by 2.

• Writing a.shape() with brackets. Correction: shape, ndim, size and dtype are attributes, written without brackets.

• Writing the printed array with commas, like a list. Correction: a printed NumPy array separates its elements with spaces: [10 20 30].